TECHNOLOGY STRATEGY: THE CASE OF THE
DIAGNOSTIC ULTRASOUND INDUSTRY
by
JOHN H. YRIAR,
III
A.B., Harvard College
(1975)
M.B.A., Harvard Business School
(1979)
SUBMITTED TO THE
ALFRED P. SLOAN SCHOOL OF MANAGEMENT
IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
September, 1986
John H. Friar, 1986
The author hereby grants to M.I.T. permission to reproduce and to
distribute copies of this thesis document in whole or in part.
Signature of Author
redacted
Signature
Sr
UI
Alfred P. Sloan School of Management
September 29, 1986
Signature redacted
Certified by
Edward B. Roberts
Thesis Supervisor
Accepted by
Arnold D. Barnett
Chairman, Doctoral Program Committee
ARCHIVES
MASS. INST. eCH
OCT 3 1 1986
R
IES
M
n[Libraves
Document Services
Room 14-0551
77 Massachusetts Avenue
Cambridge, MA 02139
Ph: 617.253.2800
Email: docs@mit.edu
http://libraries.mit.edu/docs
DISCLAIMER OF QUALITY
Due to the condition of the original material, there are unavoidable
flaws in this reproduction. We have made every effort possible to
provide you with the best copy available. If you are dissatisfied with
this product and find it unusable, please contact Document Services as
soon as possible.
Thank you.
Some pages in the original document contain pictures,
graphics, or text that is illegible.
-2-
TECHNOLOGY STRATEGY: THE CASE OF THE
DIAGNOSTIC ULTRASOUND INDUSTRY
by
JOHN H. FRIAR, III
Submitted to the Alfred P. Sloan School of Management
in partial fulfillment of the requirements
for the Degree of Doctor of Philosophy
in Management
ABSTRACT
This thesis has focused on the strength of product innovation
as a competitive weapon for the innovating firm.
Several authors
have avered that innovation is the strongest and most direct way to
achieve advantage, but studies trying to analyze such a relationship
have generated conflicting results. It was argued that the reason
for the conflicting results is the lack of inclusion of the marketing
perspective --the analysis of the reactions of potential customers to
innovation-- in the studies. A framework was presented that included
both the customers' ability to perceive technology differentiation
and the defensibility of such innovation.
It was hypothesized that
only in rather specific instances will a technological advance lead
to a viable competitive advantage.
In the other cases, a viable
competitive position can only arise through price competition or
differentiation created from capabilities in other functional areas.
Although some authors have posited that a firm can have only one
functional strength, the relationship of the alternative functional
capabilities to each other was further tested.
The diagnostic ultrasound industry was selected for study
because it epitomizes intense technological competition. Some of its
characteristics are: high levels of R&D expenditures, many new
product introductions, and shifting market shares.
If innovation
leads directly to competitive advantage, this relationship should be
demonstrable in this quintessential high-tech industry.
The data
reveal, however, not a direct link but rather a changeful
entwinement, a pavane that is intricate, delicate, and complex.
Data were collected through two surveys and interviews with
fifty-three companies. Detailed information on market conduct
variables and market performance variables was collected on a subset
of nine of the companies for the five year period ending in 1983.
Analyses were performed on two levels: at the industry level and at
the specific modality-application level.
The lesson gleaned from analyzing the diagnostic ultrasound
industry is that a singular focus on developing new technology must
be augmented by a richer array of strategic choices such as
technology acquisition and strategic linkage options. A stronger
-3interplay of the various functional strategies must occur and is
determined by the customers' perceptions of technology. Once the
marketing perspective is brought into play, the viability of a
technology thrust can be determined. Because of the physicians'
inability to differentiate products on technology criteria, in this
technologically-sensitive industry, technological advance has been
necessary but not sufficient for market success.
Thesis Supervisor:
Committee:
Edward B. Roberts
David Sarnoff Professor of Management
Mel Horwitch, Associate Professor of Strategy and Policy
Steven H. Star, Senior Lecturer in Marketing
-4-
ACKNOWLEDGMENTS
Two years ago I was poised to start down a path I had not
previously ventured and began a journey that thankfully I will not
repeat -- the doctoral dissertation. Like a knight errant at the
edge of a dark, dark forest to begin a mission, 1 doubted the
significance of the quest and feared getting lost in the woods.
I
stepped from the redoubt and peered out to see none of the trees
emblazed nor the landmarks apparent.
Luckily, there were generous and talented people who helped me
reach the final destination of this completed work. As Wart told
King Pellinore, who had become lost in the woods, "I know what
fewmets are. They are the droppings of the beast pursued. The
harborer keeps them in his horn, to show his master, and can tell by
them whether it is a warrantable beast or otherwise, and what state
it is in." Unlike Wart, I did not know a fewmet from a garous
excretion or a rancid and olidous separation. I needed occasionally
to be pointed in the right direction to ensure that I was
progressing.
The people who most directly influenced the direction of this
study were the members of my committee: Professors Ed Roberts,
Mel Horwitch, and Steve Star.
Their receptiveness to different ideas
and their constructive criticisms of my work not only helped me but
also demonstrated their considerable talents. For their guidance I
am thankful.
Faculty members who were not on my committee also provided help
and encouragement. They gave freely of their time, even though they
were under their own pressures, because they are truly concerned
teachers.
I am grateful to Professors Lisa Lynch, N. Venkatraman,
and Gordon Walker for their additional guidance on matters of
methodology and statistics.
The hot houses for new research ideas among doctoral students
are the overly crowded student offices and the over-priced bars
nearby.
The discussions in these settings occasionally touched on
academic research but more often concerned other inspirations.
I
acknowledge the contributions that the many inmates of the
institution made to my overall sanity at these times, but I would
especially like to thank John Chalykoff, Oscar Hauptman, and Nitin
Nohria for their research suggestions and help.
Although I was provided great inhouse support, the most
essential ingredient for this study was the data provided by the
participating companies.
Without their help, the fascinating story
of their industry could never have been told.
I am especially
thankful to the nine companies (Acuson, ATL, Diasonics, General
Electric, Hewlett-Packard, Philips, Picker, Siemens, and Toshiba) who
trusted me by opening their books and spending many hours discussing
their companies and their industry.
I only hope that I have captured
the flavor of the industry as they taste it.
-5My special thanks go to the editors of Diagnostic Imaging and
to the radiology staff at the University of Connecticut. D.I. opened
its archives for my perusal, which provided information on the early
participants in the industry. The University of Connecticut
radiologists, especially Dr. Mary Friar, also made available to me
their library on ultrasound procedures and equipment.
For the production of this paper, I must credit Cheryl Kelliher
and Lenner Laval for their hard work in typing and patience in
accepting changes. For quiet space in which to write, I thank
Jan Austin who found empty faculty offices for me to use.
Finally, I would like to thank
and for keeping me distracted. This
the Greek chorus of family, friends,
chanted a strophe of "Is it finished
and my thanks to you all.
Eva Guinan for making me smile
work is dedicated to her. To
and others who constantly
yet?", I can respond that it is
-6-
CONTENTS
Page
Chapter 1:
Introduction
Prologue.................................................
Research Issues..........................................
Plan of This Dissertation................................
10
12
16
Chapter 2:
Literature Review
Overview .................................................
Economics ................................................
Management of Innovation.................................
Diffusion Theory.........................................
Marketing ................................................
Strategy .................................................
Summary of First Five Fields.............................
Technology Strategy......................................
18
22
24
26
27
29
31
32
Chapter 3: Framework and Hypotheses
Framework ................................................
Hypotheses ...............................................
Choice of Analyses.......................................
42
51
55
Chapter 4: Data Collection
Choice of Industry.......................................
Participants .............................................
Surveys ..................................................
58
62
65
Chapter 5: Understanding Ultrasound
Ultrasound Technology....................................
Applications.............................................
Ultrasound Development...................................
Government Agencies......................................
Image Quality............................................
Selection Process........................................
Judging Image Quality....................................
The Industry.............................................
71
79
82
90
93
97
99
102
Chapter 6: Analyses
Overview .................................................
Licensing and Size.......................................
Major Improvement Analysis...............................
Technology Cycles........................................
Competitive Orientation..................................
Technology-Market Analyses...............................
119
119
129
134
149
154
Chapter 7: Conclusions
Discussion of Findings...................................
Further Research.........................................
Managerial Implications..................................
171
177
180
References
182
-7-
Page
Appendices... ...................................................
193
Appendix 1:
Company Profiles ...........................
194
Appendix 2:
Companies Participating in Study ..........
205
Appendix 3:
Indepth Survey ............................
209
Appendix 4:
General Survey ............................
217
Appendix 5:
Sources Used in Determining Segment and ...
221
Individual Company Sales
Appendix 6:
Additional Clustering Information .........
223
Appendix 7:
Computation and Reliability of ............
index Scales
225
Tests for Nomality of Variables and .......
233
Appendix 8:
Randomness of Residuals
Appendix 9:
Squared Semipartial Correlations ..........
236
-8-
List of Figures
2.1
Patterns of Industrial Innovation
2.2
Transilience Map
3.1
Possible Orientations of Firms as to Strategic Thrusts
5.1
Basic Elements of an Ultrasound System
5.2
Samples of Ultrasound Images
5.3
Science, Technology, and the Utilization of Their Products,
Showing the Normal Progression From One to the Other
5.4
Science, Technology, and the Utilization of Their Products,
Showing Communication Paths Among the Three Streams
5.5
Proportion of Companies That License In Technology
5.6
Cumulative Penetration of Ultrasound in U.S. Hospitals
5.7
Sales of Medical Ultrasound Equipment
5.8
Proportion of Freestanding Medical Ultrasound Companies
5.9
Sales Distribution of Firms
6.1
Medical Ultrasound Sales by Modality Type ($ million)
6.2
Medical Ultrasound Sales by Modality Type (units)
6.3
Medical Ultrasound Sales by Application Type
6.4
Market Share Positioning of the Groups of Companies Selling
Each Modality Type
6.5
Cumulative Diffusion of Real-Time Abdominal Imaging Capability
in 400 to 499 Bed Hospitals
-9-
List of Exhibits
3.1
4.1
4.2
4.3
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
5.9
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
6.10
6.11
6.12
6.13
6.14
6.15
6.16
6.17
6.18
6.19
Hypothesized Returns for a Technology Advance
Analysis of the Pool of Companies to Contact
Companies Selected for Detailed Study
Response Rate
Medical Diagnostic Ultrasound Technology Types
Preferred Initial Imaging Examinations
Medical Specialties That Use Diagnostic Ultrasound Equipment
Statist-.on R&D Expenditures for Ultrasound Firms
Leading Sellers of Medical Diagnostic Ultrasound in the U.S.
Statistics on the Age of Ultrasound Firms in the U.S.
Comparison of Large and Small Firms as to Years in Market
Foreign Companies That are Presently Competing in U.S. Medical
Ultrasound
Reasons for Entering the U.S. Ultrasound Market
Large Firm Use of External Technology Acquisition
Smaller Firms That Have Used External Technology Acquisition
Relationship Between Status as a Subsidiary of a Larger Firm
and Ultrasound Sales
Headquarters Locations for Ultrasound Divisions of Full-Line
Medical Equipment Companies
Relationship Between Size of Firm and Use of Licensing
Relationship Between Status as a Subsidiary of a Large Firm and
Licensing In
Technology Development Milestones in Diagnostic Ultrasound
Benefits From Commercialization of Major Innovations
Ability of Firms to Gain a 10 Percent Share in the Major
Technology-Application Segments
Original Application, Accepted Application, Time to Acceptance,
and Time to Peak Sales
Cluster Analysis of Relative Emphasis on R&D, Marketing, and
Price
Self-Evoked Orientation Clusters and Performance Measures
Rankings of Image Quality for a Company's Equipment in
Different Segments
Modality and Application Combinations Used in the Study
The Relationship Between Image Quality Rankings and Market
Share Rankings
Strategic Variables Used in Creating Indices of Competitive
Orientation
Relationship of Advertising to R&D Over 5 Years
Regression of Performance on Indices of Competitive Orientation
Paired Comparisons of Strategic Orientation of Overtaking Firm
to Unseated Firm
-10-
CHAPTER 1
INTRODUCTION
Prologue
Vast. The exhibitionhall was immense and yet every available square
inch of floor space had been taken by the 333 exhibitors. The exhibition
committee had limited the amount of space an exhibitorcould rent, so the
competition for space spilled to the vertical. Clean lines of chrome and neon
archedhigher and higher -- the goal now was to have one's company logo
towering above the rest. Because of the intense competition, the exhibition
committee either had to raise the building'sroof or limit the height of company
logotypes. The exhibits themselves were worthy of a place in a world's fair.
Two stories high with special meeting and hospitalityrooms hidden behind
plants and multimedia displays touting the full rangeof each company's
product lines. Forty or more smartly dressedsalespeople, well rehearsedin
their promotions, swarmed the deep-pile carpetingof each exhibit in hopes of
generatingactivity.
McCormick Place in Chicago was again hosting the annual meeting of the
RadiologicalSociety of North A merica, a misnomer because radiologistsand
techniciansfrom aroundthe world attend the scientificpresentationsand the
accompanyingtrade show. Not only do medical diagnosticequipment
companies spend tens or hundredsof thousands of dollarson exhibits, but a
like amount on receptions, entertainment, and dinners for the doctors.
The pressure to announce an advancement, an innovation, something new
is tremendous for each company, not only in general but for each product line.
If nothing but last year's equipment is shown, a company is considered
incapableof generatingany customer activity. The trade show is an orgy of new
-11-
product announcementsand professions of technologicalfirsts. One can be
easily diverted for several days by the clamor of so many companies trying to
outdo each other.
On the lower level of McCormick Place, away from the din and glamour of
the main hall, among the publishersand in front of the technical exhibits, stood
a solitary salesman and his one machine. The space was the minimum one
could rent, with no carpet over the yellowed tile floor. His exhibit consisted of a
card table piled with literature,a folding chair, and his machine. The company
logo was the cardboardsign the exhibition committee had provided to mark
each space.
I had spent the previous six months trying to number all the diagnostic
ultrasoundcompanies in the United States. I had identified over 120
companies, but I now had found one more. The company consisted of several
engineers working on a shoe string, who had developed an improved diagnostic
ultrasoundproduct. I asked about the daunting competition, the lack of a direct
sales force, the nonexistent advertising,and the general lack of profitabilityin
the industry.
He was not worried. "Image quality is all that is important. Doctors need
improved images to enhance their ability to make diagnoses. We have improved
substantiallythe image quality in our machine over the competition and so our
product will sell."
The high-tech entrepreneur-- I had found anotherone. Someone who was
attemptingthrough technologicalinnovation to create a growing company. I
did not ask, but I am sure that he hoped that his company would soon rank with
Apple, Hewlett-Packard,Wang and the myriad of other high-tech success
stories.
-12-
Research Issues
Whether or not one gives credit, as Quinn (1979) does, to the
individual inventor-entrepreneur for generating the majority of
innovations in western societies, innovation has become a central
theme for management.
The U.S. is relying on innovation in high-tech
industries to keep American firms competitive in world markets.
Innovation, moreover, has risen to the fore in the popular business
press as the key to success for the individual firm.
As an example,
Peters and Austin (1985) aver that one is conveyed down the path to
riches by constant innovation.
This thesis focuses on the relationship between innovation and
competitive strategy.
Research studies1 have illustrated the
importance of innovation for growth in economies and in industries.
On the micro or firm level,
innovation by others has been shown to be
of potential harm to the non-innovating firm.
The relationship,
however, between creating a technological advantage and creating a
competitive advantage for the innovating firm remains unclear.
The
difficulties arising from not innovating have been well documented,
but the advantages for a firm that has innovated have not been
empirically studied.
Many analysts, 2 when describing the continuing worsening of
the U.S. trade balance, lay the blame for the diminution of American
competitiveness on the neglect by U.S. firms of using product and
process innovations as competitive weapons.
These authors posit that
if U.S. firms, among other things, increase their spending on and
focus their concentration toward research and development, that
American competitiveness will rebound.
-13Other authors, 3
when analyzing competition in
technology-based industries, have relied on "the obvious" to state
that innovation can lead to increased sales and market share for a
firm.
After analyzing the process of innovation, such studies
conclude that improving the innovative process is a surefire way to
improve competitive position.
U.S. firms have responded to the increasingly competitive
environment of global trade by increasing their outlays for R&D.
Recent figures from the National Science Foundation and
McGraw-Hill 4 have illuminated that real average annual R&D spending
growth has surged since 1983 to greater than 8 percent, which is a
far cry from the miniscule 0.8 percent average from 1963 to 1978.
The link between increasing R&D expenditures and gaining
competitive advantage is not straightforward, however.
The strategy
literature is just beginning to come to grips with innovation and its
effect on
competitive position.
Horwitch (1983) discusses the
subsidiary position technology has played in the corporate strategy
literature but argues that its role is being elevated.
Porter (1983)
also argues that the study of competition has been decoupled from the
study of technological innovation so that one can merely discuss
conceptual links.
Corporate strategy, therefore, has mainly focused
on technology portfolios and on timing of entry.
When the strategy literature has attempted to couple innovation
and competitive position, innovation has been so broadly defined as
to be meaningless.
Porter (1985), in his attempt at tying innovation
to competitive advantage, defined technological innovation as being
anything new to the firm, whether it was a new secretarial procedure
or a new product or a new manufacturing process.
Such broadly
-14-
defined innovation, moreover, is said to improve a firm's value added
chain and therefore give it a competitive advantage.
We propose to use a less broadly defined concept of innovation
because innovation does not come freely to a firm --
a firm must
expend resources to create and to implement innovations.
The
innovations that we are concerned with are those that are derived
from some form of R&D expenditure and that are technological advances
that lead to new or improved products and processes.
The question
then again becomes one of coupling R&D expenditures and advances to
competitive advantage for the innovating firm.
The economics literature claims that firms invest in R&D
because they have expectations of achieving a quasi-monopoly with
accompanying supranormal profits. 5
There is, however, no
absolutely general, unambiguous answer to the shape of the benefits
function for R&D expense.
In fact, "research expenditures which may
generate great benefits to society as a whole may also generate very
small benefits - or perhaps no benefits at all - to the private firm
making such expenditures." 6
In general, then, the relationship
between technological advance and competitive advantage for the
innovating firm is not clear, is not straightforward, and has not
been analyzed.
Innovations lead to both new products and processes, but the
vast majority of new developments are for products.
Scherer (1970)
estimates that 86 percent of R&D developments are for new products.
Yet, little research has been directed at how innovation relates to
competition in a particular product. 7
Instead, most of the
research has dealt with process innovations. 8
The main reason for
-15this is the limitations of the various frameworks used to analyze
product innovation and its effect on competitive position.
Not only are most technological innovations product
innovations, they are also incremental 9 and defensive 1 0 in
nature.
are rare.
Industry shattering innovations in the Schumpeterian sense
If technological competition is more properly described by
continual, minor product advancements, then these are the types of
innovations that should be studied.
In the quest for competitive advantage through product
innovation, the success of a strategy is bound tightly to the
reactions of potential customers.
The assessment of market response
is in the domain of marketing strategy, yet the marketing strategy
literature has been developed primarily by nonmarketers.1 1
Because
the marketing framework has not yet become fully ingrained in
marketing strategy, the area is in need of further development.
The
assessment of the viability of a strategy to create competitive
advantage is in the domain of strategy content, an area that is also
in need of development. 12
The research issues that this paper deals with start with the
question of when a technological advance becomes a competitive
advantage for the innovating firm.
Although the connection between
innovation and competitive advantage might appear to be straightforward, few studies on the relationship exist and what do provide
ambiguous answers.
We have also argued that incremental product
innovations studied from the marketing and strategy content
perspectives should be analyzed.
Both marketing strategy and
strategy content need to be further developed.
-16-
Plan of This Dissertation
This work is organized in three broad sections.
section consists of Chapters 2 and 3.
The first
These chapters describe the
research from several disciplines that are germane to the study of
technology strategy or the use of innovation as a competitive
weapon.
The purpose of the literature review is to show that these
studies, although important in their respective fields, provide a
fragmented picture when applied to the question of innovation and
competition.
From this broad literature review, a framework for
study and a set of hypotheses are developed.
The second section consists of Chapters 4 and 5.
These
chapters provide information on how the data were collected and give
the background needed for understanding competition in the medical
diagnostic ultrasound industry.
The ultrasound industry is dynamic,
changing, and intensely competitive.
This competitive environment is
documented so that the reader is prepared to understand the final
section.
The third section consists of Chapters 6 and 7.
These chapters
describe the analyses that were performed to test the hypotheses from
the first section.
The results from the analyses and a discussion of
what these results tell us about technology strategy are contained
therein.
The work concludes with a discussion of topics and
questions for further research.
-17-
CHAPTER ENDNOTES
1.
See for example: Schumpeter (1934), Fellner (1971), Rosenberg
(1972), Lewis (1982), Mansfield (1982), Freeman (1982).
2.
See for example: Rosenbloom and Abernathy (1982), Lawrence and
Dyer (1983), Hayes and Abernathy (1980), Foster (1982), Lewis
(1982).
3.
Gobeli and Rudelius (1985), Peters and Waterman (1983), Foster
(1986).
4.
Business Week, June 16, 1986, p. 24.
5.
Scherer (1970), Comanor (1967), Bain (1968).
6.
Rosenberg (1972).
7.
Abernathy (1978).
8.
Mowery and Rosenberg (1979), Nelson and Winter (1977).
9.
Marquis (1969).
10. Freeman (1982).
11. Wind and Robertson (1983).
12. Rumelt (1979).
-18-
CHAPTER 2
LITERATURE REVIEW
Overview
A review of the literature on whether a technological advantage
provides a competitive advantage for the innovating firm actually
covers several fields of research.
This broad span is due to the
lack of a research paradigm in the strategy field and to the
recentness of the effort to integrate technology into strategy.
Because strategy is multidimensional and emanates from a number of
research streams, a researcher in this field has to provide a road
map of the type of strategy research that is to be performed.
In
this section we first set forth a charting of the strategy and
technology research streams before actually reviewing the literature
on technological advance and competitive advantage.
Strategy is considered to be multidimensional and situational
so that no consensus exists on even a definition of strategy. 1
Schendel and Hofer (1979) write that strategy is a young field with
no strong research tradition, especially one of empirical testing to
determine the domain of applicability and validity of the many
concepts and hypotheses put forward.
Because of this lack of
determinacy, they list 18 broad topic areas for strategy research.
The topic this study is concerned with is strategy content and
evaluation; or for a given type of environmental circumstances, what
types of strategies should a firm follow?
Not only are there many topics of study within strategy, but
also a variety of research streams with varying paradigms and units
of analysis for studying each topic.
Seven disciplines 2 have
-19contributed to strategy research.
Those disciplines have been:
industrial organization and microeconomics, organization theory,
marketing, finance, psychology, sociology, and decision sciences.
Recent papers3
have called for strategy research
cross-fertilization and research stream integration and have
discussed the lack of such research to date.
The research streams
that deal with strategy content are industrial organization economics
and marketing,4 but they approach the problem from different
levels.
In an earlier work, Hofer and Schendel (1978) laid out four
levels of policy analysis and four components of strategy.
levels were:
The
enterprise, corporate, business, and functional.
components were:
The
scope, distinctive competences, competitive
advantage, and synergy.
This paper deals with competitive advantage
at the business level.
The original question posed of when does a technological
advantage lead to a competitive
advantage for the innovating firm is
answered at the business level, competitive advantage, strategy
content confluence of the various levels, components, and topics of
strategy research.
The research stream that should be most helpful
at this intersection is marketing strategy because industrial
organization economics looks at the level of the industry rather than
at the level of the product-market.5
The marketing strategy
literature, however, has a severe handicap.
The marketing field is dominated by marketing management, which
is concerned with the marketing mix.6
Marketing research is
conducted at the brand level rather than at the product category or
business unit levels.
The stream of research that should most
-20-
readily help in illuminating the likely market response to a proposed
strategy in fact provides little research help.
from theories of diffusion.
The main aid comes
Marketing research has provided many
contributions to the strategy literature, but marketing strategy has
8
been developed by nonmarketers.
Because strategy as a whole is so amorphous and because
marketing has not directly addressed long-term competitive advantage,
a literature review has to include studies from economics, marketing,
and diffusion theory.
As will be shown, most of these studies have
dealt with technological advance and competitive advantage at the
wrong levels and without empirical support.
The other side of the technology strategy equation is that of
technology.
Technological innovation is more easily understood as a
sequence of activities, but technology has also been studied at
several levels and from distinct perspectives.
Rosenbloom (1978)
lists five levels at which technology has been studied:
sector,
industry, firm, and innovation.
society,
As with strategy,
researchers from several disciplines have analyzed technology
development.
A few are:
sociologists, historians, anthropologists,
economists, and organizational behaviorists.
Economists have
considered the aspects of the firm's or industry's environment that
spur or retard technical advance.
The other disciplines, however,
have been concerned with improving an organization's ability to
perform R&D and with linking innovation to customer needs.
Research has mainly looked at the conditions, environments, and
determinants of improving the R&D process.
Of principal concern have
been the personality traits of individuals, the flow of technical
-21-
communication, the origins and originators of new ideas, and the
influence that administrative practices and organization structure
have on technology development.
The overriding assumption is that
innovation is good, so no attempt has been made to delineate when
technology advance actually becomes a defensible competitive
advantage.
At the intersection of strategy research and technology
research is technology strategy.
Because strategy and technology are
ambiguous terms, technology strategy must also be ambiguous.
Technology strategy consists of many levels, disciplines, and
components.
Because technology strategy is just emerging as a
significant aspect of strategic management, it has yet to discover an
overarching framework or paradigm for research.
Its breadth is much
wider than innovation and competitive advantage at the firm level.
Much of the research in technology strategy has been at the corporate
and industry levels so there has been little research at the business
unit level.
Strategy is multidimensional and situational, technological
change is multidimensional and situational, and therefore technology
strategy is multidimensional and situational.
By mapping the various
dimensions of each area, some of the research streams have been
pared.
What remains for a literature review are studies in
economics, marketing, diffusion, management of innovation, strategy,
and technology strategy.
The studies, however, will have troubles
with levels of analysis and lack of empirical findings.
In critiquing the following studies, we used as a standard the
amount of information these papers elicited on the relationship
-22-
between technological advance and competitive advantage for the
innovating firm.
The studies cited are obviously important ones in
their respective fields, and they have made contributions to the
specific issue of interest.
However, because the majority of the
studies have not dealt directly with the issue of innovation and
competitive advantage, due to the fact that interest in the topic is
so recent, the contributions they do make are fragmented.
The
purpose of the literature review, therefore, is to illustrate the
present gaps in the research to date for presenting a comprehensive
discussion on the relationship of technology advance to competitive
position for the innovating firm.
The literature review is separated into two sections:
a brief
review of each of the tangential fields and a more detailed review of
the technology strategy literature.
Economics
The economics literature makes a strong case for the benefits
to society of technological advance.
Schumpeter (1934) described the
dynamic behind economic development.
He felt that it is not possible
to explain economic change by previous economic conditions alone.
Economic development arises from changes in knowledge, which lead to
new or improved products and to new processes.
New products and
processes are the results of technological advances, which,
therefore, drive growth in economies, create new industries, and
alter competition within existing industries.
Fellner (1971) in his review of the various methods of
measuring the benefits of technology advancement concludes that it is
-23-
"the main source of the significant rise of the standard of living in
Western economies."
Rosenberg (1972) in his study of nineteenth
century innovation in American industry also concludes that
technology change leads to growth.
Because of the importance of innovation to economic growth,
much of the economic literature then becomes preoccupied with the
effect public policy might have on innovativeness.
Studies analyze
innovation with the size of the firm, with the concentration of the
industry, and with the usefulness of patents and trademarks. 9
The
proper level, of funding for research and the proper agencies
--government,
industry, and academic-- to perform the research have
been analyzed along with the rates of returns to be expected. 1 0
Also, the height of entry barriers and their effect on innovation
have been studied. 1 1
Microeconomics has difficulty dealing with innovation because
innovation is considered to be exogenous to the firm rather than
under its control, and is studied from the perspective of process
innovations and their effect on productivity and market
concentration.
Even then, increasing R&D expenditures gives
ambiguous results across industries as to increasing productivity or
market concentration.
But what of the effects of product technology advancement on
the innovating firm?
The same authors who posit innovation's benefit
to society can only give ambiguous responses to innovation's effect
on the firm by drawing on the theory for the need for patent and
trademark protection.
Companies are felt to innovate to gain
competitive advantage, but the size of the advantage depends upon the
-24-
length of time for competitors to respond.
"Most of the innovative
profits can be captured before competitors can copy them."
12
But
advances may not only be copied, but also leap-frogged by further
advances.
"It is equally important to remember that progress is
strongly promoted by a high degree of competition."13
Competition,
and therefore a rapid rate of progress, will shorten advantage
times.
The slower the spread of knowledge, or the slower the
advancement of knowledge, the greater the appropriability of benefit
to the innovating firm.
Economists, then, believe that innovation advances growth for
the economy and should be of competitive advantage to a firm as long
as the innovation is not too quickly copied or improved upon.
Since
a lead time is the only real advantage to an innovator, "research
expenditures which may generate great benefits to society as a whole
may also generate very small-benefits --or perhaps no benefits at
all-- to the private firm making such expenditures." 1 4
Management of Innovation
The management of innovation literature deals with two
different processes:
the process of technical completion and the
process of commercialization. 15
A firm is assumed to have to
innovate because other firms are innovating.
A company's market
share or very existence can be threatened by another firm's
innovation so that a firm is presumed dead if it cannot generate new
products and processes.
All firms, therefore, are assumed to
need to innovate, but no judgment is made on the need for
technological improvement as opposed to duplication.
Is
-25-
technological advantage really leading to success or is it a
combination of other factors?
The process of technical completion "generates two kinds of
outputs:
new and improved products, new and improved processes." 17
The success or failure of this process, then, is judged by the
successful completion of a new product or process.
Studies on
improving the innovation process within a firm use the successful
completion of specifications18 or the judgment of managers as to
how technical groups are performing
as the criteria for success.
The process is judged, then, without regard to the competitive
advantages of an innovation.
The other process In the management of innovation literature is
that of commercialization.
Retrospective studies analyze the success
or failure of new products in the market place.
SAPPHO study. 2 0
An example is the
The study's main conclusion is that success comes
from understanding the market.
"Successful attempts were
distinguished frequently from failure by greater attention to the
education of the users, to publicity, to market forecasting and
selling and to the understanding of user requirements." 2 1
Understanding users' needs, however, encompasses much more than
just the functional performance and the embodied technology in a
piece of equipment.
Product image, for example, may be more
important than product performance in meeting users' needs.
Just
because a new product or process does well in meeting user
requirements, one cannot infer, therefore, that it does so because of
a technological advantage.
-26-
Retrospective studies of successful commercializations compared
product successes to product failures in order to analyze significant
differences.
Quality of technology was not a factor in the analyses
of product successes but was a factor in product failures.
The
SAPPHO study and another analysis of commercial failures showed that
some products failed in commercialization because a competitor
introduced a product with superior technology.
Inferior technology
caused failure 11.5 percent22 and 20 percent23 of the time for
products that did fail.
The management of innovation literature, therefore, suggests
that innovation is important for survival and that inferior
technology may lead to failure.
The literature, however, does not
analyze whether a technological advantage leads to commercial
success.
Neither analyzing the successful completion of
specifications nor analyzing the differences between successful and
unsuccessful commercialization allows any statements to be made about
technological advantage and competitive advantage.
Diffusion Theory
Diffusion studies have a tendency to look more at the adopters
than at the innovation in analyzing the rate of adoption of an
innovation.
Rogers (1971), however, in his review of studies that do
analyze differences in attributes of innovations consistently finds
that relative advantage is rated first or second as a determinant of
the rate of adoption.
Relative advantage is the degree to which an innovation is
perceived as better than the practice that it supersedes.24
A
-27-
point to emphasize is that perceived advantage and not objective
advantage is what is positively related to the rate of adoption.
Diffusion studies, also, look at innovations more as an idea or
practice than as a specific product, e.g., solar heating rather than
Company XYZ's solar heating product.
Industries or product-market
segments, then, rather than individual firms are analyzed.
Rates of
diffusion of an innovation have been found to vary by the number of
firms in an industry25 and by the capacity and profitability of an
industry. 2 6
None of the diffusion studies explicitly analyzes technological
advantage and competitive advantage (improved diffusion) for the
firm.
One could easily infer, however, that the importance of
relative advantage for an industry holds for technological advantage
and the firm.
Especially in a high-technology industry, a real
technological advantage should lead to a real relative advantage and
then to a perceived advantage.
A perceived relative advantage is
positively related to faster diffusion, which leads to greater market
share and short-term profits. 2 7
Marketing
Marketing does not really address the concept of a
technological advantage because customer value is determined by
evoked preferences. To improve the quality of a product, a marketer
must improve a product's position on a perceptual map.28
As with
diffusion theory, quality or relative advantage is perceptual and not
necessarily real.
Although marketing theory envelops all products,
the empirical work has been done with consumer products.
The
-28-
extension to high-technology products, therefore, can only be
surmised.
Marketing research also has difficulty with technological
advances when customer preferences have not yet been determined or
are changing, and when the consumer is required to adopt new behavior
patterns.
Marketing theory assumes an essentially stable, continuous
environment. 29
Quality, in marketing theory, is determined by consumers after
they have experienced a product.
By ascertaining consumer
preferences and by performing factor analyses, the determinants of
product quality are established.
In a high-technology industry, one
could easily assume that improved technology should lead to
improvements in one or all of the factors of quality.
Improved product positioning is important, as Urban (1984) has
shown, because success is determined in a consumer market by either
being first into a market or by coming in later with a betterpositioned product.
Pioneering brands are felt to gain advantages in
reduced customer uncertainty, stable preference patterns, lower
production costs, and raised barriers to entry.
These advantages are
long-term and may accrue without patent protection or long lags in
imitation.
To overcome the advantages of being first to a market,
Urban showed that a later entrant must have a superior product; a
parity product will achieve less market share than a pioneering
brand.
A technological advantage, then, should be important for two
reasons.
If a firm can-use a technological advantage to create a new
market, it will accrue the benefits of a pioneering brand.
If a firm
-29-
can improve product positioning by improving technology, it can
garner market share in an established market.
Although the marketing
literature does not analyze the implications of a technological
advantage directly, such an advantage should lead to a marketing
competitive advantage.
Although marketing research has had difficulties in analyzing
the impact of a technological innovation, Corey (1976) still felt
confident enough to say,
"Technical research has been a competitive
weapon surpassing in its power such traditional weapons as price
cutting and increased advertising."
Strategy
The strategy literature agrees with the economics literature in
that innovation can create, transform, or destroy industries and
therefore alter competitive positions. 3 0
Since technology can have
such a substantial impact on a business, a firm should maintain a
portfolio of technical capabilities.31
The technology portfolio,
however, is still analyzed separately from a business portfolio.
Strategists maintain, then, that technology development must be
linked to the other organizational areas within a firm, e.g.,
marketing and manufacturing. 3 2
Once a firm, however, has developed
a portfolio of technical capabilities and has linkages throughout the
organization, it must revert back to the generic competitive
strategies open to it before.
Porter (1980) defines three generic strategic approaches a firm
can take:
1.
overall cost leadership
-30-
-
2.
differentiation
3.
focus
Since the focus approach entails using either of the first two
approaches on a select market segment, there are really only two
fundamentally different strategies.
Technological innovation can be
used, then, to either improve one's cost position or to
differentiate.
Differentiation would entail using a technology to
create new products or to serve new customer groups. 3 3
The final strategic decision left to a firm that is
implementing a generic strategy with new technology is whether to be
offensive or defensive.3 4
An offensive strategy is to be first
while a defensive strategy is to enter later.
A later entrant can
lag behind, match, or improve on the leader's product.
Several studies have looked at the significance of entry
timing.
Freeman's (1982) review of surveys analyzing industrial R&D
in several countries concludes that most R&D is defensive rather than
offensive.
The SAPPHO study, moreover, pointed out that in medical
instruments the later entrants performed better in commercializing
their products.
The question still remains, however, of whether the
later entrants were competing with improved technology or merely
copied technology?
No studies address this question directly, but one can infer
from other studies.
The marketing studies, as shown, would suggest
that a defensive strategy still requires an Improved product.
A
corroboration of that idea would be Porter's (1979) study of large
and small firms.
In medical goods, smaller firms were found to have
rates of return much higher than the larger firms.
One could
-31-
hypothesize that since a smaller firm cannot match advertising,
promotion, and sales expenditures with a larger firm, nor compete on
price because of cost disadvantages due to scale; that the smaller
firms must be competing with improved products.
In a high-technology
industry an improved product is probably due to improved technology,
so a technological advantage does provide a competitive advantage.
The corporate strategy literature, in summary, has not
considered explicitly whether a technical advantage provides a
competitive one.
The preliminary conceptual work states that a firm
must have a range of technical capabilities and organizational
linkages.
The generic strategies do not change with innovation, but
timing of entry may matter.
In several empirical studies, later
entrants were found to perform better; the reason for the better
performance, however, Is still unknown.
Although there are no empirical studies to substantiate the
reasons for better performance, Porter claims that technology is the
driving force.
"In industries where technological change is rapid or
the level of technological sophistication is high, the technological
dimensions of competitive strategy can be the primary source of
competitive advantage in the generic strategy being followed by the
firm."35
Summary of First Five Fields
The five fields of research just reviewed all concur that
innovation is important for a firm to consider.
Existing firms may
be seriously affected by another firm's technological advantage, and
firms should benefit from innovation.
Although the failure side is
-32-
well documented, the benefit side is not.
Economists and diffusion
theorists believe that a technological advantage provides a
short-lived competitive advantage while marketers and innovation
management theorists believe that it has long-term effects.
Strategists believe that a technological advantage can be the most
important competitive advantage.
The above statements are
suppositions, however, with little empirical work to substantiate
them.
Technology Strategy
The development of the literature on technology strategy is a
recent phenomenon, even though papers first calling for the inclusion
of technology within the framework of strategy are over twelve years
old. 3 6
The development of the concept of technology strategy for
the firm has occurred from two distinct perspectives and can cause
some confusion on what is meant by technology strategy.
One perspective is the strategy of innovation management, which
is really strategy at the functional level of R&D management.
Such
studies are concerned with the individual variables that affect
innovation and deal with internal technology development and external
technology acquisition.37
Studies of this type include technology
portfolios, technology planning, S-curve analysis, make versus buy
decisions, and the locus of innovation.
The other perspective of technology strategy analyzes the use
of technology as a competitive weapon or as the main thrust of a
business unit plan.
These studies are concerned with the
relationship of R&D to the other functional areas and attempt to
-33-
match company performance with the level of R&D intensity.
This
perspective of technology strategy is more properly considered a
business strategy with innovation as the distinctive competence.
The former perspective has been discussed in the review on the
management of innovation; the latter perspective is the topic here.
The cry for the inclusion of innovation within strategy has been
followed by several studies that have shown that companies are doing
just that. 3 8
Bean et al. (1984), however, have argued that R&D's
involvement in strategy formation has largely been a matter of
managerial philosophy rather than substantive contribution.
The
reason for the lack of substantive contribution is that technology
strategy is still looking for a framework that will provide fewer
conflicting correlations to performance.
To date, technology strategy has mainly consisted of taxonomies
and frameworks, which have been untested, are descriptive rather than
normative, or provide conflicting results.
These frameworks can be
divided into five broad categories:
1.)
entry taxonomies
2.)
demand-pull versus technology-push
3.)
descriptive models
4.)
distinctive competences
5.)
technology dimensions versus market dimensions
Entry taxonomies are characteristic of studies by:
Ansoff and
Stewart (1967), Porter (1983), Miles and Snow (1978), Maldique and
Patch (1978), and Freeman (1982).
Common to the taxonomies are
titles such as pioneers, defenders, followers, and niche seekers.
-34-
Two difficulties with such taxonomies, however, are intent versus
reality and correlations to performance.
A company may have the
intention of being first to market but in reality may end up a
follower.
What actually happened gets studied as opposed to what the
company planned to do.
When the taxonomies are used to predict
performance, in some industries offensive strategies do well while in
others defensive strategies outperform the offensive ones. 3 9
The
reasons on when to use any of the taxonomies as a strategy,
therefore, have not yet been developed.
Many studies have analyzed the dichotomy between using a
demand-pull versus a technology-push strategy.
Utterback (1974).
Although demand-pull is considered to be the more
likely to succeed, technology-push
pay-off.40
For a good review see
is considered to have the higher
Mowery and Rosenberg (1979), moreover, in their review
of these studies argue that the definition of demand-pull was so
broad as to be meaningless.
Voss (1984) reinforces this view in his
study of customer-active innovation by arguing that this form of
innovation is really technology-push and not demand-pull.
The main descriptive models of technology strategy are the life
cycle models of innovation and the transilience map.
Life cycle
models describe the changes in innovation that take place over the
product life cycle and the production life cycle.41
The most well
known of these is probably that of Abernathy and Utterback (1978).
This model, which describes the changing character of innovation,
provides guidance to a firm as to what stage an industry is in.
But
because the abcissa of the model is never labelled, the trend is not
considered to be unidirectional.
A firm, then, is not given any
-35guidance on how to change the pattern of competition or in what
direction the pattern might actually be heading (Figure 2.1).
The transilience map by Abernathy and Clark (1985) plots
innovations as to their effects on markets and production systems
(Figure 2.2).
As with life cycle models, the transilience map
describes a competitive environment but provides no guidance as to
when an innovation in one of the quadrants will be successful or
not.
No work or even hypotheses on which quadrant one should strive
for have been developed, just guidance for what one should do within
a given
quadrant.
What is worrisome is that Abernathy and Clark
hypothesize that firms cannot move from one quadrant to another even
though the competitive environment may require different kinds of
innovation at the same time. 4 2
The distinctive competences literature is characterized by
studies that analyze companies competing In technology-based
industries.
Firm performance is explained by the differences in
internal capabilities.
Studies representative of this approach are:
Rothwell et al. (1974), Cooper (1984), Phillips et al. (1983), Tassey
(1983), Hitt et al.
and Wagner
(1982), Snow and Hrebiniak (1980), and Mansfield
(1975).
The difficulty with these studies is twofold:
competencies
that lead to technical completion may hinder economic return, and
competencies are situation specific.
Mansfield and Wagner separate
out three risks for successful innovation:
commercialization, and economic return.
technical completion,
They demonstrated that some
variables had directly opposite results for improving
commercialization versus economic return.
These variables were:
Figure 2.1:
Patterns of Industrial Innovation
Rate of
Major
Innovation
Process
Innovation
Product
Innovation
Source:
Abernathy and Utterback (1978)
Figure 2.2:
Transilience Map
disrupt existing/create
new linkages
Niche Creation
0
0,
'U
C
0
0,
U
Architectural
N
Ford
Model T
(1908)
0
U
'U
Ford Model A
(1927)
I
Technology/Production
di srupt/obsolete
ex isting competence
conserve/entrench
existing competence
Ford V-8 Engine
(1932)
U
Electric Starter
(1912)
closed steel
body
(1923)
U
Lacquer Painting
System
(1923)
RegularZ
conserve entrench
existing linkagesI
Transilience map and selected automotive innovations.
Source:
Abernathy and Clark (1985)
Rev olutionary
-38-
project ideas from marketing, quantitative project selection, and
ambitious technical projects.
The other studies did not
differentiate between commercialization and economic success.
If economic success is the criterion for comparing capabilities
of firms, the studies differ as to the following dimensions:
1)
Marketing research
2)
Venturesome projects
3)
Ideas from marketing
4)
Offensive approach
5)
Focus in markets and technologies
6)
Demand-pull
Mansfield, Rothwell, and Cooper recommend strong marketing
research while Snow and Hitt do not.
Mansfield and Cooper suggest
venturesome technical projects while Rothwell is mixed.
Cooper is
for the marketing department generating project ideas while Mansfield
is not.
Cooper recommends offensive strategies, which was discussed
before, and no focus on market segments, which will be discussed
later.
Finally, Rothwell found demand-pull to be significant while
Mansfield did not.
The last of the five broad categories is the comparison of
technology characteristics to market characteristics using the firm
as the orienting point.
These frameworks have focused on new
business development, with the majority concerning diversification
strategies.
For an extensive review of this literature see the
discussion in Meyer (1986).
The best known of these writings is
probably Abell's technology-application-function cube which helps a
firm to define itself and its competition.
-39-
The studies that have tried to empirically test different newbusiness strategies (Roberts and Berry (1985), Meyer (1986)) have
used a framework of the newness of the market and technology to the
firm.
Firms trying to move away from their core technologies and
core markets have had difficulty.
These studies suggest that firms
moving away from their established technical base and/or trying to
enter new markets face lower economic returns.
This finding is
counter to Cooper's mentioned previously.
The difficulty with these studies for linking technological
advance and competitive advantage is that the point of reference is
newness to the firm and not to the customer.
These studies are not
ones of competition in a market using one's strength, but of moving
away from such competition.
Technology strategy, in summary, is not only confusing because
it is somewhat amorphous due to the various research streams, points
of reference, levels of analysis, and frameworks; but also due to the
conflicting results of many of the studies.
The reason many of the
studies provide conflicting results is that technology strategy is
situation specific and these studies have not yet captured the
correct situational parameters in the various frameworks.
In the next chapter, an attempt is made at defining what are
the important situational parameters to consider.
From this
discussion a framework and a set of hypotheses are developed for
further testing.
-40-
CHAPTER ENDNOTES
1.
Hambrick (1983), Chaffee (1985).
2.
Thomas and Venkatraman (1985).
3.
Jemison (1981), Schendel and Hofer (1979a).
4.
Jemison (1981).
5.
Wind and Robertson (1983), O'Shaughnessy (1984), Biggadike
(1981), Jemison (1981).
6.
Wind and Robertson (1983).
7.
Rao (1982).
8.
Wind and Robertson (1983).
9.
Freeman (1982).
10. Lewis (1982).
11. Caves (1972).
12. Caves (1972).
13. Fellner (1971).
14. Rosenberg (1972).
15. Mansfield and Wagner (1975).
16. Freeman (1982).
17. Roberts (1981).
18. Allen (1977).
19. Katz and Tushman (1981).
20. Rothwell (1974).
21. Freeman (1982).
22. Myers and Sweezy (1978).
23. Rothwell (1974).
24. Rogers (1971).
25. Mansfield (1982).
-41-
26. Metcalfe (1981).
27. Metcalfe (1981).
28. Urban and Hauser (1980).
29. Wind and Robertson (1983).
30. Cooper and Schendel (1976).
31. Petrov (1982).
32. Ansoff and Stewart (1967).
33. Abell (1980).
34. Freeman (1982).
35. Porter (1983).
36. Prahalad (1974), Rosenbloom (1978), Kantrow (1980).
37. McGinnis and Ackelsberg (1983), Foster (1986), Friar and Horwitch
(1985).
38. Liberatore and Titus (1983), Frohman (1982), Friar and Horwitch
(1985).
39. Freeman (1982), Porter (1979), Snow and Hrebiniak (1980).
40. Nelson and Winter (1977).
41. Wheelwright (1978), Hayes and Wheelwright (1979), Utterback and
Abernathy (1975), Moore and Tushman (1982).
42. Freeman (1982), "Firms that cannot perform radical innovation
fail in the long run."
-42-
CHAPTER 3
FRAMEWORK AND HYPOTHESES
Framework
The major difficulty with the divergent literature on
technological innovation and competitive advantage is the same as the
problem with the strategy literature in general -stream integration.
lack of research
To understand performance in an industry, one
must analyze market structure, the conduct of sellers, and the
conduct of buyers.
The various research streams each separately
analyze one or two aspects of performance but not all three
together.
This separation leads to a priori theories yielding
ambiguous predictions and empirical studies providing conflicting
results.
The industrial organization literature considers the structure
- firm conduct - performance paradigm, with emphasis on structure.
The organization studies literature analyzes firm conduct and
performance, and the marketing literature analyzes buyer behavior.
All three research streams are needed together to explain
performance.
In this section, we draw on resources from all three research
streams to develop a line of reasoning as to what technological
competition really means and how it affects market performance.
Many
pundits have proclaimed that all that U.S. firms have to do to gain
competitive advantage is invest in R&D.
It was demonstrated in the
previous section, however, that gaining advantage through innovation
is not always appropriable to the innovating firm.
The question
still remains as to what it means to compete with technology and when
is innovation a viable competitive weapon?
-43-
In many industries it is plain that competition among firms
centrally involves their R&D policies, successes, and failures. 1
For a firm to gain competitive advantage through innovation, it must
have the capability to innovate.
Within every industry,
organizations are recognized as having distinctive competences, that
is, capabilities that their competitors do not. 2
For an innovating
firm, then, to gain competitive advantage, it must be better able to
innovate than its competitors.
Although studies about the distribution of capabilities within
and across organizations are rare, there is a growing body of
literature that is beginning to illuminate that firms have different
distinctive competences and that they vary by type of strategy
employed.
This should not be surprising in that it means that a
firm's effectiveness at realizing intended strategies depends
significantly on the existence of a match between strategy and
organization. 3
Some of the early work in this area was performed by Lawrence
and Lorsch (1969), who demonstrated that depending on the competitive
requirements, there will be greater or lesser differentiation between
functional groups and these groups will have different relative
importance.
In high-performing companies where technology was
important, R&D had the greatest influence of the functional
departments.
Snow and Hrebiniak (1980) analyzed ten functional areas and
their relative importance for various strategies.
The only
distinctive competence that could distinguish which strategy a firm
-44-
was employing was R&D.
Firms that emphasized R&D, moreover, as a
distinctive competence had no associated competence in marketing.
Hitt, Ireland, and Palia (1982) found that R&D was
significantly more important than marketing or production for firms
emphasizing internal growth.
Firms emphasizing external acquisitive
growth had distinctive competences in marketing and finance.
Further studies by Gupta and Govindarajan (1984) and Tassey
(1983), again demonstrated that firms emphasize different functional
skills within the same industries.
Tassey was able to further
distinguish the innovation orientation of a firm by its greater
emphasis on R&D over marketing.
There is evidence, therefore, that a firm's varying of emphasis
on functional capabilities is closely aligned with its business
strategy.
The business strategy or thrust should give rise to some
critical advantage for some market within the business domain. 4
Rothschild (1979) claims that a thrust of a business can be based on
product innovation, marketing, manufacturing, finance, or executive
expertise.
O'Shaughnessy (1984) claims that a firm can actuate only
one thrust at a time.
In its purest form, then, a firm competing with technology is a
firm whose business thrust is innovation and whose distinctive
competence is R&D.
Such firms can be determined by their relative
emphasis on R&D as compared to the other functional areas, and R&D
led firms are distinct from firms with other thrusts.
Now, for the other half of the question, we must look at when
an R&D thrust is a viable strategy.
To gain competitive advantage,
according to the industrial organization literature, a firm can
-45-
either compete on price or differentiate.
For a firm competing with
innovation, then, it can either use innovation to reduce price or to
differentiate product.
Abernathy and Wayne (1974) claim that a firm
cannot actuate both a cost strategy and a product R&D strategy at the
same time because the former precludes the latter.
Porter (1980),
moreover, claims that a firm in general can only do one or the other
whether the company is innovation based or not.
A firm using
Innovation to create new products, then, is actuating a product
differentiation strategy.
Differentiation is a catchall term for products that are
nonhomogeneous or imperfect substitutes for each other.
Differentiation may occur from several sources, however.
Caves
(1972) considers two broad sources of differentiation between
products:
differentiation of fundamentally homogeneous products and
real product differentiation.
Bain (1968) decries the use of
differentiation as too broad a category and lists five sources of
product differentiation:
1.
Product quality
2.
Buyer ignorance
3.
Sales-promotion
4.
Prestige goods
5.
Seller location
Only within product quality would Bain consider real product
differentiation.
Product quality, however, may have many different meanings.
Although a recent Wall Street Journal 5 poll lists product
-46-
performance as the number one determinant of product quality to
consumers, quality may also be determined by lack of manufacturing
defects 6 or better product inputs, customer service, reliability,
longevity, and image.
Economists7 do not consider there to be any
difference between changing quality of a product through innovation,
inputs, or advertising.
Studies that analyze product quality and its
relationsip to market performance also do not make any distinction as
to how product quality is achieved.8
We know from the earlier discussion, however, that firms who
differentiate their products in one of the many ways listed have
different organizational capabilities and different thrusts.
A firm
with a manufacturing thrust can achieve both improved quality and
lower price, 9 a position not previously thought possible because of
the dichotomy in the definition of price competition and
differentiation (or non price competition).
The level of abstraction
of the concept of the two generic strategies is so high that actually
putting into effect a strategy based on either of them is impossible
as is trying to study the success of such a strategy.
A firm's
strategy must be analyzed from its primary source of differentiation.
A firm whose thrust is R&D is most likely involved in product
innovation.
new products.
Scherer (1970) figures that 86 percent of R&D goes to
Nelson and Winter (1977) claim that firms rarely do
process innovation but rather adopt supplying firms' products.
firm, therefore, is more likely to be attempting product
differentiation through innovation.
Freeman (1982) is more
straightforward and states that R&D intensive competition means
improving product performance.
A
-47-
The purpose for investing in R&D is the expectation of
achieving a monopoly with accompanying supranormal profits.
Economists, however, consider there to be a range of monopoly power:
from a pure monopoly to oligopoly to monopolistic competition, in
decreasing order of market power.
Monopolistic competition provides
only short run profits to competitors; in the longer run monopolistic
competition is the same as pure competition.
The goal of R&D, then,
is to increase seller concentration maximally.
To increase seller concentration, however, an innovation must
be both defensible and unsubstitutable.
Caves (1972) noted that
innovative profits are captured before competitors can copy them.
This protection from copying can come from patent protection, trade
secrets, or first mover advantages.
The second condition is that the
innovation must have no ready substitutes, or in other words, it must
be truly differentiated from other products.
The level of differentation from technological innovation comes
from creating a totally new product or from improving performance on
existing products.
The level of differentiation is determined from
the perspective of the buyers, not the sellers, and is the element of
analysis that is most often missing from discussions of technology
strategy.
Differentiation, however, is not a simple continuum in the
sense of the more the better.
Differentiation advantages arise from
the level of consumer ability to analyze products. 10
If buyers can
exactly appraise the differences in products and they all agree, then
the market result is the same as if the product is undifferentiated.
-48-
Competition in this case consists of simple price - performance
trade-offs.
At the other extreme, if buyers cannot tell the difference in
products or the differences are small, then again no real
differentiation will have occurred.
The market will consist of
monopolistic competition in which there will be no returns to the
innovating firm.
When consumers typically lack the skill to evaluate
the different brands, they may form their preferences on the basis of
superficial appearance, advertising claims, and the like.
In fact,
in an industry where performance is not readily apparent, the
technical dimension of competition may disappear.
For true product differentiation to occur, buyers must have
different preferences or some appraisal skill but not exact.
If
buyers' preferences are different, however, innovation provides niche
benefits rather than total market power.
The expected returns from innovation, then, very much depend on
buyers' awareness and competitors' ability to replicate.
As is
listed in Exhibit 3.1, very few cells have large expected returns to
innovators.
The conditions for large expected returns require
protection gained from patents, trade secrets, or first mover
advantages and require that buyers can appraise products but have
different tastes.
In every other case, insufficient returns to cover
R&D expenditures are expected unless a firm can gain advantage
through pricing or differentiation from other than product
performance.
In other words, the technical dimension of competition
abates and other functional competencies come into play.
Exhibit 3.1
Hypothesized Returns for a Technology Advance
Level of Protected
Technological Advantage
Commonality of
Buyers' Appraisals
of the Technical
Dimensions of
Competing Products
Results
great
not applicable
quasi-monopoly
some
some
some
some
perfect agreement
some agreement
little agreement
cannot
differentiate
price/performance competition
oligopoly
niche
monopolistic competition
none
not applicable
pure competition
-50Some of the complexity of garnering profit from product
innovation becomes apparent when the responses of buyers are taken
into account.
Only in rather specific instances will a technological
advantage lead to a competitive advantage.
In the other cases,
competitive advantage must come from thrusts in the other functional
areas if any advantage is to be gained.
An awareness of customer
responses, moreover, can help in explaining some of the conflicting
results discussed earlier.
To understand when an innovation is likely to generate positive
returns, the level of analysis must go to that of product-market
segments.
Buyers' responses and level of differentiation only become
clear at that level, and so one would expect the potential of either
positive or negative returns to an innovating firm.
The viability of any given thrust can also be explained at this
level.
If the technical dimension diminishes, then one of the others
must take its place.
This would help explain why in some studies
marketing was important or not, and why R&D ability was important or
not.
Also, if marketing is the viable thrust in a technologically
competitive industry, then market focus and timing of entry would
also be important.
The significance of venturesome technology
projects could also be explained at this level.
In industries where technological competition is prevalent,
companies can be expected to have differing strategies depending on
their distinctive competences.
The strategic choice becomes one of
varying relative significance in functional abilities.
Depending on
consumer behavior, an R&D thrust may not be the most viable one.
Instead, a different thrust or a different combination of functional
capabilities may better fit the environmental conditions.
-51Hypotheses
As was postulated in the previous section, if a firm is to use
product innovation as a competitive weapon, the viability of such a
strategy depends on both the level of protection and the ability of
customers to recognize technological advantage.
If the innovation is
of a type where no strong competitive advantage arises, then a firm
must either differentiate the product by using a different functional
capability or compete on price.
In order to test whether such a framework is reasonable,
technological competition in the medical diagnostic ultrasound
industry was studied.
The alternative functional capability to R&D
that was analyzed was sales and marketing.
This was chosen because
Miles and Snow (1978) found it to be the top strategic function other
than R&D in the electronics industry and so was assumed to hold for
diagnostic ultrasound as well.
The researcher's previous experience
in the industry also led him to limit the number of potential
functional alternatives thusly.
The definitions that marketing can take have a broad range -some would claim that all business decisions are really marketing
ones.
The studies that split R&D capability from marketing, as this
one does, define marketing in a more limited way.
For this purpose,
R&D is concerned with developing or changing the product while
marketing is concerned with targeting customers and selling the
product.
Marketing efforts consist of market research, advertising
and promotion, sales force, distribution, and service
considerations.11
Even price, a normal consideration of the
marketing mix, is always split out from marketing as a strategic
-52-
option.
When the term marketing is used in this paper, therefore,
the more limited definition is understood to be in use.
This study tests the importance of improvements in technology
to success in the market place for diagnostic ultrasound
manufacturers.
Technological competition appears to predominate in
this industry and market shares have been changing dramatically at
the same time.
The hypothesis to be tested, then, in this industry
where, on the surface, technological advance seems to be affecting
market shares is whether the following can be rejected:
Hl: Technological advance leads to improved market performance.
If,
in this industry where technological competition seems to
predominate, the above is rejected, alternative hypotheses might more
fully explain the reality.
H2: Some other factor, e.g. sales intensity, might more fully
explain the changes in market share.
H3: A combination of factors rather than unidimensional
competition is needed to explain changes in market share.
If the second hypothesis were to hold, it would fly in the face
of much that has been written about competition in innovation-based
industries.
Porter (1983) and Krugman (1983) have both stated that
innovation is the main competitive advantage in high-technology
industries.
A study by Tassey (1983), however, has shown that
investment in advertising can be more important than investment in
R&D for improved market performance in technology-based industries.
A corroboration of this hypothesis would raise doubts as to just how
"innovation-based" a so called "innovation-based industry" is.
-53-
If the third hypothesis were to hold, it would contradict
O'Shaughnessy's (1984) assertion that a company can actuate only one
strategic thrust at a time; that it is organizationally impossible
for a firm to gain advantage from more than one competitive
dimension.
It would contradict, moreover, Snow and Hrebiniak (1980)
who found that companies strong in R&D had no associated competence
in marketing.
Support for this hypothesis would explain why the
strategy literature has had difficulty in analyzing competition in
technology-based industries:
to succeed in such an industry a firm
must be able to compete in a manner thought impossible before --
to
be able to actuate combined generic strategies.
In graphic form (Figure 3.1), what is being tested is whether a
company positioned to compete just through R&D (or marketing or
price) can maintain a viable competitive position.
Furthermore, we
want to discover whether firms emphasize a single competency or
combined competencies.
Market performance is used as the test of the
viability of the various positions a firm can take on the cube,
whether these positions are just on the axes or also within the cube.
As a further refinement to the hypotheses for this industry, in
a previous study (Friar (1984)) the author determined that diagnostic
ultrasound customers could not readily differentiate technical
performance.
If the framework is to hold, therefore, an R&D thrust
alone should not be a viable strategy in this industry.
Rather, the
better performers should be the companies that use a different thrust
or that combine competencies.
Figure 3.1: Possible orientations of firms as to strategic thrusts
MARKETING
p
PRICE
R& D
-55-
Choice of Analyses
To answer the above questions, several analyses were performed.
The specific questions answered by each analysis are discussed below.
An industry description is furnished to detail the level of
technological competition in the diagnostic ultrasound industry.
This description provides data on the intensity of R&D efforts and
the rapidity of technology advancement.
An analysis of the patterns of technology development or
acquisition that the firms have used provides data on the use of
technology-sharing mechanisms. Such mechanisms may dampen the use of
technology as a purely competitive weapon.
Market performance of the
firms that share technology is compared to those who do not.
This
analysis is a further test of how important technological competition
is.
An analysis of the major improvements or milestones in the
industry considers the defensibility of innovation.
This analysis
examines the length of time a firm had a major breakthrough to itself
and whether having innovated helped the firm more than the rest of
the industry.
An investigation into the technology cycles underlying
diagnostic ultrasound provides insight into the difficulties of
expanding into new areas.
The importance, also, of having the right
technical capability is documented.
The last analyses test directly the relationship to market
performance of a firm's positioning on the strategic-orientation
cube.
Tests are made at the level of the industry and at the level
-56-
of the specific technology-application segment.
Both actual
positioning and intended positioning are examined.
These analyses taken as a group are used to test the hypotheses
put forward in this chapter.
After a discussion in the next chapter
on the methods of data collection, the later chapters describe the
industry, lay out the analyses, and discuss the findings.
-57-
CHAPTER ENDNOTES
1.
Nelson and Winter (1982).
2.
Snow and Hrebiniak (1980).
3.
Gupta and Govindarajan (1984).
4.
O'Shaughnessy (1984).
5.
WSJ, January 6, 1986, p. 23.
6.
Wheelwright (1978).
7.
Caves (1972).
8.
Phillips et al. (1983), Gale and Klavans (1985).
9.
Phillips et al. (1983).
10. Caves (1972).
11. Cochran and Thompson (1964).
-58-
CHAPTER 4
DATA COLLECTION
Choice of Industry
In choosing an industry to analyze within the framework of
technological competition, we selected an industry that strongly
evinces all the key characteristics of a high-tech industry.
Although an exact definition of a high-tech industry is difficult to
formulate, there are three features that differentiate competition in
the advanced technology area from competition in more traditional
industries.1
()
High-tech industries are typically based on products or
processes which make use of recent fundamental advances in
science or technology.
ii)
The rate of technical progress --including both product
and process improvement-- is very high.
iii) Firms spend heavily on the acquisition of knowledge.
An industry that is considered to be prototypically high-tech
is the medical electronic equipment industry.
The medical electronic
equipment industry is characterized by rapid growth, high
fragmentation, and intense technological competition.
Competition is
waged through technical innovations rather than cost.
The market has
been highly responsive to new products and product improvements so
changing technology has spurred growth through product innovation,
which has led to a proliferation of product lines.
This continued
product innovation, however, has reduced product life spans to less
than three years.
Introducing new products by advancing technology,
moreover, is the entree for new firms, which now number over 400
worldwide.
The high research costs, short product lives, and number
-59-
of firms have driven average profitability in the industry to low and
risky levels.
The medical electronic equipment industry can be divided into
five broad segments:
diagnostic, monitoring, therapeutic,
prosthetic, and surgical support.
The medical electronic equipment
market in the U.S. was estimated at $3.7 billion in 1983, with the
diagnostic segment accounting for 45 percent of that total. 2
Worldwide sales are estimated to be twice that of U.S. sales, so the
worldwide market was about $7.4 billion in 1983.
The worldwide
market is estimated to be growing at 15-20 percent per annum,
although some of the subsegments are growing at twice that rate.
Because the markets are large and growing rapidly, many firms
have entered these markets.
Entry barriers are low.
Manufacturing
consists of component assembly; the components typically make up 85
percent of the cost of the equipment, so manufacturing value added is
very low. 3
Distribution and service can be provided by
manufacturers' representatives until a firm reaches a scale to
perform direct sales.
Capital has been available from venture
capitalists although high levels are not needed because the research
costs are often bootlegged from large firms, universities, and
government research groups.
Many new firms are divisions of
companies not originally in the medical electronic equipment industry
but use expertise from other fields to enter these markets.
The product value added that a company can provide is in its
proprietary assemblage of common components.
Although patent
activity in medical technology is twice that of any other
-60industry, 4 products are readily duplicated
and hard to defend.
In
some product lines, therefore, over 50 companies compete.
No single company markets products in all the major
categories
of products, although some co;'panies emphasize breadth
of product
line.
No company, therefore,
dominates the indust.y.
No
moreover, dominates any single product line because no
company has
been able to transfer product line dominance across countries.
Only
two firms, GE and Siemens, are considered to have market
shares
greater than 10 percent.5
The industry, then, is highly
fragmented.
The industry is not nascent in that it was started
in 1895 and
so is over 90 years old.
The worldwide market has always been
fragmented and the number of firms is still growing.
The industry,
then, is not one that is likely to go through
an industry cycle of
birth, growth, maturation, and decay.
Expectations, also, are for
continued product technology advancements, both
major and
incremental.
A major shake-out of the industry, then, is
highly
unlikely.
Because the medical electronic equipment industry
is so
diffuse, a discussion of one of its subsegments
would be more
manageable.
The subsegment to be analyzed is diagnostic ultrasound
in the United States.
As with the medical electronic industry as a
whole, the competition of improving technology
has led to the
continued fragmentation of the ultrasound market.
Over fifty
companies now compete in diagnostic ultrasound, with
the top five
companies holding about 50 percent of the market in
1983.
companies have entered in the last two years.
More new
-61-
Advances in the state-of-the-art have come quickly.
Obsolescence of equipment by improved technology happens so quickly
that the industry has a cycle time for replacement of a generation of
scanners of only one or two years.
This drive to constantly improve
the technology is expected to continue.
Technological competition
is, in fact, so intense that diagnostic ultrasound has been called
one of the most technologically sensitive industries in the world
(Frost and Sullivan (1982)).
The advantages to studying a subsegment of the medical
electronic industry are several.
Competition within a given segment
is mostly that of incremental innovation, which is what we argued
should be studied.
The level of detail needed, moreover, to
understand competition can only be reached by analyzing indepth the
multitude of variations that make up a technology type.
Most studies
of technological competition only go to the level of the generic
technology.
For medical diagnostic equipment this would be
equivalent to comparing competition among X-ray, CT, NMR, and
ultrasound products.
As will be demonstrated, ultrasound consists of
many varieties and applications that are important to understand and
would be missed at a higher level.
The medical diagnostic ultrasound industry, which exhibits
intense technological competition through incremental improvements to
products. has had very volatile market shares.
Cook (1983) argues
that shares of units sold are the "spoils" of strategic investments.
On the surface, therefore, if there exists an industry in which
technological competition leads to improved market performance, this
should be it.
-62-
We argue that not only does medical diagnostic ultrasound
exhibit the characteristics of an industry we wish to study, and that
by studying it indepth we will get to the needed level of detail, but
we also argue that individual industry characteristics must be
analyzed before aggregation with other industry studies.
Several
studies 6 have questioned the validity of analyzing performance
across industries when there is either inter-industry or
intra-industry heterogeneity as to structural characteristics.
By
studying just diagnostic ultrasound, that problem is avoided.
Participants
The difficulty with attempting to study the medical diagnostic
ultrasound industry is that it is so dynamic and changing, that one
literally needs a scorecard to keep track of the players.
Because so
many of the companies in the industry have sales of less than $5
million, they do not readily appear on any industry listing.
These
companies are often tiny divisions of large corporations or small,
privately-held companies so that little public information exists.
The medical specialties that ultrasound overlaps, moreover, each have
distinct professional organizations so what supplier information they
do keep is fragmented and incomplete.
The first task of data
collection, therefore, was to create a scorecard of the companies
involved in the U.S. medical diagnostic ultrasound industry.
The sources that were used to identify the companies were
medical journals, trade journals, medical supply catalogs, and trade
shows.
Journals such as Diagnostic Imaging, Medical Ultrasound,
Journal of Clinical Ultrasound, The Medical and Health Care
Marketplace Guide, as well as 22 other sources, were scanned for any
-63-
mention of a company in an article or advertisement from the years
1977 to 1985.
Through these sources, 131 companies were identified
as having participated in the U.S. ultrasound market in this time
frame.
Even after a lengthy presurvey search, five more companies
were identified during the study for a total pool of 136 companies.
The purpose of the study is to analyze market response to a
technological advantage, so only companies that sold mainframe
equipment to the end user under their own brand names were
considered.
The purpose of this limitation was to allow testing at a
unit of analysis that the customers perceived.
If Company A sold
Company B's product yet labelled it as Company A's, then it was
considered Company A's.
Likewise, if Company A sold Company B's as
Company B's, Company A was excluded.
Under this condition the
following types of companies were screened from the pool:
1)
Peripheral or non imaging equipment vendors.
2)
Regional or Canadian distributors.
3)
Component suppliers.
4)
Companies with no known address.
This screening left a pool of 81 companies that were thought to
be locatable manufacturers and distributors, or national distributors
of ultrasound mainframe equipment (Exhibit 4.1).
were approached in two different ways.
These 81 companies
The companies that were
leaders in one of the major subsegments of diagnostic ultrasound were
contacted to participate in a detailed analysis of their experiences
in each of the segments.
The major segments to be analyzed are
listed in Exhibit 6.14, and ten companies were approached.
These ten
Exhibit 4.1
Analysis of the Pool of Companies to Contact
Companies originally identified in
literature search
131
New ones discovered
Total
5
136
Eliminations
55
Companies to contact
81
*Reasons for elimination:
(1)
Regional or non imaging equipment vendors
(2)
Regional or Canadian distributors
(3)
Component suppliers
(4)
No known address
-65-
companies comprise over 95 percent of the market in each of the six
segments.
Of the ten companies, listed in Exhibit 4.2, only Johnson
& Johnson refused to participate.
This may have been due to the fact
that the medical equipment divisions of Johnson & Johnson (Technicare
and J&J Ultrasound) were for sale and have recently been acquired by
General Electric.
The 71 remaining companies were approached to provide more
general information on themselves and the industry.
In the course of
the survey collection, it was found that 21 companies had either been
acquired by another ultrasound company, were OEM suppliers only, or
were defunct.
The response rate, therefore, was 88 percent for the
overall study (Exhibit 4.3).
listed in Appendix 2.
The companies that participated are
It is estimated that these companies comprise
over 90 percent of the market.
Surveys
Although much more detail on the survey questions will be
provided in each section of analysis, a few words on the creation of
the surveys is in order here.
The general survey (Appendix 4) was
designed to discover who was participating in the market, in which
subsegments, and by what means.
The surveys were addressed to the
presidents or division general managers of the companies.
Follow-up
calls were made to clarify any ambiguous or confusing answers.
In
addition to the general managers, at least one other person from each
company was interviewed on the history of the company and the
industry.
These interviewees were either in marketing or sales.
Exhibit 4.2
Companies Selected for Detailed Study
Acuson
Diasonics
General Electric
GEC (Picker)
Hewlett-Packard
Johnson & Johnson (Technicare, Irex, Echo Labs)
Philips
Siemens
Squibb (ATL, ADR)
Toshiba
Companies in parentheses are names of the ultrasound divisions if
different from parent's
Exhibit 4.3
Response Rate
Surveys and interviews
37
Interviews based on survey
16
7
No response
60
Acquired, overlapping name,
foreign supplier or defunct
21
81
Non response:
7/60 = 12%
-68-
The indepth survey (Appendix 3) asked for detailed company data
for the years 1979 through 1983.
reasons.
The dates were chosen for two
Firstly, in 1979 the various subsegments began to overtake
the established B-scan market and so most of the ten companies had
entered by then.
Secondly, 1983 was chosen as the final year because
a major change in health care reimbursement took effect in October,
1983.
The results, therefore, would not be confounded by this major
change.
The change that was enacted was the shift by the federal
government in reimbursing hospitals for treating Medicare patients to
a prospective payment based on diagnostically related groups.
Two
potential changes to the ultrasound market that could affect
competition would be the dampening of the hospital market and the
strengthening of the role of hospital administrative personnel in the
equipment purchasing process.
Although respondents felt that the
latter had happened and sales of other imaging devices have dropped,
sales of diagnostic ultrasound equipment have continued to grow
(Hambrecht and Quist estimates, November, 1985).
How strong the
potential for confounding the results, therefore, is still uncertain.
Informants were chosen by the companies and the researcher
after the companies had consented to participate.
The respondent was
either the general manager, product manager, or director of marketing
and planning.
Each company was visited by the researcher to review
and clarify the survey answers.
At least two other people from each
company were interviewed; the interviews generally lasted two hours
apiece.
These interviewees were from marketing and R&D.
-69-
The indepth survey was used to collect data on two types of
variables:
the expenditures a manager or a firm can control in
exercising market conduct and actual market performance data for each
segment.
The market conduct variables are similar to the ones used
in other studies attempting to explain changes in market share
(Zeithmal and Fry (1984); Phillips, Chang, and Buzzell (1983)).
For
a review of the studies that demonstrate the relationship between a
strategic variable and its impact on market share, see the Zeithmal
and Fry article.
A prestudy survey of radiologists as to what factors they
consider when purchasing ultrasound equipment was performed.
The
strategic variables were tailored in this way to the ultrasound
industry.
Survey collection was conducted from July,
1986.
1985 until April,
Some of the respondents did not fill out the surveys
themselves but consented to interviews that asked questions based on
the survey.
Confidentiality of individual company data was assured
so that in many instances company identities are hidden.
Only when
information was already public will specific companies be mentioned.
-70-
CHAPTER ENDNOTES
1.
Krugman (1982).
2.
Figures are from:
3.
Abell (1980).
4.
Arthur Young & Co., Study for National Center for Health Services
Research, 1982.
5.
Eberstadt and Co., estimates.
6.
Frazier and Howell (1983), Bass et al.
Electronics, 57.1, January 12, 1984.
(1977, 1978).
-71-
CHAPTER 5
UNDERSTANDING ULTRASOUND
Ultrasound Technology
Ultrasound refers to the generation of sound waves whose
frequencies are above 20,000 Hertz.
In medical diagnostic ultrasound
imaging, sound waves are used to produce images of the body.
The
technique was first used in 1943 in an attempt to map the brain.
The
technique was borrowed from the military's use of SONAR, developed in
World War I, and from industry's use of ultrasound to detect metal
flaws, developed in the thirties.1
By the early 1960s the first
commercially produced diagnostic ultrasound units were available. 2
Ultrasound images are used by physicians as an aid in
establishing diagnoses.
Ultrasound can provide clinically useful
images of most parts of the body.3
To create an image, ultrasound
uses reflected sound waves as the means of visualizing anatomic
structure.
Soft tissues have varying acoustic characteristics so
that by measuring the changes in speed, attenuation, and scatter of
the echo, one can form a mapping of the various tissues.
Diagnostic
ultrasound images are used to provide information on the location and
size of organs rather than on the physiologic functioning of those
organs.
The information provided to a physician is a cross-sectional
display of anatomic images.
The basic elements of any diagnostic ultrasound system are a
transducer, a transmitter, a receiver, a signal amplifier, and a CRT
(Hamilton (1982)).
More recently, microcomputers have become an
increasingly important part of a system; not only to perform simple
measurements but also for controlling the firing sequence and
focusing of array transducers (Figure 5.1).
Figure 5.1:
Basic Elements of an Ultrasound System
PHASED ARRAY
I
UNEAR ARRAY
MECHANICAL SECTOR
DEDICATED
SIGNAL
SCANNER
PROCESSING
ELECTRONICS
ELECTRONICS
DIGITAL
SCAN
CONVERTER
AND
MEMORY
IMAGE IS
FORMED IN A FIXED
TRANSDUCER LENS
Source:
CONVENTIONAL ULTRASOUND SYSTEM
RNM Images (January, 1985)
DISPLAY
-73-
The transducer acts as a sender and receiver of sonic waves; it
functions as a receiver 99.9 percent of the time. 4
The transducer
converts electrical energy into sound waves and vice versa by means
of a piezoelectric crystal.
Depending on the intended application,
the size, shape, and frequency range of a transducer will vary.
In
medical ultrasound, transducers usually have a frequency range of
between 1 and 15 MHz.
The lower the frequency, the greater the beam
penetration, but also the lower the resolution.
There is a direct
trade-off, therefore, between depth and resolution.
The transmitter regulates the sonic waves through the
transducer.
A timer in the transmitter controls the frequency and
duration of ultrasonic pulses emitted by the transducer.
Each echo
received by the transducer causes a small voltage to appear across
the piezoelectric crystal.
If the emitted pulses are intermittent, a
single crystal can both send and receive the signal.
If the outgoing
pulse is continuous, then another adjacent crystal must be used to
pick up the reflections.
The voltages caused by the returning pulse
are amplified, rectified, and smoothed.
Rectification is the removal
of all the negative components of the signal.
These processed
signals are then displayed on a CRT or oscilloscope.
The image is assembled one bit a time.
Each returning echo
generates one bit of data, and many bits together form the electronic
image.
In early equipment, each bit was considered to be either
black or white because the early displays could not differentiate
shades of brightness.
In 1973, the first scan converters were
introduced that could display shades of gray.
In 1977, digital scan
converters were introduced with enhanced gray scale.
Some present
-74-
systems display 64 shades of gray.
Gray scale capability resulted in
much more distinct images and made it possible to depict organs that
had not previously been visualizable.
There are three basic types of ultrasound systems (Exhibit
5.1).
These are pulse-echo, doppler, and duplex.
Pulse-echo
scanners send out intermittent sound waves while doppler scanners
most often emit continuous bursts.
Doppler always is used to detect
velocity in moving objects by measuring "Doppler-shift."
Duplex
combines pulse-echo and doppler techniques into one image.
There are three different types of pulse-echo scanners:
A-mode, M-mode, and 2-D imaging.
A-mode is a one dimensional
representation of the amplitude modulation of the reflected pulse.
The height of the image on the screen is related to the intensity of
the echo.
This system can acurately lccae
sources of all
echoes.
M-mode, or motion mode, is a method of display in which tissue
depth is displayed along one axis and time is displayed along the
second axis.
If a structure of interest, e.g., a heart valve, is
moving toward or away from a transducer, its movement will be
accurately followed.
Two-dimensional scanning takes the basic A-mode information and
extends the one-dimensional image into two dimensions by sweeping the
ultrasound beam across a field of view.
This plane is then
equivalent to a cross-section of tissue that can be imaged by
successively plotting the echo amplitudes along the adjacent range
directions and effectively connecting the dots along the second
dimension.
2-D systems are classified by the manner in which the
beam is swept through the field of view.
Exhibit 5.1
Medical Diagnostic Ultrasound Technology Types
I.
Pulse-Echo
a.
A-mode
b.
M-mode
c.
2-D
1.
B-scan
2. Mechanical sector scan
3. Mechanical linear array
4. Electronic sector scan or phased array
5. Electronic linear array
6. Annular array
7. Trapezoidal array
8. Curvilinear array
II.
Doppler
1.
Continuous wave
2. Pulsed
III.
Duplex
1.
Echo-doppler
2. Color flow mapping
-76-
The oldest of the two-dimensional systems is the B-scanner.
A
B-scanner uses an articulated arm to build a static picture in
several sweeps.
Position and angle sensors in the scanning arm
identify the coordinate position of the imaging plane relative to the
With each pass, the new data is added to the image display.
patient.
In opposition to the creation of a static picture with a
B-scanner, real-time systems produce image frames fast enough to
allow motion to be followed.
Real-time systems do not use an
articulated arm to keep track of position, but rather scan in
whatever direction the transducer is positioned.
Real-time systems
are divided into mechanical and electronic systems.
Mechanical scanners produce an image by mechanically rotating
the transducer, transducer reflector, or group of transducers so that
its beam scans the field.
Electronic systems electronically switch
between transducer elements to sweep the field.
The geometric shape of the image display is the final
classification of the pulse-echo scanners.
Pie-shaped images are
called sector scans while rectangular images are known as linear
arrays.
These may be either mechanically or electronically
produced.
Other electronic images are trapezoidal arrays, annular
arrays, and curvilinear arrays (Figure 5.2).
Within each type of ultrasound classification, the various
manufacturers use a wide variety of methods to create the ultrasound
image.
An example which illustrates this is the variety of
transducer configurations for the mechanical sector scanners
available.
Products presently have:
1) a single stationary transducer with oscillating mirror;
or
2) a single oscillating transducer;
Figure 5. 2:
Samples of Ultrasound Images
a) B-scan
b) Sector scan
Figure 5.2 (continued)
c) Linear array
d) Curvilinear array
-79-
or
3) a single rotating wheel with 2, 3, or 4 transducers;
or
4) multiple rotating wheels with 3 transducers.
The technical alternatives, therefore, in choosing an
ultrasound scanner are broad and diverse.
Each modality has certain
advantages and disadvantages compared to the others when trying to
image a certain part of the body.
None of the modalities has yet
been retired from clinical practice, although there is tremendous
competition to determine which modality is better in any given
application.
Applications
The clinical use of ultrasound is widespread and accepted.
It
has become a normal part of a standard radiology residency program
and is required for board certification.
Ultrasound use, however,
spreads far beyond radiology and into other medical specialties.
Ultrasound must compete with other imaging modalities to become the
test of choice.
Its main rivals are conventional radiography, CT,
radionuclide scanning, and MRI.
Ultrasound is considered to be a safe and inexpensive imaging
alternative.
It is noninvasive, nonradioactive, and requires no
contrast media.
For the time duration and frequency range in which
it is used, ultrasound has not yet been shown to have any significant
biological effects in mammalian tissues.
It therefore involves no
known health hazards.
Ultrasound studies are not limited in their plane of view, and
they can differentiate soft tissues of different densities, which
-80radiographic studies cannot.
Ultrasound is not only considered to be
safer than other modalities but sometimes more effective.
The disadvantages of ultrasound are its limits from
obstructions and its difficulty in use.
Ultrasound will not pass
through bone or gas and therefore cannot image behind these
obstructions.
Likewise, ultrasound images are relatively difficult
to create and to interpret and therefore require greater skills from
the user in reading the image.
With the wealth of imaging possibilities available, the medical
profession has not determined definitive pathways for studying
various diseases in patients.
Because of this, the various
modalities are more often used in conjunction with each other rather
than to the exclusion of the other.
Since several modalities can
image the same parts of the body, the competition between modalities
becomes one of being the preferred initial examination.
An example
of the preferred initial examinations for certain ailments is
presented in Exhibit 5.2.
Not only does ultrasound compete with other modalities, it
competes within itself.
The various ultrasound modalities can image
the same organs and there is no definitive course for choosing one
over the other.
A physician, therefore, has flexibility when
ordering or performing diagnostic tests in choosing among general
imaging modalities and among ultrasound modalities.
Of the many parts of the body that ultrasound is of use in
imaging, the easiest way to categorize them is by the medical
specialty that is concerned with that part of the body.
The one
exception is radiology, which is interested in imaging all parts of
Exhibit 5.2
Preferred Initial Imaging Examinations
Disorder
Examination
Abdominal or pelvic mass of
unknown origin
Plain roentgenography and ultrasound
Abdominal pain of unknown origin
Contrast examinations (intravenous (IV)
urogram, upper gastrointestinal tract
studies, barium enema examination) tailored
for suspect organ system (exception:
ultrasound for specific question of
gallstones)
Suspected adrenal problem
Ultrasound; computed tomography (CT) if
sonography is not satisfactory
Acute cholecystitis
Nuclear scintigraphy
Gallstones
Ultrasound
Cholestasis
Ultrasound and nuclear scintigraphy
Hepatic enlargement
Ultrasound; CT or angiography if sonography
shows focal solid masses
Pancreatitis or pseudocysts
Ultrasound; CT if sonography is not
satisfactory
Neonatal renal disorder or mass
Ultrasound; nuclear scintigraphy preferred
over IV urography for evaluating renal
function in first few days of life
Renal cystic disease
Ultrasound
Renal trauma
CT is most sensitive, although IV urography
is more often performed for expediency
Abscess
Ultrasound; CT or nuclear scintigraphy if
sonography is not satisfactory
Source:
American Journal of Diseases of Children, Vol. 135, Oct. 1981, p. 961.
-82-
the body.
Because of the historic development of ultrasound,
radiologists were at first interested in using ultrasound to scan the
upper abdomen.
When radiology is listed as a specialty, for
ultrasound it is considered to mean scanning of the abdomen.
In
Exhibit 5.3 are listed the major medical specialties that use
ultrasound devices.
This list is by no means complete because unlike other major
diagnostic equipment, ultrasound equipment is purchased and used by
specialists other than radiologists.
Although radiologists have been
the major user group of ultrasound, the other specialties have
developed their own uses for ultrasound.
Manufacturers, in turn,
have designed specialized equipment for these specialties.
Other
areas not listed that are beginning to use ultrasound are:
dermatology, pathology, surgery, and dentistry.
Ultrasound Development
The use of ultrasound as a medical imaging device was developed
over a twenty year period by medical researchers before any
commercial products were introduced. 5
The development took place
in several countries and was performed mostly by university-based
researchers.
In 1943, an unsuccessful but first attempt at using ultrasound
occurred when Dr. Dussik tried to map the brain.
In the late forties
and early fifties, Drs. Howry, Wild, Reid, Bliss, and Holmes, all in
the U.S., began to image internal structures in a cross-sectional
depiction of different parts of the body.
Their equipment was
adapted U.S. Navy surplus sonar equipment or self-made instruments.
Exhibit 5.3
Medical Specialties That Use Diagnostic Ultrasound Equipment
Radiology (Abdomen)
Cardiology
Obstetrics/Gynecology
Opthalmology
Neurosurgery
Neonatology
Neurology
Cardiovascular Medicine
Endoscopy
Urology
-84-
In Sweden, cardiologist Inge Edler began imaging the heart.
Also in Sweden, Dr. Leskell, a neurosurgeon began imaging the brain.
Ultrasound was also tested for gynecological and obstetric purposes
at the University of Glasgow.
In Japan, Dr. Santomura first used
doppler techniques in medicine.
All these researchers published
their findings in the 1950s and built their own machines to perform
the studies.
In the 1960s commercial products were made available, but the
companies used designs developed by university or NIH researchers.
The first two-dimensional scanner for sale in the U.S. was developed
by a research group at the University of Colorado and was marketed by
a company named Physionics.
The first real time unit was also
developed at the University of Colorado and marketed by Magnaflux.
Further advancements by medical researchers led to the use of
ultrasound in opthalmology, breast imaging, echoencephalography, and
vascular imaging.
Technical developments included the development of
contact scanners, mechanical sectors, rotating wheel transducers,
linear arrays, and phased arrays.
Many of the ultrasound
technologies and applications, therefore, had first been developed by
medical practitioners long before the commercial market was ever
established.
In the early 1970s, the National Science Foundation (NSF), in a
program proposal, concluded that U.S. manufacturers were not
undertaking enough development work for commercializing medical
ultrasound. 6
The NSF had recognized ultrasound's great potential
for medicine and was worried that instruments were being vigorously
developed abroad.
In 1973 ultrasound was growing more visibly
-85-
outside the U.S., and this growth was spurred by government support.
Australia, in fact, had set up a separate research institute for the
development of ultrasound instrumentation.
The NSF in 1973, therefore, decided to create an incentives
program to spur American industry's involvement in ultrasound.
The
NSF created performance specifications for an instrument that would
presumably meet a real medical need and that was technically
feasible.
Any company creating an instrument to specifications would
have clinical testing sites in the VA system made available.
program was announced in February,
30,
The
1974, and had a deadline of April
1978.
Although 12 companies did sign up for the program, no one
developed a product to NSF specifications.
In a review of the
incentives program, Arthur D. Little compared the growth of
commercial activity in four sectors of diagnostic medicine.
They
concluded that the development time for ultrasound was normal, that
the NSF's basic hypothesis was incorrect, and that the incentives
program really had provided no incentives.
Ironically, at the same time as the NSF program, a small
company in California, Robe Scientific, developed the first practical
stored video gray scale in a B-scan device.
Because gray scale in
different versions had been tried for several years, several
companies had stored gray scale capability within a year.
The
introduction of gray scale has been credited for the initial growth
in commercial ultrasound activity.
With the growth of commercial activity, much of the technology
development shifted to the industrial sector and to U.S. companies.
-86-
In the 1970s, peak detection, gray scale, scan conversion, digital
processing, automatic scanning, and beam focusing were all developed
by companies in the U.S.
The development of ultrasound, seemingly first by the
scientific community and then by the industrial community, has not
followed the straightforward pattern pictured in the simple model of
development as described by Allen (1977) and shown in Figure 5.3.
The practical need for diagnostic imaging is well understood and
human anatomy has been stable for quite some time.
That parts of the
human body need to be seen before an autopsy is apparent enough.
What is not so well known a priori is the relationship between
certain diagnostic findings and pathologies.
An example is the
visualization of a fetus in utero, which really began with ultrasound
scanning.
As more studies were performed, fetal growth, dating,
behavior, and abnormalities could be determined.
An illustration of
the advancement of findings to pathology is the presence of fetal
breathing (a periodic see-saw movement of the fetal chest).
Fetal
breathing was first correlated to gestational age and then to
favorable perinatal outcomes.
Later, fetal breathing was shown to be
of use in identifying false-positive contraction stress tests.
Fetal
breathing is now considered a regular part of an antepartum fetal
evaluation.
As certain pathologies are better understood, the need to
better image certain organs grows.
Likewise, as imaging techniques
improve, applications become apparent.
The proper application,
therefore, of a new technology in medicine is often not apparent at
the start.
New devices need testing to develop ways of using them
Figure 5.3:
Science, Technology, and the Utilization of Their Products,
Showing the Normal Progression from One to the Other
Science
Body of
Knowledge
Technology
State of
the Art
Practical
Need and
Use
Utilization
Time
Source:
Allen (1977)
-88-
and to modify them if necessary.
The normal procedure for diagnostic
ultrasound has been that clinicians test the new technology in
various applications until clinical efficacy is demonstrated.
Examples of the need to search for the proper application
abound.
The importance to a company of finding the correct
applications is illustrated by two examples.
ADR, a small startup
company in 1973, developed one of the first linear arrays for the
U.S. market.
ADR developed the linear array for cardiac applications
but found it not useful in clinical trials.
By accident ADR found
that it was useful for imaging the fetus and created the obstetric
ultrasound market.
Another example is ATL, another startup that introduced early
duplex scanners in the U.S.
ATL's original intent was to develop
cardiac applications, but they almost went bankrupt doing so.
ATL
then shifted to radiology applications and became successful.
Several years later, cardiologists began accepting the clinical
utility of duplex scanners.
In medical ultrasound, as in other innovation-based industries,
technological improvement and clinical skill advance together.
Because clinical practice is so closely aligned with scientific
research, the development of ultrasound more closely resembles
Allen's complicated development pattern (Figure 5.4).
The scientific
community, the clinical community, and industry work closely together
and push each other ahead.
The purpose of this section has not been the testing of
theories on technology development, but rather the describing of a
rich environment of scientific and technological advance.
The
Figure 5.4:
Science, Technology, and the Utilization of Their Products,
Showing Communication Paths Among the Three Streams. (a) The normal
process of assimilation of scientific results into technology. (b) Recognized
need for a device, technique, or scientific understanding. (c) The normal
process of adoption of technology for use. (d) Technological need for understanding of physical phenomena and its response (from Marquis and Allen,
1966).
Body of
Knowledge
Science
Sd
d
Technology
State of
the Art
C
Practical
b
Need and
flzto
Use
Time
Source:
Allen (1977)
-90-
challenge to a company has not only been the development of
technology but a matching of the technology to an application.
Scientists and clinical researchers have also been developing
instruments and applications, which has led to the creation of new
markets.
In a later chapter, data collected during the study are
used to illustrate the importance of matching technologies to
applications.
This section has illuminated the many avenues by which
it has been done.
Government Agencies
Although many government agencies have jurisdiction over
medical ultrasound devices, the agencies that support the advancement
of ultrasound have been more influential than the agencies that
regulate its sale.
After considering the regulators, we will discuss
the supporters.
The Food and Drug Administration has two programs that could
affect diagnostic ultrasound:
the Bureau of Radiologic Health (BRH)
and the Bureau of Medical Devices (BMD).
The BRH is responsible for
protecting the public from unnecessary exposure to electronic product
radiation.
The potential radiation hazard from diagnostic ultrasound
devices, however, was not deemed serious enough to impose any
standards.
The FDA did suggest that the industry develop its own
voluntary standards, which we will discuss later.
The BMD is responsible for assuring the safety and efficacy of
all medical devices.
Diagnostic ultrasound is considered a Class II
product which means that it does not have to pass any premarket
approval process.
As long as new ultrasound products are not
-91-
significantly different from existing products, only a 90 day notice
of intent to sell is required.
The Federal Communications Commission has authority to minimize
radio frequency radiation emitted by electronic instrumentation.
Although medical ultrasound equipment technically must comply with
FCC regulations, very few manufacturers have done so. 7
Certifir
:tna (CON)
programs require a hospital to
demonstrate a need for purchasing certain equipment.
Although these
programs are usually run at the state level and have varying
criteria, the minimum purchase price for consideration generally was
over $100,000.
This minimum was raised to $400,000 in 1981.
Diagnostic ultrasound machines generally cost less than $100,000 and
so were not restricted by CON requirements.
The U.S. Department of Health and Human Services has a group,
the Health Care Financing Administration (HCFA), that is responsible
for overseeing the Medicare and Medicaid programs.
The National
Center for Health Care Technology (NCHCT) is responsible for
reviewing new technologies for the HCFA and recommending whether
Medicare should reimburse for these procedures.
The NCHCT was
authorized by Congress in 1978 and to date has not reviewed
ultrasound procedures that were in practice before 1980.
These
procedures, therefore, have continued to be reimburseable under
government plans.
Two trade groups, the National Electrical Manufacturers
Association (NEMA) and the American Institute of Ultrasound in
Medicine (AIUM) have been working together in a joint task force
since 1980 to develop voluntary performance standards for ultrasound
-92-
equipment.
Although interim standards have been developed, they are
still voluntary.
A policy for obstetrical ultrasound exams,
moreover, was not adopted by the AIUM until October, 1985.
These
standards are in fact really guidelines to which manufacturers have
varying levels of compliance.
The National Institutes of Health (NIH), and to a lesser
degree, the National Science Foundation promote research in the
development of medical ultrasound instrumentation and applications.
In 1982, these agencies provided $7.3 million of research
funding.8
In addition to providing research support, the NIH in
conjunction with the FDA sometimes assembles panels of experts to
review medical procedures.
In February, 1984, such a consensus panel issued a report
recommending that the use of ultrasound be limited in obstetrical
practice.
The panel said that ultrasound imaging appears to be safe
and useful in most cases; but that in the absence of conclusive test
results, a hypothetical risk must be presumed.
The panel listed 27
indications for which ultrasound benefits have been documented and
for which ultrasound is recommended.
The head of the panel estimated
that one-third of all pregnancies evince at least one of the
indications. 9
However, because ultrasound is used to examine
between 15 and 40 percent of all pregnant women, the panel's warning
is not expected to alter usage patterns.
Although several agencies have authority over medical
ultrasound equipment, none of these agencies has really slowed or
hindered the introduction and sale of ultrasound equipment.
On the
contrary, the NIH has been encouraging the advancement of ultrasound
-93-
technology by spending millions of dollars annually for research in
ultrasound instrumentation.
Image Quality
Of the several people who can be involved in a capital goods
purchasing decision for a hospital or a medical practice, the person
who evaluates diagnostic equipment as to its performance capabilities
is the physician.
A physician is highly educated and sophisticated
as to medical practice and diagnostic needs.
When purchasing
equipment, the physician participates in a decision process that is
long and involved.
In an attempt to make an informed decision on an
expensive and important purchase, the physician performs an
information search, asks for recommendations, and compares products.
E. Von Hippel (1976) defines innovation improvements to
scientific instruments to mean those technology advances that improve
the functional utility of the products.
For ultrasound equipment,
one of the main benefits of technology advance would be increasing
the diagnostic information one could receive from an ultrasound
scan.
To provide more diagnostic information, an ultrasound scan
must produce better anatomic images.
The importance of image quality in this industry is beyond
dispute.
Image quality has improved dramatically over the last
several years, which has led to improved clinical utility. 10
Because the advances in technology were to improve image quality,
technology has been considered the most important factor in industry
growth. 1 1
Advancing product technology and improving equipment
performance is considered the main form of competition among
-94-
suppliers.1 2
Image quality, for the buyers moreover, is considered
the number one purchasing criterion.13
The present images from ultrasound scans are still difficult to
create and to interpret.
quality.15
Users continue to seek better image
To gain increased diagnostic information from
ultrasound scans, however, manufacturers are trying two tacks:
improving the images and creating computer analyses of the images.
Most of the foreseeable advances in technology that will improve
ultrasound's diagnostic accuracy, however, will come in improving
image quality.)
6
"The flow of novel developments continues today, making
diagnostic ultrasound one of the most technologically-sensitive
industries in the world." 1 7
Nine studies 1 8 of the ultrasound
industry all concur that the industry is technology driven, that
competition rests on improving image quality, and that image quality
is the main purchasing criterion.
Pratt (1978) considered image quality to be composed of two
parts:
image fidelity and image intelligibility.
Image fidelity is
defined as the difference in a processed image from that of some
standard image; the closer the processed image is to the standard,
the higher the fidelity.
Image intelligibility is the ability of a
person or a machine to extract relevant information from an image;
the more intelligible an image, the easier one can extract
information.
To improve functional capability in an ultrasound machine, an
engineer must improve the image intelligibility of the final output.
Because the human visual system is so poorly understood, formulating
-95-
an intelligibility measure based on a perceptual model is not
presently possible.
The most common and most reliable judgment of
image quality, at this time, is subjective rating by human
observers.19
Although human judgment is considered the best rating of image
quality, other methods that might correlate well with the subjective
testing have been tried but have not yet been successful.
Such
methods have included fidelity testing, testing on phantoms, clinical
trials, and measurement of individual parameters of technical
performance.
The difficulty in measuring image fidelity is that at present
there is no official standard against which to measure
performance. 2 0
The relationship of image fidelity, moreover, to
diagnostic utility is currently moot.
An example in current debate
is the ability to image eyelid hairs on a fetus in utero.
What
relevant information is gained?
In an attempt to establish some standards, phantoms (mechanical
representations of tissues) have been used to test fidelity.
Phantoms, however, are static while tissues move.
Phantoms,
therefore, do not test realistic images and are only used to assess
depth calibration and electromechanical alignment. 2 1
Clinical trials are not definitive and take too much time.
clinical trial would test several machines on the same patients.
A
The
same machine, however, can give a good image of an organ on one
patient and not on another patient.
"The favorite explanation for
this, at the moment, is that the image is affected by the
organization of fat in the patient."
22
Clinical trials, therefore,
-96-
need to test many patients.
Besides not being able to test all the
various types of equipment, a clinical trial to compare products
takes a long time.
The rapid technological advances in ultrasound
technology, however, have caused obsolescence of equipment, thereby
rendering clinical trials on that equipment out of date before the
reports are published.
Several difficulties arise in measuring individual parameters
of technical performance.
The main difficulties are judging
resolution trade-offs and judging unspecified design options.
To improve the image of any specific organ, an engineer must
increase the ratio of image signal to "noise."
Ultrasound image
signals, however, carry a relatively high degree of noise.
comes from two main sources.
Noise
Ultrasound does not travel well through
gas or bone and is therefore of little use in imaging structures
behind them. 2 3
Another problem is that tissue is not static.
Tissue motions cause variability in attenuation and absorption,
refraction, and back scatter. 2 4
To improve resolution, one must trade off improvements on five
types of resolution -contrast.25
lateral, radial, temporal, spatial, and
The trade-offs are between aperture size, bandwidths,
movement detection, location, and penetration.
Improving the
resolution in imaging one organ may impair the resolution in trying
to image another organ.
A designer, therefore, must set a resolution
mix as to which organs or types of scans the machine will image
better.
Some rating scheme would have to be determined which could
compare overall machine quality by its ability to improve resolution
on a number of specific organs.
A rating matrix has not yet been
-97-
determined that can compare instruments having different resolution
mixes.
The unspecified design options cannot be judged because they
are not explicit and because no one has yet determined which methods
are better.
Examples are quantization options for analog to digital
conversion and interpolation methods because of undersampling of the
image.
Studies that have measured individual parameters of technical
performance, which are open to testing, have shown a very wide
variation in the magnitude of some parameters among
manufacturers.26
The problem still remains that no one has yet
determined which of those parameters that are measurable lead to
improved image quality.
The use of diagnostic ultrasound is an accepted clinical
technique.
Diagnostic information comes from ultrasound's creation
of anatomic images.
To improve clinical utility, one must improve
the image quality of ultrasound.
The state of the art in judging
image quality, however, relies on the subjective evaluation of human
judgment.
No one has yet determined some other measures to test for
ultrasound image quality.
Selection Process
The selection of ultrasound equipment is a complex process.27
The number of people involved in the selection and the length of time
needed to evaluate equipment ensure a serious, thoughtful
investigation of equipment performance as well as the many other
criteria considered:
ease-of-use, cost, reliability, serviceability,
-98-
and compatibility.28
Because hospital radiology departments have
been the main purchasers of ultrasound equipment,29 this section
will describe a typical hospital's evaluation process.
The performance evaluation of ultrasound imaging equipment is
usually contemporaneous with or lags the capital appropriation
process.
The appropriate personnel in the hospital, therefore, are
aware of the intended purchase.
Along with the radiology ultrasound
specialist, the radiology department head and other specialists will
decide on the applications the unit must perform.
Once the
application mix is determined, the modality type can be selected.
A search then ensues for the products that are available in the
selected modality, with some preliminary screening as to which
products to investigate further.
Trade shows, medical journals, and
peer recommendations are the main sources of preliminary
information.
Prospective buyers can try the equipment at the trade
shows or at another hospital that has a unit.
Once salespeople are alerted that a radiologist is interested
in purchasing some unit, the evaluation process will still take
another six to nine months. 3 0 ,31
During this time, manufacturers
are asked to bring equipment to the hospital for comparison trials.
Some hospitals will request that a unit be left on loan for a week to
a month for even more clinical evaluation.
At this time, the
ultrasound technicians will also get involved in the selection
process since they will perform the trial scans and will be asked for
recommendations.
The ordering process, then, is both lengthy and complex.
purchaser tries for some breadth and then depth in evaluating
The
-99-
equipment.
After an initial search as to what equipment is
available, a number of instruments are selected for clinical trials
and comparisons.
Several people are involved in the decision
process, which is actuated over a long time period.
The decision,
therefore, is an informed one among professional people.
Judging Image Quality
Although physicians claim in survey after survey that the main
purchasing criterion for selecting ultrasound equipment is its
performance capability or image quality, and although the selection
process to choose the best equipment is lengthy and involved; there
is some question as to whether physicians actually can differentiate
performance capability.
As was previously discussed, there is no
objective way to rate image quality.
Even those who claim that the
information content of an image can be objectively measured still
include a subjective portion called "image aesthetics" when
describing the perceived quality of an image.32
These authors
would urge the user to ignore his preferences in image presentation
to "see through" the aesthetic factors.
Either way, the user is the
one who must "read" the image in order to gain diagnostic
information.
Because so many physicians, and not just radiologists, use
ultrasound scanners, physician training ranges from several years to
none.
Surveys indicate that as many as 75 percent of the physicians
with ultrasound scanners in their offices rely on technologists to
conduct exams. 3 3
The head of the American Institute of Ultrasound
in Medicine, moreover, claims that physicians simply do not have the
time to learn ultrasonography well and so must rely on technicians.
-100Physicians must rely on technicians to do more than just perform the
scan, however.
Unlike radiographic studies, the later reading of the
hard copy by someone not in attendance at the exam is difficult.
The problem with reading the hard copy of the sonogram is that
there is no way to localize the scanning plane (for real-time
units).
In other words, someone looking at the picture would not
know where the transducer probe had been placed and in what direction
it was facing.
Any single real-time image may not be understandable
without identifiable landmarks in the image and records of the
transducer's position.
Unlike other diagnostic techniques, there are no standard
positions or techniques by which the sonographer can capture the area
of interest.
Procedures vary because patients vary as to body fat,
gas, and other anatomic obstructions.
Each scan has two parts:
the
survey scan and the specific scan for detailed imaging information.
In the survey scan, a sonographer must decide when a particular
image is of literest or not.
field of view and tech
Ultrasound has a relatively narrow
cal Lrade-offs between depth and resolution.
The operator must continually adjust the transducer direction and the
instrument controls.
It is up to the sonographer to decide whether
the image is in fact the organ of interest or an artifact, or whether
an organ is hidden or atrophied.
If the technologist misses
something important, it will not be recorded.
While positioning the transducer for the detailed study, the
ultrasonographer observes the echoes on a video display until the
transducer is correctly oriented and instrument sensitivity is
properly set.
Only then are images recorded.
But because the
-101-
scanning plane can be moved in any direction and will be because of
patient differences, a common problem for later readers is that the
image is not referenced to any particular position on the body.
Because every ultrasound examination is almost a unique event,
the skills level of the ultrasonographer is considered to be higher
than that of other technicians.
Barley (1984) in his study of
hospital radiology departments found that sonographers were held in
higher esteem than their counterparts in the other areas.
He also
found that sonographers claimed that physicians could not understand
what ultrasound could and could not do, and that they had to teach
the physician how to read a scan.
Furthermore, Barley found that
sonographers were interpreting sonograms and making diagnostic
decisions, which is solely the duty of the physician.
The issue that physicians rely too heavily on technicians to
perform and interpret sonograms has been raised in the medical
literature but not researched indepth.
What has been debated more
openly, however, is whether technologists have the requisite skills
to conduct an ultrasound exam.
Only 22 of an estimated 150 training programs for ultrasound
technologists in the U.S. have been accredited by the Joint
Commission for Diagnostic Medical Sonography. 3 4
Technologists need
not graduate from an accredited program to practice ultrasonography,
nor are they required to have credentials from organizations that
test minimum standards for training and skill.
There are real
issues, therefore, as to whether sonographers are properly trained
and whether physicians have abdicated their roles.
-102-
Ultrasonography is not another extension of a radiographic
technique.
It requires new skills in performing and reading a scan.
Sonographs are radically different from radiographs and are difficult
to interpret.
Ultrasonography also demands that the examination be
carried out and a diagnosis made concurrently, yet technicians, who
may not be properly trained, are performing the exams.
There is some
doubt, therefore, as to whether ultrasound users can really judge
equipment performance, or whether the technology is beyond their
present capabilities.
In a previous study, Friar (1984) showed that application
experts from ultrasound manufacturing concerns consistently rate
ultrasound equipment as to its image quality and feel that there is a
statistically significant difference in that image quality.
Forty-five hospital radiology departments who were actively comparing
ultrasound products in a preselection search were also asked to rate
image quality on those products.
Radiologists are thought to be the
best-trained specialists in ultrasound.
Friar found that the
technicians knew more of the products on the market, but that neither
radiologists nor technicians could differentiate product as to image
quality.
The users, then, with the best training in ultrasound could
not differentiate product on a performance measure.
The Industry
Although the levels of image quality from ultrasound units may
be difficult to compare objectively, the importance of image quality
is undisputed.
Not only does customer survey after customer survey
list image quality as the most important purchasing criterion, all
-103-
the interviewees from the manufacturers list it as the main agent in
shifting relative market shares.
Only those companies that keep up
with the state-of-the-art are thought to maintain a viable
competitive position.
That R&D is important in this industry can be seen in Exhibit
5.4.
The median proportion of R&D to sales is 16 percent, with a
range from 2 to 40 percent.
To augment internal R&D, 47 percent of
the companies license in products from other suppliers (Figure 5.5).
That physicians appreciate improved performance is demonstrated
in the rapidity with which ultrasound diffused.
The introduction of
gray scale in 1973 is credited with improving ultrasound images to a
clinically acceptable level and allowing more widespread use.
The
introduction of digital scan converters in 1979 provided a further
boost for more general acceptance.
The rate of diffusion of
ultrasound, once it reached a stage of practical clinical usefulness,
was very quick --
from less than a 20 percent penetration of
hospitals in 1974 to over a 90 percent penetration in four years
(Figure 5.6)35
Sales of medical diagnostic ultrasound equipment in the U.S.
grew at an annual rate of 39.7 percent over the decade after the
introduction of gray scale.37
As can be seen in Figure 5.7, after
a gestation period of twelve years with minimal sales, dollar sales
volume increased exponentially.
Even after most hospitals had an
ultrasound device, sales continued to grow at an annual rate of
between 20 and 30 percent36 from 1979 to 1983.
market was estimated at just over $300 million.
In 1983, the total
Exhibit 5.4
Statistics on R&D Expenditures for Ultrasound Firms
Proportion of R&D to Sales
N = 29
mean
18.86%
median = 16.00%
minimum
2.00%
maximum = 40.00%
standard deviation = 10.87%
--
Figure 5.5: Proportion of companies that license in technology
(N = 30)
0
License in
10
20
30
40
-
Do not
percentages
50
60
50
60
Figure 5.6:
Cumulative penetration of ultrasound
in U.S. hospitals (by hospital bedsize segment)
100%-
300
-
or
more
200
to
299
75%-
100
--
to
50%-
199
25%
50
to
99
0
to
49
T
1973
1975
1977
I
I
I
1979
Year
Source:
Diagnostic Imaging (October, 1982)
A
1981 Number
I
of
beds
FICURE 5. 7.
SALES OF MEDICAL ULTRASOUND EQUIPMENT (*MILLION)
------ rr-r.-r-r-i
~-T-i
350r--ip
900
200
150
0I
0
1
9
a
0
1
9
a
1
1
9
a
2
I
9
8
3
1
9
8
4
+---: I
1
1
9
9
8
8
6
5
I
1
9
8
7
-T
-- ~T-~+-I4I41
t
1
9
9
9
7
a
8
0
9
8
1
9
7
1
-t
1
9
7
2
II.
1
9
7
3
I
I---I.1
1
9
9
7
7
5
4
I
1
9
7
6
I
1
9
7
7
I..
1
9
7
8
1
9
7
9
... L.1
1
1
9
9
9
8
8
2
1
0
1
9
PERIODS
Source:
Electronics Annual Survey
.
S
A
L
E
S
D
0
L
-
250
1
9
B
a
1
9
B
4
-108-
Because advances in the state-of-the-art have come in quick
succession, market shares have changed dramatically.
As can be seen
in Exhibit 5.5, many of the companies on top in 1978 were not even in
the top ten in 1983.
Many of the companies, moreover, with leading
market shares now were not even in the industry in 1978.
The median age in 1986 of the ultrasound companies or divisions
is 7 years, with a range from several new entrants to one company of
23 years (Exhibit 5.6).
There is no relationship between age of firm
and size of firm (Exhibit 5.7).
The older companies, therefore, have
not been successful at establishing defensible positions and
precluding entry.
The attractiveness of rapid growth has brought many entrants
into the ultrasound market.
Berggren (1985) estimated there to be 37
companies competing worldwide in 1976.
Drew (1981) estimated 36
companies in 1980, and Frost and Sullivan (1982) estimated 34 in 1981
(both estimates are for the U.S. market only).
McKay (1983) claimed
that there was a major egression of companies in 1981.
These data,
however, show that. the number of companies in the U.S. has continued
to grow since 1980.
Although there is constant churn in the industry as to
acquisition, divestiture, entry, and exit, the net number of
companies in 1986 is 54.
Forty-nine companies actually sell under
their own brand names and another five are foreign suppliers who
source more than one company.
This adjustment is needed to match
Drew's criteria for enumerating.
The net number of companies, then,
has grown from about 36 in 1980 to 54 in 1986.
Exhibit 5.5
Leading Sellers of Medical Diagnostic Ultrasound in
the U.S. (Dollars)
Position
1
2
3
4
5
6
7
8
9
Source:
1978
1983
Picker
Unirad
Rohe
ADR
Varian
Searle
ATL
Diasonics
Technicare/Irex
HP
Philips
Biosound
Picker
GE
Toshiba
1978 figures - Hamilton (1982)
1983 figures - Friar
Exhibit 5.6
Statistics on the Age of Ultrasound Firms in the U.S. (1986)
(years)
N = 30
mean = 8
median
7
minimum
0
maximum
=
23
standard deviation
=
5.8
Exhibit 5.7
Comparison of Large and Small Firms
as to Years in Market
Two Sample t-test, variances unknown
Firm Size
Mean
Small
(less than $5million)
7.53
7.05
17
Large
(greater than $5million)
8.62
3.80
13
t-test of ml
=
Standard Deviation
m 2 ; t = -0.54; not significant
N
-112-
Although this study only considers competition in the U.S.,
competition in ultrasound is truly international.
All of the
companies in the survey sell products outside of the United States.
Of the companies selling in the U.S., twenty companies are foreign
and represent 11 different countries (Exhibit 5.8).
In response to an open-ended question, the survey respondents
stated their reasons for entering the U.S. ultrasound market (Exhibit
5.9).
These reasons were coded into five categories.
prominent reasons were:
The two most
the development of a new technology and the
desire to maintain status as a full-line medical equipment company.
The development of new technology occurred from two sources:
companies in non medical ultrasound who used their technical bases to
enter medical ultrasound; and startups with a better idea.
That many
companies started expressly to serve the medical diagnostic
ultrasound market can be seen in the proportion of freestanding
companies (Figure 5.8).
Sixty percent of the companies were
freestanding medical ultrasound companies.
The continued entry of new firms and the international
technological competition have led to a growing fragmentation of the
industry.
The four firm concentration level has decreased to 53
percent from 70 percent 3 7 over the period of 1978 to 1983.
Two-thirds of the companies had U.S. sales in 1985 of less than $5
million (Figure 5.9).
The allure for a company of entering into a rapidly growing
market has also been a bit of a trap.
The fragmentation and high
levels of R&D expenditures have created an extremely unprofitable
industry.
The president of one company cited an article that claimed
Exhibit 5.8
Foreign Companies That Are Presently Competing in U.S. Medical Ultrasound
Country
Company
Australia
Austria
Denmark
England
Ausonics
Kretz
Bruel & Kjaer
Picker (U.S. based)
Sonicaid
Thomson - CGR
Siemens
Elscint
Aloka
Hitachi
Matsushita
Shimadzu
France
Germany
Israel
Japan
Toshiba
Netherlands
Scotland
Switzerland
Yogokawa
Organon Teknika
Philips (U.S. based)
Pie Medical
Diagnostic Sonar
GL
Kontron
Exhibit 5.9
Reasons for Entering the U.S. Ultrasound Market
Reason
Proportion of Firms
Outgrowth of technical capability
.33
Foreign firm entering the U.S.
.15
Serve present customers with new technology
.07
Maintain status as a full-line medical equipment
company
.33
Niche opportunity
.11
N = 27
Figure 5.8: Proportion of freestanding medical ultrasound companies
(N = 30)
0
Freestanding
10
20
30
40
50
60
70
..-IIIII
Subsidiary
percentages
Figure 5.9: Sales distribution of firms
(N = 30)
Sales in millions
Less than $5
:$5 and
$10
0
10
I
20
30
40
50
II
Greater than $10
percentages
60
70
-116-
the industry lost between $50 and $100 million in 1984 on sales of
about $300 million.
Although I could not find the cited article, few
companies admitted to being profitable themselves and fewer still
believed that anyone else was making money.
The survey information, in summary, corroborates what earlier
studies on the ultrasound industry have said.
Technological
competition is intense and has led to rapid market growth and
industry churn.
What the survey appears to refute is that more firms
have left the industry than entered.
This study has found continued
growth in the number of firms and a generally unprofitable
environment because of it.
-117-
CHAPTER ENDNOTES
1.
Klein (1980).
2.
Frost and Sullivan (1982).
3.
Hamilton (1982).
4.
Discussion of technology relies on:
(1980), Berggren (1985).
5.
For the early history of development, we have relied on:
Hamilton (1982), Klein (1980), Berggren (1985).
6.
Discussion of NSF relies on:
(1979).
7.
Frost and Sullivan (1982).
8.
Frost and Sullivan (1982).
9.
Diagnostic Imaging, March, 1984, p. 13.
Hamilton (1982), Klein
Drew (1981), Arthur D. Little, Inc.
10. Technology Marketing Group (1981).
11. Klein (1980).
12. Theta Technology Corporation (1981).
13. Technology Marketing Group (1981).
14. Frost and Sullivan (1982).
15. Technology Marketing Group (1981).
16. J. Lloyd Johnson (1981).
17. Frost and Sullivan (1982).
18. Studies not already cited:
Doz (1975).
Bernstein (1982), Eberstadt (1983),
19. Pratt (1979).
20. Department of Health (1980).
21. Madsen (1978).
22. Personal correspondence with W. N. McDicken, University of
Edinburgh.
23. J. Lloyd Johnson (1981).
-118-
24. McDicken (1983).
25.
Interviews with engineers at Elscint and Philips.
26. McDicken (1983).
27. Hamilton (1982).
28. Fusfeld (1978).
29. J. Lloyd Johnson (1982).
30. Doz (1975).
31. Frost and Sullivan (1982).
32. Maslak (1985).
33. Freiherr (1985).
34. Freiherr (1985).
35. J. Lloyd Johnson (1982).
36. Low figure my estimate, high figure Electronics.
37. Hamilton (1982) for 1978 figure.
-119-
CHAPTER 6
ANALYSES
Overview
In order to answer the questions posed in Chapter 3, this
section contains a multi-level and multi-approach attack.
were performed at two levels:
Analyses
at the ultrasound industry level and
at the specific technology-application level.
Industry competition
was analyzed by using frameworks often found in the literature on
technology strategy.
These frameworks include:
mode of technology
acquisition, firm size, and major improvement advantages.
Two
further analyses considered the effects of technology cycles and the
effects of differences in competitive orientations of the firms.
The competitive orientations of the firms were further analyzed
within several ultrasound subsegments.
This level of analysis was
required to assess the importance of a technological advance when the
products are direct substitutes for each other.
A discussion of the combined findings from all the analyses is
left to the next chapter.
Licensing and Size
Two dimensions of technology strategy that have been much
discussed in the literature are:
internal versus external
development, and large firm activity versus small firm
entrepreneurial activity.
A previous work by Friar and Horwitch
(1984) has shown that corporations have increasingly been acquiring
technology through external mechanisms such as licensing and joint
development.
This trend has been bemoaned by Hayes and Abernathy
(1980) and others who purport that by using external mechanisms the
long-term competitive viability of a firm is undermined.
On the
-120-
other side, Roberts (1981) discussed that the benefits from R&D
activity can be more fully realized through corporate involvement
rather than through licensing or sale of technological knowhow.
A
firm should, in summary, neither license in or out.
Others have argued (Horwitch (1986), Friar and Horwitch (1985))
that for an industry to be innovative there must be a blending of
large-firm and small-firm modes of development.
Further, as this
blending occurs, there will arise a complex array of linkages and
relationships among the companies in an industry.
Although these linkages must be present in an innovative
industry, no tie-in to individual firm management of these linkages
and performance is hypothesized.
What has been hypothesized by
others is whether a small firm or a large firm is better able to
appropriate innovation benefit.
The debate over firm size and the benefits of R&D has been long
standing.
Schumpeter (1934) claimed that large firms could more
readily benefit from a technological advance because they already
have marketing and distribution economies of scale.
Many others have
argued that firms need to be entrepreneurial and small to both
innovate and benefit from it.
The blending of large and small firms, as Horwitch described,
has occurred in this industry and will be documented.
Then,
questions of licensing in and firm size will be investigated.
As can readily be seen in the company profiles (Appendix 1) and
in Exhibit 6.1, almost all the large medical equipment, medical
instrument, and pharmaceutical companies that compete in ultrasound
have used external methods of technology acquisition
.
(See Friar
Exhibit 6.1
Large Firm Use of External Technology Acquisition
Company
Colgate-Palmolive
GE
GEC
H-P
J&J
Philips
Roche
Rorer
Siemens
Squibb
Toshiba
Acquisition
for Entry
Acquisition
for Technology
Medasonics
Electro Physics
Picker (Physionics)
J&J Ultrasound
Technicare (Unirad)
Rohe
Grumman
Sonometrics
Searle
ATL
Ekoline
Irex, Echo Labs
ADR
Other
External
Yes
Yes
Yes
No
Yes
Yes
No
No
Yes
Yes
No
-122-
and Horwitch (1985) for a discussion of these mechanisms.)
of the major firms has not used external mechanisms.
Only one
Nine of the
eleven firms used the acquisition of small startup firms as their
means of entering the industry.
What may be surprising is that many of the small startup firms
have also acquired technology from external sources (Exhibit 6.2).
Several firms have licensed technology from universities and research
institutes.
Examples are:
ATL, Hoffrel, Unirad, and Imex.
Others
have sourced some of their product line from external sources:
Corometrics, Storz, Ultrasonix, and National.
Firms with as little
as $2 million in sales, moreover, have acquired other firms for their
technology:
High Stoy, Biosound, Fischer, Cooper, Multiscan, Xonics,
and Diasonics.
The diagnostic ultrasound industry, therefore, is characterized
by a rich and varied technology-strategy environment.
Many small
startup fJms are vying for market share against large corporations.
But both the small firms and the lar'e corporations are using a broad
range of techniques to develop and to acquire technology, including
strategic linkages.
Even in an apparently fluid technology-intensive
setting, a blending of previously distinctive modes is taking place.
To analyze whether firm size is important, we used absolute
size and growth as performance indicators.
Of the eight firms who
have greater than $10 million in sales in 1985, five are parts of
large corporations and three are startups.
Firms that are divisions
of larger companies are as likely as freestanding firms to have less
than $5 million in ultrasound sales (Exhibit 6.3).
Of the seven
large corporations in the detailed sample who acquired startups, four
Exhibit 6.2
Smaller Firms That Have Used External Technology Acquisition
Company
Approach
Biosound
Cooper
Corometrics
Diasonics
Acquired Honeywell Ultrasound
Acquired Xenotec
Distribute Pie Data
Acquired Varian Ultrasound,
Distribute Hitachi
Acquired EMI Ultrasound
Acquired Fischer
Acquired Interspec
Licensee of University of Pittsburgh
Licensee of University of Colorado
Acquired Life and Echomed
Distribute Pie Data and Diagnostic Sonar
Distribute High Stoy
Distribute CGR
Acquired Litton Ultrasound and SKI
Fischer
GL
High Stoy
Hoffrel
Imex
Multiscan
National
Storz
Ultrasonix
Xonics
Exhibit 6.3
Relationship Between Status as a Subsidiary of a Larger Firm
and Ultrasound Sales (1985)
Status
Subsidiary
Less than t5million
7
(9.1)
Freestanding
10
Total
17
(7.9)
Sales
Greater than 15million
9
(6.9)
Total
16
X2 = 2.33; not significant at p = 0.05
Expected counts in parentheses
4
(6.1)
13
14
30
-125-
of the companies have lost market share since their acquisition, one
has
maintained share, and only one has gained share but not until 5
years after the acquisition.
(One acquisition occurred after the
data's time frame and was excluded.)
Large firms, then, have not helped the small firms they have
acquired, have not been able to use corporate size to grow in
ultrasound, and have not been able to exclude small startups from
growing large in ultrasound.
Large corporate size would not appear
to be of much importance.
A corroboration of the lack of influence of the size of the
overall corporation is the fact that in every case the ultrasound
groups are separate entities in the large corporations.
Even the
firms that are full-line medical equipment companies have separate
business units for the ultrasound groups, often with the headquarters
located in
different cities from those of the medical equipment
groups (Exhibit 6.4).
Not only is management separate, but the sales
force, to varying degrees, is dedicated to ultrasound products.
The
companies themselves, then, have not tried to gain much marketing
synergy outside of company name.
In analyzing ultrasound companies as separate entities and
using the size of the ultrasound group itself as an indicator of
performance, we found no relationship between licensing and size and
between licensing and status.
Small and large firms were equally as
likely to license in equipment (Exhibit 6.5).
Freestanding firms,
moreover, were as likely to license in equipment as were firms that
were subsidiaries of larger corporations (Exhibit 6.6).
Exhibit 6.4
Headquarters Locations for Ultrasound Divisions of Full-line
Medical Equipment Companies
Company
Ultrasound
Headquarters
Medical Equipment
Headquarters
GE
HP
Philips
Picker
Siemens
Technicare
Toshiba
Rancho Cordova, CA
Andover, MA
Santa Ana, CA
Northford, CT
Iselin, NJ
Ramsey, NJ
Tustin, CA
Milwaukee, WI
Same
Shelton, CT
Cleveland, OH
Same
Cleveland, OH
Same
Exhibit 6.5
Relationship Between Size of Firm and Use of Licensing
License In
Small
(Less than $5million)
No
10
(9.1)
Yes
Large
(Greater than $5million)
6
7
(6.9)
(6.1)
7
Total
17
(7.9)
Size
Total
16
X2 = 0.48; not significant at p = 0.05
Expected counts in parentheses
14
13
30
Exhibit 6.6
Relationship Between Status as a Subsidiary of a Large Firm
and Licensing In
License In
No
9
(8.5)
Yes
7
(7.5)
Total
16
Free
7
(7.5)
7
(6.6)
14
Total
16
14
30
Sub
Status
X2 = 0.12; not significant at p = 0.05
Expected counts in parentheses
-129-
The data would suggest, in summary, that the ultrasound
industry is characterized by a rich array of linkages among firms.
In trying to relate corporate size and licensing to performance, we
found no relationships.
Large firms and small firms were equally
likely to succeed and were equally likely to license in.
Major Improvement Analysis
In a previous section it was argued that most innovations are
incremental advances to technology and that this is the type of
innovation that should be studied.
Nonetheless, innovations that are
considered to be milestones exist for diagnostic ultrasound.
In
order to arrive at a listing of these milestones, the marketing
managers and R&D managers of the ten firms studied indepth were asked
to list what they considered to be milestones.
If an event were
mentioned by at least three people, it was included in Exhibit 6.7.
Although most of the major milestones occurred before 1979 and
therefore antedated the detailed performance data, the time it took
for others to respond could be determined.
Likewise, the change in
overall relative position of the innovating company after the
introduction of the innovation could also be determined.
The first milestone was the introduction of the first
commercial, two-dimensional contact scanner.
The first 2-D scanner
was a bistable B-scan and was introduced in 1963 by Physionics.1
Physionics did not design the unit but rather marketed a scanner
designed at the University of Colorado.
Because several other
researchers had designed 2-D scanners, Physionics did not create a
technical breakthrough.
Also, because the image quality was poor,
Exhibit 6.7
Technology Development Milestones in Diagnostic Ultrasound
Year
1963
1969
1972
1973
1975
1976
1977
1983
Milestone
Commercial, 2-D contact scanning
Mechanical real time
Electronically switched real time
Stored gray scale
Electronic focus
Microprocessor control
Digital scan converter
Computed sonography
-131-
Physionics sold few units and eventually gave the marketing rights to
Picker in 1968.
The second milestone was the introduction of the first
mechanical real-time unit.
As with the first B-scan, the first
real-time unit was designed at the University of Colorado.
The unit
was introduced commercially in 1969 by Magnaflux Corporation but was
unsuccessful in the marketplace.
The next milestone was the introduction of the first
electronically switched real-time unit.
Medical researchers in the
Netherlands developed the first linear array, which was marketed by
Organon Teknika in 1972.
The product was withdrawn a few years
later.
In 1973, Rohe Scientific introduced stored gray scale imaging
into its B-scan unit.
Gray scale is credited with improving
ultrasound image quality to an acceptable level.
As can be seen in
Figure 5.7, the diagnostic ultrasound market began to take off in
1973.
Within the same year, however, Litton had introduced its own
gray scale, and Hughes was supplying all the other manufacturers with
gray scale converters.
Previous to Rohe, other firms had introduced
different solutions to gray scale that were not stored in the
converter but
emulated in the camera, for instance.
imitation,
then, was almost immediate because so many companies had been working
on developing the capability.
Assessment of what competitive advantage Rohe's introduction of
gray scale provided is difficult because no extensive industry data
were collected at the time.
Rohe was the third largest seller of
B-scans at that time and remained so for six years, when they
-132-
introduced a B-scan and sector combination.
Rohe then was able to
gain relative market share in B-scan sales.
The introduction of gray
scale, therefore, did not seem to make an impact for Rohe.
The first electronically focused scanners were introduced in
1975.
Diagnostic Electronics Corporation developed and introduced
the first phased array, which was based on a design by medical
researchers in the Netherlands.
Off the same prototype design,
Grumman Corporation also introduced a phased array in 1975.
A third
company, Varian, introduced a phased array unit in the same year.
Imitators of Diagnostic Electronics appeared immediately.
The
reason, again, was that medical researchers had developed the first
phased array in 1968 and so several companies were independently
advancing the design.
As for market effect, the third company in,
Varian, was able to dominate the phased array segment for five
years.
The other two companies never created a presence in the
phased array segment.
None of the companies extended electronic
focusing into the other modalities.
Because phased array was a
relatively small segment until 1982, Varian never dominated the total
ultrasound market but did reach $12 million in sales in 1981.
The
ultrasound division of Varian was acquired in that year by Diasonics.
Scanners that included microprocessor control were introduced
in 1976 by Searle Ultrasound.
Microprocessors were introduced to
control peripherals, select pre- and post-processing maps, provide
measurements and calculations, and place alphanumerics on the CRT.
Searle introduced microprocessor control in its B-scan and was
quickly followed by others.
-133-
Searle entered the ultrasound industry in 1976 and rapidly made
an impact.
To separate out the effect of its microprocessor-
controlled units is difficult because in 1977 Searle introduced the
first digital scan converter, which is the next milestone.
Digital scan converters improved gray scale images because they
provided stability and a greater number of shades.
They also
provided enhanced pre- and post-processing of the image.
As with
analog gray scale, Hughes was able to supply digital scan converters
to all the other manufacturers in the same year.
In consecutive years, Searle introduced two major advancements
that were quickly copied.
Searle was able to garner a 7 percent
share of the overall market in 1978 and about 15 percent of the
B-scan market.
Searle was never able to garner better than a fourth
position in B-scans, and in fact achieved its highest share in 1978.
After 1978, Searle began losing share and was acquired by Siemens in
1980.
Searle's introduction of two major innovations, then, provided
it with some immediate impact that was soon dissipated.
The final milestone to date is the introduction of computed
sonography.
The image in such scanners is formed in a computer under
software control.
made this possible.
Enhancements to microprocessor capabilities have
Acuson, a startup company, introduced computed
sonography in 1983.
Because of the expense of such units, no one has yet imitated
Acuson.
Acuson has been able to go from
a startup with its first
deliveries in the last quarter of 1983 to one of the leading
companies in dollar terms in 1985.
In units, Acuson is considered
-134-
the leading seller of linear arrays to hospitals, but does not lead
in selling either phased arrays or linear arrays to the total market.
The introduction of a major technological innovation, in
summary, has consistently been made by small startup firms.
Almost
all the innovations were quickly imitated, some so rapidly as to be
simultaneous.
strong.
The market impact for the innovating firm has not been
Of the seven firms and eight milestones, five firms received
little or no market advantages from innovating.
Two firms have
received short-term benefit (Acuson is still too new to determine).
One firm, Varian, that was an almost simultaneous imitator, had a
longer-term market impact from innovating (Exhibit 6.8).
Another point of interest is that each of the innovators worked
basically in one modality.
Except for Acuson, other companies took
the initial innovation and applied it to the other ultrasound
modalities.
An example is that Toshiba and Aloka introduced the
first electronically focused linear arrays; none of the three
original innovators of electronic focus ever entered the other
modalities.
Technology Cycles
Although the ultrasound industry has experienced smooth, rapid
growth since 1973, this sales curve belies the turbulence experienced
by each of the underlying technologies and applications.
As can be
seen in Figure 6.1, the major imaging modalities have gone through
rapid cycles of growth, maturity, and decay in dollar volumes.
same is true in unit volumes (Figure 6.2).
The
The applications for
these technologies, moreover, have been shifting in importance
(Figure 6.3).
Units designed for abdominal/radiology applications
Exhibit 6.8
Benefits from Commercialization of Major Innovations
Amount of Benefit
Company
Little or None
Physionics
Magnaflux
Organon Teknika
Rohe
Diagnostic Electronics
Short Term
Searle
Acuson
Longer Term
Varian
Figure 6.1:
Medical ultrasound sales by modality type ($million)
150
i
I
I
I
I
I
I
125 L.
A
/\
/\
V.
/
100
1
I
I
75
I.
50
2
0
1980
-A
1981 1982 1983 1984 1985 1966 1987 1988 1989 1970 1971
1972 1979 1974 1975 1978 1977 1978 1979 1980 1981
PERIODS
BSCANS
LINEAR
..........
-SECTOR
.
.
. .--
PHASED
Source:
See Appendix 5
1982 1983 1984
Figure 6.2:
Medical ultrasound sales by modality type (units)
9.500 F-"*
T
T~
3. 000
2.500 L
2.000
1. Soo
/
7'
---
--.
---
-
t -
~
~/..
-
5CL II
/
1. 0C I
7/
1975
1976
1977
1Q78
1Q80
1979
1981
1982
YEARS
BSCANS
...M SS
LA
PA
Source:
See Appendix 5
1989
19 84
Figure 6.3:
Medical ultrasound sales by application type ($million)
125
V,
- ""-- -- ---
V -,
!
100
75
50
2
0
1975
1976
.-.
1977
_--
1978
1979
1980
1981
1982
Source:
See Appendix 5
1989
19
84
YEARS
...
.
RADIOLOGY
CARDIOLOGY
OBGYN
VASC
NOW-
-139-
once predominated but now have given way to cardiac applications.
Cardiac and vascular applications are growing while obstetrical and
radiological ones have appeared to level off.
As was discussed in a previous section, ultrasound can be used
for many applications, and the various modalities compete to become
the modality of choice for a given application.
A company competing
in this industry, therefore, can attempt to grow out of a given
segment by trying to serve new customers with the same technology, or
by bringing new technology to its present customer base.
Abell
(1980) discussed at length the strategy options of trying to serve
new customers, develop new technologies, or create new functions for
the technology.
In Exhibit 6.9, a matrix is shown that takes the four main
ultrasound modalities and pairs them with the three largest
applications for ultrasound.
The technologies have been sold for
each of the applications so theoretically a firm could sell, for
example, phased arrays for radiology, cardiology, and obstetrics
applications.
Firms that have had a major share in any given
technology are listed.
a 10 percent share,
if a firm had a minimum presence, defined as
in any application, it is highlighted.
As can be seen in the exhibit, only two firms have been able to
sell the same technology in more than one sector, although almost
everyone has tried.
Likewise, of the firms that have been successful
in more than one technology, only two firms have done so through
internal development.
Exhibit 6.9
Ability of Firms to Gain a 10 Percent Share in the Major
Technology-Application Segments (Through 1983)
Technologies
Applications
Radiology
B-scans
OB/GYN
*
*
*
*
*
GE [Electro Physics]
Philips
Picker
Searle (Siemens)
Unirad (J&J)
Cardiology
*
*
*
*
*
*
*
*
ATL
Diasonics
Hoffrel
Irex (J&J)[Aloka]
v/Philips
Picker [Ausonics]
SKI
*
Mechanical Sector
Phased Array
*
*
*
*
*
*
Acuson
GE [Yokogawal
HP
Irex
-Toshiba
Varian (Diasonics)
Linear Array
*
*
*
ADR
GE [Aloka, Yokogawa]
/Toshiba
/
)
Supplying companies in [ I
Acquiring companies in (
Internal development only
-141-
Within what is considered to be one technology, ultrasound,
there are many modality-application segments.
Only two firms have
been able to expand into new applications with a given technology,
and only two other firms have been able to develop new modalities for
their given application base.
There is a high level of impedance,
therefore, for companies trying to expand into new technologies and
markets.
It is not surprising, then, that a company's fortune rides with
the cycle of one technology if it is not able to move out of it.
The
overall market shares of the industry have been determined by having
the right technology at the right time.
Companies were ranked by overall market share for the years
1979 through 1983.
The first-place company was given 10 points, the
next nine, and so on.
technology.
The companies were then grouped by their
The B-scan companies dominated the market in 1979, the
sector companies took over in 1981, and the phased array companies
were gaining over the period.
As can be seen in Figure 6.4, the
group of companies in a technology moved up or down in the same order
as the underlying technologies.
The rankings of the groups were further tested by using
Kendall's W (W = 0.28, not significant at p = 0.05).
This shows that
the firms could not maintain position over the time period -there was no entrenchment or defense of market share.
that
Rather, firms
moved as their technology cycle did.
The importance of having the right technology designed for the
right application can be further clarified by analyzing the history
of the major technology-application segments.
This analysis
Figure 6.4:
Market share positioning of the groups of companies selling each modality type
4
/
3.5
N.
2.5/
2
-
---
1.
-
>7
5
7
7/
1
~
1979
1980
1981
YEARS
BSCAN
-. - ..... .... LINEAR
MSS
PHASED
2'
K.
.4-
198
2
1983
-143-
considers three dimensions for each technology:
the intended
application at first commercialization versus the actual application
at acceptance; the time from first commercialization to acceptance;
and the length of time from acceptance to peak sales.
Acceptance was defined by the technology reaching a 20 percent
penetration rate of the specific specialty.
The level of penetration
was chosen because diffusion of ultrasound equipment has been shown
to have followed an S-shape pattern (J. Lloyd Johnson (1982)).
Buyers who are considered to be pioneers or experimenters
are
defined as the first 15 to 20 percent of adopters (Rogers (1971)).
Acceptance, therefore, has been defined to mean the point when the
general population starts to adopt.
Using different percentages and
different definitions of the start of the acceptance phase altered
the date by at most one year.
In 3.5 of 4 instances, the major application for an ultrasound
modality was not the one originally envisioned (Exhibit 6.10).
The
one-half was determined in mechanical sector scans; although
cardiology applications began sooner, radiology applications soon
dominated.
The length of time, moreover, from first
commercialization to acceptance has been about ten years, except in
one case.
On the other hand, the length of time from acceptance to
peak sales has only been 3 or 4 years.
The development of a technology, therefore, has required a long
gestation period in which pioneering users have experimented with
various applications and the manufacturers have steadily improved
image quality.
Once a modality reached an acceptable level of
efficacy for a given application, the growth window was relatively
Exhibit 6.10
Original Application, Accepted Application, Time to Acceptance,
and Time to Peak Sales
Technology
B-scan
Linear Array
Mechanical
Sector
Phased Array
Annular Array
Introduction
Original
Application
1963
1973
1969
OB/GYN
Cardiology
Cardiology
1975
1977
Neurology
Radiology
First
Accepted
Application
Radiology
OB/GYN
Radiology and
Cardiology
Cardiology
?
Time to
Acceptance
(years)
Time to
Peak
(years)
11
4
10
4
3
4
8
>9
?
?
-145-
short.
The growth phase was so short because market penetration has
reached a 90 percent level in about a three to four year time frame
(Figure C.5).
Along with the short sales windows, the peaks of the different
modalities have come at only 2 to 3 year intervals.
A company that
is facing a maturing segment must respond almost immediately to enter
the new growth segment.
If a company waits until its market matures
and then tries to shift segments, it faces entering either segments
that will mature very quickly or that will not take off for a number
of years.
An example to further illuminate the dilemma is the response
of
the major B-scan manufacturers to mechanical sector scans.
B-scan
sales peaked in 1978 and mechanical sector scans for radiologists
became accepted in 1979.
By the time the B-scan manufacturers
realized the situation and could develop a new product, which
they
all tried to do internally, the earliest they could deliver
mechanical sectors was in 1981 or 1982.
This, of course, was right
before the peak of sector scan sales.
What has further exacerbated the delay of companies responding
to new technologies is that competition has been framed in
technological terms.
The debate in the medical literature was
whether mechanical sectors would obviate the need for B-scans in
radiology and vice versa.
Today the debate is whether phased array
will supplant mechanical sector, and, likewise, whether annular array
will eventually unseat phased array.
Because the debate is framed
this way, manufacturers have not given up on their technologies and
thus acceded to the upstarts.
None of the major radiology mechanical
Figure 6.5:
Cumulative diffusion of real-time abdominal
imaging capability in 400 to 499 bed hospitals
100%,
Static B
75%-
-Any
50%-
-Linear
array
real-time
(sector or
linear array)
25%
-Sector
I
1973
I
U
1975
I
I
1977
I
I
1979
I
I
1981
Year
Source:
Diagnostic Imaging (October, 1982)
-147-
sector companies have introduced phased array devices for
radiologists.
In hindsight, one can make the argument that each modality has
gone through an independent product life cycle.
All of the
modalities are still in clinical use so one has not obviated the
other.
Each of the technologies has diffused very quickly once a
level of acceptance was reached, and not one has been cut short by a
new modality.
Because the equipment has a useful life of 7 to 10
years, most of the older placements have not needed replacement yet.
Not only do firms have difficulty deciding when to go to a new
technology and thus have a slow response, they also have additional
pressures from the fluidity of their sales forces.
Salespeople have
jumped from company to company, depending on which ones have the
brightest near-term outlook.
Because the buyers of ultrasound equipment are so often
first-time buyers, especially for a given modality, the ultrasound
salesperson must spend a good amount of time training the physician
and spreading clinical information.
This task is similar to that of
pharmaceutical detail men, which has been well documented. 2
Ultrasound salespeople, therefore, must be well-trained in
ultrasound and very often are former ultrasound technicians.
Because
these people are in short supply, trained salespeople are always in
demand. Several aggressive startup firms have had explicit policies
of skimming the cream of their competitors' sales forces.
One
company had an explicit policy of only interviewing candidates who
had earned at least $70,000 in commissions the previous year.
The
-148-
raiding became so bad at one point that Philips actually sued
Diasonics over the issue.
Salespeople will look to another company if their present
company does not have a unit available, or shortly available, in the
"hot" technology-application segment.
B-scan companies lost many of
their salespeople to mechanical sector scan companies.
The sector
scan companies, moreover, have recently been raided by phased array
companies.
One company that is about to introduce an annular array
has admitted that it has already been talking to the better
salespeople of the phased array companies.
A company faces double pressures.
As its core technology
matures, it must decide when and how to change to a new segment.
In
the meantime, however, its sales force may become quickly depleted,
which means that sales spiral downward even faster because it cannot
sell its present equipment.
The leading companies in an old segment,
moreover, are hit harder than their competitors, and therefore may
lose relative market share in their core segment because of
organizational dismemberment.
A company needing to introduce new products often faces the
double task of both developing a technology and redeveloping its
organization.
Because of this phenomenon within diagnostic
ultrasound, companies will rehire salespeople even after they have
quit and gone to a competitor.
As an example, GE will rehire a
person three times before she has worn out her welcome.
The importance of having the right technology, in summary,
cannot be overstated.
Firms with the growing technology take over
total market dominance; firms without it face quickly maturing
-149-
segments and collapsing sales organizations.
The impedance to a firm
trying to enter into a new segment has been very high, especially
when the shift is attempted through internal methods of development.
Competitive Orientation
As was discussed in Chapter 2, the two generic strategies
available to a firm are to compete on price or differentiate.
But
from the discussion in Chapter 3, differentiation is too broad a
category and belies the organizational differences behind the various
types of differentiation.
Two organizational competencies that have
been shown to be of importance in the medical equipment industry are
R&D and marketing.
These competencies are assumed to be the thrusts
that can lead to differentiation, but are also thought to be
exclusive.
This latter hypothesis will be tested.
Competitive orientation, therefore, can be thought of as a
trade-off in competing through innovation, through pricing, or
through sales and promotion.
The technology-strategy decision for a
firm becomes one of positioning the organization as to which approach
to emphasize.
The question still remains of whether these are really
exclusive positions or ones that can be a combination of two or
three.
In other words, can a firm try to both innovate and compete
on price, or have strong R&D and strong marketing?
What needs to be
tested is whether firms mainly compete in one area or try to balance
among more than one.
A further question to be asked once the competitive
positionings are determined is whether having a competitive stance is
important from an industry-wide standpoint.
Do all the companies
-15C-
within a given strategic group outperform the other groups or not?
Or is it a matter of actuating well whatever strategy one chooses?
To test these questions, respondents were asked to allocate
budget expenditures among marketing, R&D, and price promotions, as to
their relative importance for improving sales.
for proportions that would total 100 percent.
The question asked
Strategic expenditures
have been shown to be relatively constant 3 (and this will be
further corroborated in the next section).
Asking for the relative
importance of strategic expenditures at one point in time should
provide the relative standing over the firm's recent history.
A further point to consider is whether one respondent can
actually know what a firm considers to be relatively important.
To
counter this possible weakness, in half of the cases more than one
respondent was asked his opinion on the strategic expenditures for
each firm.
This was done with the respondents sitting together and
debating the figures.
In the other cases, the surveys were completed
by the manager who would have the final approval or a strong
influence in actual budget decisions.
This person would have been
privy to any discussions on setting overall strategy.
These relative expenditures were tested using cluster analysis
to see if the firms group together in any way along the three
dimensions.
For a discussion on the use of cluster analysis to
determine strategic groups, see Harrigan (1985).
The clustering
method used was the Howard-Harris clustering technique.
The
significance of the clusters was tested by using the Calinski and
Harabasz ratio (Everitt (1980)).
to company size and growth.
The clusters were then compared as
-151-
Four significant clusters were found and are broadly defined as
those companies who emphasize marketing, or R&D, or both marketing
and R&D equally, or who thought that price was important (Exhibit
6.11).
No company emphasized price as the predominate approach, but
the companies in the last cluster felt that price was relatively more
important than the other groups did.
A three cluster break out could also have been taken, which
would have consisted of groups who emphasize only one dimension -marketing, R&D, or price.
Because the test statistic does not really
discriminate between three or four groups, however, we elected to use
four groups.
The fourth group was almost entirely a subset of the
marketing group but differed because its emphasis on marketing was
matched by an equal emphasis on R&D.
Likewise, the one company that
was pulled out of the R&D group and put into the fourth group had a
similar balance between R&D and marketing.
Using the extra group,
therefore, seemed to make sense.
That the companies' self-expressed positioning as to generic
strategy gives rise to distinct clusters is consistent with many
other studies as to the presence of distinct generic groups in an
industry.
That companies also perceive themselves as being different
because of their generic positionings gives support for the use of
such classifications.
What needs to be further determined is whether
these classifications help predict performance at the industry level.
Two tests of performance were used:
whether a company had
grown to any significant size in the industry and whether companies
were growing faster than the market.
As can be seen in Exhibit 6.12,
Exhibit 6.11
Cluster Analysis of Relative Emphasis on R&D, Marketing, and
Price (N=25)
1
2
3
4
5
*peak values
Total Sum of
Squares
74.9999
37.9981
24.4849
17.7361
14.2185
C Ratio
-
Number of
Clusters
22.39
22.67*
22.66*
21.41
Exhibit 6.12
Self-Evoked Orientation Clusters and Performance Measures
Cluster Orientation
Marketing
Centroids
(normalized)
MTKG
RD
Price
1.05508
-1.13292
-0.193626
-0.754281
1.08386
-0.45056
8
6
Numbers in
Clusters
Proportion of
firms with greater
than t5million in
sales
Performance
R&D
.375
-,-,
+ grew faster than market
0 grew with market
- grew less than market
.50
Price
-1.17068
0.54373
1.68714
5
.40
Marketing/R&D
0.323076
-0.026405
-0.697218
6
.67
-154-
the number of small and large companies divided randomly among the
four groups. Although not statistically significant when tested as a
binomial distribution, the marketing and R&D group had the highest
proportion of companies who have attained large market size.
Growth could only be tested on a subset of the companies
because we did not have sales data over time for all of them.
It was
fortuitous that the companies with known growth figures were evenly
divided among the four groups.
states:
Companies were divided into three
those who grew faster than the market, those who grew with
the market, and those who grew slower or lost sales.
It is interesting to note that the three companies in the
marketing and R&D group grew faster while those in the marketing only
group did not keep up.
There is some evidence, then, that having a
balance between R&D capability and marketing is stronger than having
an emphasis in only one area.
Technology-Market Analyses
Overview
At the level of the industry, the previous analyses have
considered the relationship of strategic options to market
performance.
The most significant finding has been the need to have
the right technology at the right time --
companies selling a
specific modality have dominated or fallen together with the
underlying technology cycle.
The difficulty for a company of moving
from one modality or application to another has also been
illustrated.
The other analyses have not shown much about individual
firm performance except that there is some evidence that companies
-155-
using an equally balanced approach between R&D and marketing have
performed better.
A shortcoming of the self-evoked positioning analysis is that
such positions may not be maintained in each segment.
An example is
that a firm may want to be a technology leader but cannot possibly do
so in all the various segments it chooses to enter.
A further
illustration of this possibility is the ratings of image quality4
for firms in several segments (Exhibit 6.13).
Only one firm has
maintained high relative rankings (top 3) by itself in more than one
segment.
The other companies that have more than one product in the
top three have done so through acquisition of another firm or outside
sourcing.
In analyzing market performance at the technology-application
level, this section considers the relative positioning a firm
actually achieved in each segment as opposed to its overall plan.
The following analyses test whether relative competitive position has
changed in response to technological improvements or in response to
other market conduct variables.
Not only have relative market shares
shifted at the industry level, they have shifted within the six major
segments.
These segments are listed in Exhibit 6.14.
At the segment level the technological competition has been one
of incremental product improvements.
Competition within each segment
has been intense because of the presence of so many firms.
In the
B-scan segment twelve firms competed at one time; in mechanical
sector-radiology, fourteen; in linear array, sixteen.
Exhibit 6.13
Rankings of Image Quality for a Company's Equipment
in Different Segments
Company
A
B
C
D
E
Rankings in Each Segment
1, 6, 8
1 2, 3*
3 , 8, 1
2, 7, 5, 1*
5, 4
F
G
H
4, 9, 2*, 6, 2*
I
6, 5, 3, 5
*Products in top 3 that were externally sourced
Exhibit 6.14
Modality and Application Combinations Used in the Study
Modality
Application
Mechanical Sector
Mechanical Sector
Phased Array
Phased Array
B-Scan
Linear Array
Radiology
Cardiology
Radiology
Cardiology
Radiology
Obstetrics
-158-
With the number of companies selling in each segment, customers
have become confused.
As was previously discussed, Friar (1984)
found that customers actively purchasing equipment knew only a small
proportion of the existing companies.
Of the companies the customers
did recognize, they could not differentiate the equipment as to image
quality.
Because customers have had difficulty differentiating equipment
as to image quality, the relationship between functional performance
and market success is not straightforward, if there is one.
As a
surrogate for the users' ratings, the manufacturers' ratings of image
quality were compared to the average market share rankings of the
firms in each segment.
Although the significance level is not
strong, the market leaders tend to have units with higher image
quality, as judged by their competitors (Exhibit 6.15).
Technology
capability, therefore, is still an important variable to consider
even if customers have difficulty seeing it.
Two different tests of performance were performed at the
segment level.
One test analyzed performance as defined by the
number of units sold over a five year period.
The second test
compared companies that superseded the segment market leaders.
The
independent variables in both cases were indicators of strategic
posturing as discussed in the framework section --
R&D, marketing,
and price.
The strategic variables or market conduct alternatives used in
this analysis were culled from the literature and tailored by
prestudy interviews.
(See data collection section.)
These variables
are listed in Exhibit 6.16 and will be discussed below.
Exhibit 6.15
The Relationship Between Image Quality Rankings and Market
Share Rankings
Friedman Two-Way Analysis of Variance
Image Quality Rank
1
2
3
4
5
A
1
2
3
4
5
B
4
2
5
3
1
C
1
3
5
4
2
D
2
1
3
4.5
4.5
E
1
3
2
5
4
F
1
4
2
3
5
Ri
10
15
20
23.5
21.5
Segments:
r = 7.96*, 4 degrees of freedom, p
=
.10
Exhibit 6.16
Strategic Variables Used in Creating Indices of Competitive Orientation
Variable
Direction of Stronger Orientation
R&D
Entry into segment
Rating of image quality
Number of updates
Type of development
Earlier
Higher
Greater
Internal
Marketing
Sales force size
Sales technique
Service force size
Service force technique
Advertising and promotion
Product assortment
Larger
Dedicated
Larger
Dedicated
Greater
Greater
Pricing of product
Pricing of service
Financing
Lower
Lower
Available
Price
-161-
Many authors have argued that the strategic posturing of a firm
--that is the relative importance of one functional area to another-remains relatively stable over time.
Fry (1984), and Tassey (1983)).
(See for example Zeithmal and
They argue further that four or
five-year averages should be analyzed rather than yearly change
variables.
There is some limited corroboration of this relative
stability for this industry (Exhibit 6.17), and so averages over five
years were used in the first analysis.
Although
a firm's posture may stay stable within itself, its
relative market performance has not.
Because relative market shares
have changed in each segment, looking at average performance over the
time period does not capture the dynamics of competition.
Instead of
using average performance, therefore, the second analysis looked at
changes in relative position over time.
Indicators and Measures
The R&D index consisted of summing three quasi-Likert ordinal
scales and one raw score.
The companies in each segment were rank
ordered as to entry into a segment and as to image quality
ratings.4
Entry and quality ratings were both needed to determine
not only whether a firm was offensive, but also what type of follower
(imitator, defender, etc.). 5
Development was also rated on a scale
.
depending on level of firm involvement 6
a.
Internal duvekopmient
b.
Technology acquisition with major internal
c.
Evenly split between external and internal
d.
Mostly external
e.
Completely external
Exhibit 6.17
Relationship of Advertising to R&D over 5 Years
Company
Range of Proportions
A
.10 - .15
B
C
D
E
F
.09
.10
.20
.27
.26
- .13
- .12
- .26
- .33
- .27
-163-
The raw score was for the number of updates introduced for each
product.
This was included as an indicator of continuing commitment
to improving product performance.
An interesting side note is that the companies with the highest
quality ratings (4 or greater) introduced 90 percent of the updates.
This means that companies that already had the first or second-rated
equipment continued to improve it.
rarely upgraded or improved upon.
The lesser-rated equipment was
Since companies had such
difficulty getting into the top two ratings of more than one segment,
the same companies were not upgrading across the board.
Rather, the
performance leaders in each segment tried to continually improve
while the others watched.
The aggregate index of R&D orientation was computed by summing
the four scores.
The reliability of the additive index based on the
above four measures was sufficiently high to justify its use as a
measure of a single construct (Cronbach's alpha equal to 0.76).7
The marketing index also consisted of the summing of three
quasi-Likert ordinal scales.
The assumption for industrial sales is
that a salesperson needs to spend a large amount of time with
relatively fewer accounts.
The larger the sales force, then, the
stronger the sales orientation.
size of the service organization.
reflect sharing of services.
Likewise, the same is true for the
Both of these were adjusted to
(See Appendix 7).
Dollars spent on
advertising and promotion were adjusted by segments and rated on a
five point scale.
0.91.
Again, the Cronbach alpha was sufficiently high,
-164-
The assortment of products offered by a firm in a segment was
at first thought to be another measure of marketing orientation
because of a desire to further segment the market.
No real changes
in technology were required to put forth a variety of products.
This
measure, however, reduced the reliability score and was therefore
excluded.
The price index was straightforward in that the total price of
an average unit was calculated for each year.
Total price consisted
of the unit price, the service contract, and any financing benefits.
The median price in each segment in each year was determined and
points were given for each $1,000 below or above the median another
product was.
The points were totalled and averaged in the following
manner for each company in a segment:
raw score =
where:
-
median -price(j)
n = number of years
j = specific company
The scores were then rank ordered.
The indices were tested for normality using the Ryan and Joiner
test.8
All three were found to be normal at the p = 0.05 level
(Appendix 8).
Average Performance Analysis
Performance was measured as the average of the units sold over
a five year period.
To improve linearity, the natural log of the
scores was used and again was tested for normality.
A second
performance measure, the rank order of the total number of units sold
-165-
over the period, was also tested for linearity and normality, and was
found to be so (Appendix 8).
The model to be tested is an attempt at explaining market
performance by the relative mix of expenditures and focus that firms
actually made over the time period.
To test whether the data could
be pooled across segments, dummy variables were included for each
technology-market segment. 9
The equation then became:
Per = po+ P1R + P2M +P3P + P4D1 + PsD 2 + P6 D3 + P7D 4 + P8D 5
where: Per = performance (log e, ranking)
R = R&D index
M = marketing index
P = price ranking
D 1 through Ds are dummy variables
Linear regressions were run using ordinary least squares techniques
and the residuals were tested for randomness.
Results
The results were consistent using both definitions of
performance.
The fully described equations had none of the dummy
variables significant so the pooled data were rerun without them.
The explained variance was very high in every case (Exhibit 6.18 and
Appendix 9).
The only significant variables in the runs were the R&D and
marketing indices; price was not significant in any of the runs.
That price was not significant corroborates earlier statements about
Exhibit 6.18
Regression of Performance on Indices of Competitive Orientation
log e
ave. unit sales
Constant
R&D
Marketing
Price
D3
N
R2
F
DW
1.999
(1.79)
0.14079
(3.48)**
0.20074
(4.09)**
-0.0678
(-0.51)
1.0173
(1.93)
-0.4944
(-0.99)
0.1669
(0.31)
-0.3667
(-0.64)
-0.4795
(-0.91)
1.4559
(2.26)*
0.12603
(2.97)**
0.19254
(3.59)**
-0.0970
(-0.67)
rank of total
sales
-0.6322
(-0.90)
0.20071
(4.72)**
0.23422
(4.54)**
0.0019
(0.01)
-0.7767
(-1.39)
0.20547
(5.57)**
0.23506
(5.06)**
0.0127
(0.10)
0.1338
(0.24)
0.1281
(0.24)
-0.2082
(-0.37)
-0.3278
(-0.55)
-0.2007
(-0.36)
30
72.4%
30
57.1%
30
79.0%
30
77.8%
7.21*
2.34
11.98*
10.34*
2.20
31.54*
2.12
ratio in parentheses
= 0.05
= 0.01
1.80
-167-
the industry in that there is little price sensitivity.
As further
support, in 60 percent of the cases in this data, the market leader
or company that became the market leader had the highest priced
unit.
In two-thirds of the cases after a company became the new
market leader, it raised prices to then have the most expensive
units.
That both marketing and R&D together are significant provides
further corroboration that a combined approach is needed --
companies
that perform better have stronger investments in both areas.
This
finding also refutes the idea that this market is strictly
technology-driven; both capabilities are significant.
This analysis, however, only looked at average performance -on average over a five year period, the better companies have had
strong capabilities in R&D and marketing.
This analysis did not
compare changes in relative performance, which is what will be
discussed next.
Paired Comparisons
There were eight cases across five of the six segments of
companies that entered the market and wrested away market leadership
from the previous leader.
There was also one case of the market
leader not becoming unseated, and so the next best company was
compared to the leader.
This provided nine cases of paired
comparisons.
The three indices of R&D, marketing, and price were used in
comparing relative position.
The differences between the winner and
loser are listed in Exhibit 6.19.
As can be seen in the exhibit, an
advantage in marketing was a factor in every case and in pricing in
Exhibit 6.19
Paired Comparisons of Strategic Orientation of Overtaking Firm
to Unseated Firm (Index (overtaking) - Index (unseated))
Case
R&D
Marketing
A
B
C
D
E
F
G
H
1
+2
-9
-11
-4
+1
+3
+4
-5
-2
+4
+8
+4
+10
+2
+7
+4
+5
+5
Mean
standard deviation
T for m = 0
P value
Price
0
+3.5
+3.5
+2
+1
+1
+3
0
+2
-2.33
5.44
5.34
2.46
1.37
-1.31
6.65
3.89
0.001
0.005
N.S.
1.78
-169-
most of the cases.
In over half the cases, however, the companies
that became market leaders had an R&D disadvantage to the companies
they displaced.
In this analysis, technology was not a significant
factor while marketing and pricing advantages were.
Summary
By looking at performance in two ways --one being average sales
over a five year period and the other a paired comparision of a
winner and a loser-- some interesting results appear.
In the first
analysis, stronger investments in both marketing and R&D were
required for better performance.
In the paired comparisons,
marketing and pricing were important.
What is significant about the
two concomitantly is that marketing was important in both analyses,
while R&D was important only in one.
A strong investment in R&D
alone was never significant, and an R&D disadvantage could be
overcome with stronger marketing and pricing.
The significance of
these findings, and of all the findings in this chapter, will be
discussed further in the next chapter.
-170-
CHAPTER ENDNOTES
1.
Klein (1980), Hamilton (1982), Berggren (1985).
2.
See for example: Coleman et al. (1966), Frost and Sullivan
(1982), McKay (1983).
3.
Tassey (1983), Zeithmal and Fry (1984).
4.
Ratings of competitors' equipment performed by application
specialists from ultrasound companies. See Friar (1984).
5.
Freeman (1982).
6.
Friar and Horwitch (1985).
7.
Cronbach (1951).
8.
Ryan and Joiner (1982).
9.
Hannan and Young (1977).
-171-
CHAPTER 7
CONCLUSIONS
Discussion of Findings
This thesis has focused on the strength of product innovation
as a competitive weapon for the innovating firm.
Several authors
have avered that innovation is the strongest and most direct way to
achieve advantage, but studies trying to analyze such a relationship
have generated conflicting results.
It was argued that the reason
for the conflicting results is the lack of inclusion of the marketing
perspective --the analysis of the reactions of potential customers to
innovation-- in the studies.
A framework was presented that included
both the customers' ability to perceive technology differentiation
and the defensibility of such innovation.
It was hypothesized that
only in rather specific instances will a technological advance lead
to a viable competitive advantage.
In the other cases, a viable
competitive position can only arise through price competition or
differentiation created from capabilities in other functional areas.
Although some authors have posited that a firm can have only one
functional strength, the relationship of the alternative functional
capabilities to each other was further tested.
The diagnostic ultrasound industry was selected for study
because it epitomizes intense technological competition.
characteristics are:
Some of its
high levels of R&D expenditures, many new
product introductions, and shifting market shares.
If innovation
leads directly to competitive advantage, this relationship should be
demonstrable in this quintessential high-tech industry.
The data
reveal, however, not a direct link but rather a changeful
entwinement, a pavane that is intricate, delicate, and complex.
-172-
The major improvement analysis demonstrated that innovation has
not been defensible in the medical diagnostic ultrasound industry
(Exhibits 6.7, 6.8).
Major technological advances were copied
immediately and the pioneers were not able to sustain any long-term
advantage.
The age of the companies in the industry, moreover, had
no relationship to market share (Exhibit 5.7).
The older companies,
therefore, have not been successful at establishing defensible
positions.
In an earlier study, the author found that customers could not
differentiate ultrasound equipment on a performance criterion.
Technological innovation in this industry has been shown to be
neither defensible nor perceptible.
It was hypothesized that the
technical dimension of competition should abate, but to what relative
position was further examined.
That technology is important was shown in the analysis of the
technology cycles.
Having the right technology designed for the
right application explained much of the shift in industry market
shares (Figures 6.1 to 6.4).
Developing the correct union of
technology and application, moreover, was demonstrated to be
difficult and risky:
the gestation periods of modality development
to an acceptable level were long; the originally intended
applications were wrong; and the technology pioneers rarely benefited
(Exhibit 6.10).
In the test of the relationship of competitors' ratings of
image quality to market share, a weak relationship was found (Exhibit
6.15).
Better performing companies had better performing equipment.
Firms were unable, however, to transfer easily their performance
-173-
leadership in one segment into the other segments (Exhibit 6.13).
The technical dimensions of competition were proven to be important,
therefore, but they were also shown to be insufficient in themselves.
Technological competition in diagnostic ultrasound took place
on two distinct levels:
entering the various modality-application
segments and competing once there.
The use of a strong R&D
orientation, in and of itself, was not sufficient for creating
differential advantage.
Firms had little success in broadening their
business scopes through internal R&D development.
Once in a segment,
moreover, firms with a balanced approach between marketing and R&D
outperformed those firms with a more narrow focus on R&D.
The
technical dimension was not the primary source of competitive
advantage on either level.
Because of the short sales cycles in the ultrasound segments,
companies had to enter new segments to maintain growth.
In trying to
broaden the scope of their businesses, firms encountered tremendous
difficulty.
Part of the difficulty in expanding from one technology
into others can be explained by firms focusing on technological
competition --worrying whether one modality will become the dominant
design-- rather than focusing on serving customer groups and their
specific applications.
Once the firms realized that they needed to
broaden their scopes, they then began to use external methods of
technology acquisition.
In the analysis of licensing and acquisition, a large
proportion of the firms were found to be engaged in outside sourcing
of technology (Figure 5.5, Exhibits 6.1, 6.2, 6.5, 6.6).
The
majority of firms that entered new segments and achieved highly rated
-174-
equipment and/or a significant market presence did so through
external technology acquisition.
Firms with large market shares were
as likely to license in technology as not.
What became important for firms trying to enter new segments
was not a stronger emphasis on R&D but on marketing.
Firms began to
acquire external technology; in addition, they needed to manage fluid
sales forces.
Firms either had to maintain extant sales forces or
build new ones quickly.
parts of marketing.
These additional skills are considered to be
Hitt, Ireland, and Palia (1982) claim that the
organizational strengths required for acquisitive growth are in
distribution and finance.
internal growth.
They claim that R&D is essential for
The shift by companies to outside sourcing,
therefore, shifted strategic orientations away from technology
thrusts.
It was important for a firm to have a presence in the right
technology segment, but it was not important how a firm got there.
Within the specific modality-application segments, firms that
performed better were shown to have strong R&D capabilities, but not
necessarily the best (Exhibits 6.18 and 6.19).
Firms that won
segment leadership were often at a disadvantage to the firms they
unseated as to technical capability.
In every case, however, the
ascending firms had stronger investments in marketing.
In three
other tests, firms that held marketing and R&D in equipose
outperformed the firms with different competitive orientations
(Exhibits 6.11, 6.12, 6.18, Appendix 9).
That all these test results
gave the same indications --that marketing and R&D capabilities
combined are characteristic of high performers-- supports the
-175-
hypothesis that unidimensional competition is not enough.
A combined
approach is needed.
This finding is counter to the statements of O'Shaughnessy
(1984), Rothschild (1979), Tassey (1983), Snow and Hrebiniack (1980),
and others who have argued that companies with an R&D orientation
cannot or do not have an equal strength in marketing.
This finding
is also counter to Porter's (1983) statement that in industries where
technological change is rapid the technological dimension is the
primary source of competitive advantage.
A technological orientation
by itself was never shown to be viable.
This finding reinforces the conclusions of the SAPPHO study and
Miles and Snow (1978) in that the marketing dimensions of competition
were very important for the commercial success of innovation.
This
should not be surprising, however, given the discussion in the
framework section.
If an innovation provides a technological
advantage that is neither defensible nor perceptible, then any
attainable competitive advantage must come from combining strengths
on other dimensions.
In the case of diagnostic ultrasound, the other
dimension was sales and promotion.
The intricacy of technological
competition in diagnostic ultrasound arises from the need to have a
strong R&D capability combined with an equal marketing capability.
As with R&D alone, a marketing orientation alone did not prove to be
effective.
The two are enmeshed.
The more one investigates this industry, which is
hype:-sensitive to technology, the more one realizes that competition
is much more complex than funneling more resources into R&D.
The
main reasons for this complexity are that customers did not perceive
-176-
differences in the technical dimensions of the equipment in specific
ultrasound segments and innovation has not been defensible.
Because
technology alone was not enough, equal skills in marketing were
essential.
As firms, moreover, tried to broaden their scopes,
additional skills in marketing became even more important.
The findings of this study should be generalizable to
industries with similar characteristics of technology development and
performance evaluation.
The long development times and times to
acceptance, the mistaken original applications, and the large number
of companies creating new products are consistent with other
high-tech industries.
See for example the Freiberger and Swaine
(1984) history of the personal computer industry or the Rosenbloom
and Abernathy (1982) history of the video tape recorder industry.
The most essential industry characteristic, however, customer
inability to evaluate functional performance, also holds for many
industries.
That technology cannot be judged through objective assessment
of functional performance even though a comprehensive preselection
process occurs is true for other industries.
In consumer goods many
tests have shown that customers cannot tell apart products such as
beer, cars, or ice cream when taken from the same price class and
with labels removed.
But even more exhaustive tests and
sophisticated testers sometimes get fooled.
Consumer Reports once
rated as having different performance levels the same videocassette
recorder sold under two different brand names (Nohria, forthcoming).
Even in industries that are older and more stable, performance
criteria are hard to judge.
For judging power and speed in mainframe
-177-
computers, there is still debate over what to judge.
Benchmarks,
known by such terms as MIPS, MFLOPS, Whetstones, and Linpacks, can
have contradictory and misleading results.
There is also debate over
whether these conflicting standards or benchmarks actually test
anything close to actual usage.
The claims of improved performance
in mainframe computers, therefore, are really impossible to judge
because no measure can predict how fast a computer will do hundreds
of different jobs under various conditions.
The results of the study are generalizable, therefore, to
industries in which technical performance is not easily judged or is
judged to be identical across competitors.
Further Research
Any peroration should not only emphasize the findings of a
study but also the weaknesses, which will suggest areas of further
research.
A strength of this study is the level of detail reached by
analyzing indepth a single industry, but this is also a weakness
because it is still only one industry.
More industries must be
studied to determine which organizational thrusts or combinations
lead to true differentiation when innovation by itself is not viable.
This study only examined one cell of the framework put forth in
Chapter 3.
Industries with different characteristics of technology
advance perceptibility and defensibility must also be studied to test
the framework.
Pooling of data from studies of industries with
similar competitive characteristics of consumers' perceptions may
provide fewer ambiguous findings of positive returns for innovating
firms.
The level of analysis needed to understand competition in
diagnostic ultrasound was at the specific modality-application, which
-178-
is a level of detail not found in cross-sectional studies.
Because
customer perceptions can only be understood at this level, the
viability of a competitive thrust can only be understood at this
level.
An example of another industry in which innovation has not been
defensible is computer memory chips.
An industry of intense
technological competition, it experiences fantastic price
competition.
In its first year of introduction, the 256K RAM dropped
in price 75 percent (Business Week, May 20, 1985).
The need for
strong capabilities in manufacturing as well as R&D would be a
reasonable hypothesis to test.
At the other extreme, industries such as petrochemicals and
pharmaceuticals have seemingly enjoyed strong patent protection.
As
companies in these industries move into biotechnology, in which
innovation is harder to defend, a reasonable hypothesis would be that
a different mix of functional capabilities will have to be developed.
The move by pharmaceutical firms into biotechnology raises
another interesting issue --
the ability of firms to enter new
technology and new market segments.
The finding of the inability of
firms in diagnostic ultrasound to move easily into new markets or new
technologies reinforces the findings of Roberts and Berry (1985), and
Meyer (1986).
What is interesting to note is the level at which
impedance to shifting technologies arises -modality level.
at the ultrasound
Uitrasound firms have had difficulty developing
different types of ultrasound products.
Firms have been incapable of
developing highly rated equipment in more than one segment.
Likewise, they have had difficulty in garnering large market shares
-179-
in more than one segment.
Because of this, firms have used external
methods of technology acquisition and have shared technology.
Many of the major pharmaceutical companies have created
strategic linkages both with other large firms and with startup firms
to develop technology (Friar and Horwitch (1985)).
As these linkages
for jointly developing and sharing technology continue to develop,
further research into the sharing of technology and its effect on
competitive position is needed.
There are many reasons why firms would share technology, but
one hypothesis is that they do so when technology itself is not that
important in creating long-term competitive advantage.
Bain (1968)
analyzed technology sharing in the automotive industry.
He found
that General Motors licensed technology to many of its competitors.
He concluded that GM did so because it gained no real competitive
advantage by keeping the technology inhouse.
When the technological
dimension of competition does not provide a viable thrust, then firms
are more likely to share technology.
The evidence for the diagnostic
ultrasound industry tends to support this hypothesis.
Several other research questions could be generated from using
the market perspective in analyzing technological competition.
A
final example is that industry development may be determined by
customer perceptions of technology.
When the buying public has
little familiarity with a technology, many competing alternatives vie
in the marketplace, with the technical dimensions waning in
importance.
As customers become more experienced, tastes may
converge so that a dominant design will emerge, and competition will
shift to price and manufacturing considerations.
If tastes do not
-180-
converge, or again diverge, a dominant design may not exist.
The
likelihood of a dominant design arising, therefore, may have the
relationship to increasing customer sophistication of an inverted
U-shaped curve.
The most essential point for further research is that the
marketing perspective must be included.
Not only must research
consider potential customer reactions to a new generic technology,
e.g. personal computers, but also consider whether potential
customers can differentiate the myriad alternatives within a
technology type.
Rarely is technological competition monolithic, in
that the first firm to design a technology-type wins.
Rather,
competition in industries where the technology is changing rapidly
consists of many firms with diverse technology alternatives.
Customer reactions must be considered in the attempt to create
product differentiation through innovation.
Managerial Implications
The recommendation for researchers also applies to managers -to consider the marketing perspective when competing through product
innovation.
A narrow focus on R&D led competition will only work in
certain circumstances.
The rest of the time, the managerial
challenge is much broader and more difficult.
Firms should test potential customers on whether they really
can differentiate technology dimensions of performance.
Although
several surveys of physicians found that they rated performance of
diagnostic ultrasound units as the number one purchasing criterion,
when they were actually tested on it, they could not tell any
difference.
The technical dimensions of competition remained
-181-
important, but not singularly so.
The management challenge that
occurred in this superfluid technology-intensive industry was that of
balancing skills in R&D and marketing, a posture that is not easy to
maintain.
Another managerial challenge that arose was the difficulty of
entering new technology and new product segments.
A revealing
finding was that firms could not maintain technology leadership in
very many segments.
The lengthy development times and short sales
growth cycles of each segment created a most difficult situation for
broadening business scopes, and yet these short sales cycles required
firms to shift segments if they did not want to shrink.
The result
was the need to manage strategic linkages for the development and
acquisition of external technology.
Diagnostic ultrasound firms used
the full range of technology development and acquisition approaches.
(See Friar and Horwitch (1985)).
The lesson gleaned from analyzing the diagnostic ultrasound
industry is that a singular focus on developing new technology must
be augmented by a richer array of strategic choices such as
technology acquisition and strategic linkage options.
A stronger
interplay of the various functional strategies must occur and is
determined by the customers' perceptions of technology.
Once the
marketing perspective is brought into play, the viability of a
technology thrust can be determined.
Because of the physicians'
inability to differentiate products on technology criteria, in this
technologically-sensitive industry, technological advance has been
necessary but not sufficient for market success.
-182-
REFERENCES
Abell, Derek. Defining the Business.
Prentice-Hall, Inc., 1980.
Englewood Cliffs:
Abernathy, William J. The Productivity Dilemma.
Johns Hopkins University Press, 1978.
Baltimore:
The
Abernathy, William J. and P.L. Townsend. "Technology, productivity and
process change," Technological Forecasting and Social Change, 7.4,
1975.
Abernathy, William J. and James M. Utterback.
innovation," Technology Review, 80, 1978.
"Patterns of industrial
Abernathy, William J. and Kenneth Wayne. "Limits of the learning
curve," Harvard Business Review, September-October, 1974.
Abernathy, William J. and Kim B. Clark.
"Innovation: Mapping the
winds of creative destruction," Research Policy, 14, 1985.
Allen, Thomas.
Press, 1977.
Managing the Flow of Technology.
American Hospital Association.
Cambridge:
Hospital Statistics.
MIT
Chicago:
ARA,
1983.
American Hospital Association.
Chicago: ARA, 1983.
Guide to the Health Care Field.
Ansoff, H. Igor and John M. Stewart. "Strategies for a
technology-based business," Harvard Business Review,
(November-December), 1967.
Bain, J. Industrial Organization.
Inc., 1968.
New York:
Bain, Joe S. Barriers to New Competition.
University Press, 1971.
John Wiley & Sons,
Cambridge:
Harvard
Barley, Richard Stephen. "The professional, the semi-professional,
and the machines: The social ramifications of computer based imaging
in radiology," M.I.T. Ph.D. Thesis, 1984.
Bass, Frank M., Phillippe Cattin and Dick Wittink. "Market structure
and industry influence on profitability," Strategy and Structure
Equals Performance, ed. Hans Thorelli. Bloomington: Indiana University
Press, 1977.
-183-
Bass, Frank M., Phillippe Cattin and Dick Wittink. "Firm effects and
industry effects in the analysis of market structure and
profitability," Journal of Marketing Research, 15, 1978.
Bean, Alden S., et al. "Involvement of R&D in corporate strategic
planning: Effects of selected R&D management practices," Strategic
Management of Industrial R&D, eds. Barry Bozeman, Michael Crow, and
Albert Link. Lexington: Lexington Books, 1984.
"CT scanning and ultrasonography: A comparison of two
Berggren, Ulf.
lines of development and dissemination," Research Policy, 14, 1985.
Bernstein Research.
document, 1982.
"Ultrasound market analysis," prospectus
Biggadike, E. Ralph.
"The contributions of marketing to strategic
management," Academy of Management Review, 6.4, 1981.
Buzzell, Robert D. and Mark J. Chussil.
Sloan Management Review, 26.4, 1985.
"Managing for tomorrow,"
Caves, Richard E. "Industrial organization, corporate strategy and
structure," Journal of Economic Literature, 18, 1980.
Caves, Richard E. American Industry:
Structure, Conduct,
Performance. Englewood Cliffs:
Prentice-Hall, Inc., 1972.
"Three models of strategy," Academy of
Chaffee, Ellen Earle.
Management Review, 10, 1985.
Cochran, Betty and G. Thompson.
"Why new products fail," The National
Industrial Conference Board Record 1, October, 1964.
Coleman, James S., Elihu Katz, and Herbert Menzel.
Innovation.
Indianapolis: Bobs-Merrill, 1966.
Medical
Comanor, William. "Market structure, product differentiation, and
industrial research," Quarterly Journal of Economics, November, 1967.
Cook, Victor J., Jr.
"Marketing strategy and differential advantage,"
Journal of Marketing, 47, 1983.
Cooper, Arnold and Dan Schendel.
"Strategic responses to
technological threats," Business Horizons, February, 1976.
Cooper, Robert G. "The myth of the better mousetrap: What makes a
new product a success?" Business Quarterly, Spring, 1981.
Cooper, Robert G. "The new product process: An empirically-based
classification scheme," R&D Management, 13.1, 1983.
Cooper, Robert G. "The strategy-performance link in product
innovation," R&D Management, 14.4, 1984.
-184-
Corey, E. Raymond. Industrial Marketing.
Prentice-Hall, Inc., 1976.
Englewood Cliffs:
Cronbach, L.J. "Coefficient alpha and internal structure of tests,"
Psychometrika, 16, 1951.
Strategic Market Planning: The Pursuit of Competitive
Day, George S.
Advantage. St. Paul: West Publishing Co., 1984.
Department of Health and Social Security - Great Britain. "Evaluation
of real-time ultrasonic scanners," Health Equipment Information, 81,
1979.
Department of Health and Social Security - Great Britain. "Evaluation
of real-time ultrasonic scanners," Health Equipment Information, 89,
1980.
Doz, Yves and Richard Hamermesh. "Medical products group (A,B,C),"
Harvard Business School Case, 2-375-313, 1975.
Drew, Philip G. "The great ultrasound incentives program," Diagnostic
Imaging, 3.8, 1981.
Durbin, J. "Incomplete blocks in ranking experiments,"
Journal of Psychology - Statistical Section, 4, 1951.
Eberstadt & Co., Inc.
document, 1983.
Everitt, Brian.
British
"Ultrasound market analysis," prospectus
Cluster Analysis.
New York:
Halsted Press, 1980.
Fellner, William. The Economics of Technological Advance.
General Learning Press, 1971.
New York:
Fombrun, Charles J. and Edward J. Zajac.
"Structural and perceptual
influences on industry stratification," Academy of Management Journal,
forthcoming.
Foster, Richard N. Innovation:
Summit Books, 1986.
The Attacker's Advantage.
New York:
Foster, Richard N.
"A call for vision in managing technology,"
Business Week, May 24, 1982.
Frazier, Gary L. and Roy D. Howell.
"Business definition and
performance," Journal of Marketing, 47, 1983.
"Intraindustry marketing strategy
Frazier, Gary L. and Roy D. Howell.
effects on the analysis of firm performance," Journal of Business
Research, 10, 1982.
-185-
Freeman, Christopher. The Economics of Industrial Innovation.
Cambridge: The MIT Press, 1982.
Freiberger, Paul and Michael Swain.
Osborne/McGraw-Hill, 1984.
Fire in the Valley.
Berkeley:
Freiherr, Greg.
"Doubts arise over quality of ultrasonography
training," Diagnostic Imaging, 7.7, 1985.
Friar, John and Mel Horwitch.
"The current transformation of
technology strategy: The attempt to create multiple avenues for
innovation within the large corporation," Sloan School Working Paper
1618-84, December 20, 1984.
"The emergence of technology strategy:
Friar, John and Mel Horwitch.
A new dimension of strategic management," Technology in Society,
7.2+3, 1985.
Friar, John.
"A test of a methodology to rank quality of technology
in medical equipment," MIT Working Paper 1624-85, 1984.
Frohman, Alan L. "Technology as a competitive weapon," Harvard
Business Review, January-February, 1982.
Frost & Sullivan, ed. Medical Ultrasound Imaging Equipment Markets in
the U.S. New York: Frost & Sullivan, Inc., 1982.
Fusfeld, Alan.
"How to put technology into corporate planning,"
Technology Review, 80.6, 1978.
Gale, Bradley T. and Richard Klavans.
"Formulating a quality
improvement strategy," Journal of Business Strategy, Winter, 1985.
Gobeli, David H. and William Rudelius.
"Managing innovation: Lessons
from the cardiac-pacing industry," Sloan Management Review, 26, 1985.
Gupta, Anil K. and V. Govindarajan.
"Business unit strategy,
managerial characteristics, and business unit effectiveness at
strategy implementation," Academy of Management Journal, 27.1, 1984.
Hambrick, Donald C. "Some tests of the effectiveness of functional
attributes of Miles and Snow's strategy types," Academy of Management
Journal, 26.1, 1983.
Hamilton, Betty, ed.
F&S Press, 1982.
Medical Diagnostic Imaging Systems.
New York:
"Estimation in panel models:
Hannan, Michael T. and Alice A. Young.
Results on pooling cross-sections and time-series," Sociological
Methodology 1977, ed. David R. Heise. San Francisco: Josey-Bass,
Inc., 1977.
-186-
Harrigan, Kathryn Rudie. "An application of clustering for strategic
group analysis," Strategic Management Journal, 6, 1985.
Hayes, Robert H. and Steven C. Wheelwright. "Link manufacturing
process and product life cycles," Harvard Business Review, 57.1, 1979.
Hayes, Robert and William Abernathy. "Managing our way to economic
decline," Harvard Business Review, July-August, 1980.
Heany, Donald F. "Degrees of product innovation," The Journal of
Business Strategy, 3.4, 1983.
Hitt, Michael A., R. Duane Ireland and K.A. Palia. "Industrial firms'
grand strategy and functional importance: Moderating effects of
technology and uncertainty," Academy of Management Journal, 25.2, 1982.
Analytical
Horwitch, Mel and C.K. Prahalad. "Managing technological innovation
three ideal modes," Sloan Management Review, Winter, 1976.
-
Strategy Formulation:
Hofer, Charles W. and Dan Schendel.
Concepts.
St. Paul: West Publishing Co., 1978.
Horwitch, Mel. "Changing patterns for corporate strategy and
technology management," paper presented at Mitsubishi Bank Foundation
Conference, March, 1983.
Horwitch, Mel.
"The blending of two paradigms for private-sector
technology strategy," Competitiveness Through Technology: What
Business Needs from Government, ed. Jerry Dermer.
Lexington:
Lexington Books, Inc., 1986.
J. Lloyd Johnson Associates. Ultrasound Systems Market Analysis.
Northfield: J. Lloyd Johnson, 1981.
J. Lloyd Johnson Associates.
J. Lloyd Johnson, 1982.
The Future of Ultrasound.
Northfield:
Jemison, David B. "The importance of an integrative approach to
strategic management research," Academy of Management Review, 6, 1981.
Kantrow, Alan M. "The strategy-technology connection," Harvard
Business Review, 58.4, 1980.
Katz, R. and M. Tushman. "An investigation into the managerial roles
and career paths of gatekeepers and project supervisors in a major R&D
facility, " R&D Management, 11.3, 1981.
Kendall, Maurice. Rank Correlation Methods.
Publishing Company, 1962.
New York:
Hafner
Kerin, Roger A. and Robert A. Peterson. Strategic Marketing
Problems.
Boston: Allyn and Bacon, Inc., 1984.
-187-
Klein Biomedical Consultants.
The Medical Diagnostic Ultrasound
Market. New York: Klein Biomedical, 1980.
Krugman, Paul.
"Defining high-tech," unpublished note, 1982.
Krugman, Paul. "New theories of trade among industrial countries,"
AEA Papers and Proceedings, 73.2, 1983.
Kruskall, William H. "A nonparametric test for the several sample
problem," The Annals of Mathematical Statistics, 23, 1952.
Kruskall, William H. and W. Allen Wallis.
"Use of ranks in
one-criterion variance analysis," Journal of the American Statistical
Association, 47.260, December, 1952.
Lawrence, Paul R. and Davis Dyer.
York: The Free Press, 1983.
Renewing American Industry.
Lawrence, Paul R. and Jay W. Lorsch.
Homewood, IL: Irwin, 1969.
New
Organization and Environment.
Lewis, Jordan. "Technology, entreprise, and American economic
growth," Science, 215, 1982.
Liberatore, Mathew J. and George J. Titus.
"Synthesizing R&D planning
and business strategy: Some preliminary findings," R&D Management,
13.4, 1983.
Little, Arthur D., Inc.
"An analysis of the National Science
Foundation's medical instrumentation experiment," report from the
National Technical Information Service, June, 1979.
Loether, Herman and Donald McTavish. Descriptive and Inferential
Statistics. Boston: Allyn and Bacon, Inc., 1974.
Madsen, Ernest, et al. "Tissue mimicking materials for ultrasound
phantoms," Medical Physics, 5.5, 1978.
Maidique, Modesto A. and Peter Patch. "Corporate strategy and
technological policy," Harvard Business School Case, 1978.
Mansfield Edwin.
Norton, 1968.
The Economics of Technological Change.
New York:
Mansfield, Edwin and Samuel Wagner. "Organizational and strategic
factors associated with probabilities of success in industrial R&D,"
The Journal of Business, 48, 1975.
Mansfield, Edwin, et al. Technology Transfer, Productivity, and
Economic Policy. New York: W.W. Norton & Company, 1982.
Mansfield, Edwin.
59.6, 1981.
"How econmists see R&D," Harvard Business Review,
-188-
An
"Technological change and market structure:
Mansfield, Edwin.
empirical study," AEA Papers and Proceedings, 73.2, 1983.
"The anatomy of successful innovations," Innovation,
Marquis, Don.
November 1969.
Maslak, Samuel H. "Computed sonography," Ultrasound Annual 1985, eds.
R.C. Sanders and M.C. Hill. New York: Raven Press, 1985.
"A moving
McDicken, W.N., D.C.M. Morrison and D.S.A. Smith.
tissue-equivalent phantom for medical ultrasonics," University of
Edinburgh working paper, 1983.
"Effective innovation
McGinnis, Michael A. and M. Robert Ackelsberg.
management: Missing link in strategic planning?" The Journal of
Business Strategy, 4.1, 1983.
McKay, Niccie Lee. "The economics of the medical diagnostic imaging
equipment industry," MIT Economics Ph.D. Thesis, 1983.
Metcalfe, J.S. "Impulse and diffusion in the study of technical
change," Futures, 13.5, 1981.
Meyer, Marc H. "New product strategy in the technology-based firm:
product technology, market strategy, and performance," MIT Ph.D.
Thesis, 1986.
Miles, R.E. and C.C. Snow. Organizational Strategy, Structure and
Process. New York: McGraw-Hill, 1978.
Mintzberg, Henry.
Science, 1978.
"Patterns in strategy formation," Management
"Managing innovation over
Moore, William L. and Michael L. Tushman.
the product life cycle," Readings in the Management of Innovation,
eds. Michael L. Tushman and William L. Moore. Marshfield: Pitman
Publishing Inc., 1982.
"The influence of market demand
Mowery, David and Nathan Rosenberg.
upon innovation: A critical review of some recent empirical studies,"
Research Policy, 8.2, 1979.
Myers, Sumner and Eldon E. Sweezy.
Technology Review, 80.5, 1978.
"Why innovations fail,"
"The Schumpeterian trade-off
Nelson, R.R. and S.G. Winter.
revisited," American Economic Review, 72, 1982.
Nelson, Richard R. and Sidney G. Winter. "In search of a useful
theory of innovation," Research Policy, 6, 1977.
-189-
Nohria, Nitin. "International strategic alliances: Some theoretical
and empirical propositions," unpublished Sloan School paper.
O'Shaughnessy, John. Competitive Marketing:
Boston: George Allen & Unwin, 1984.
Peters, Thomas J. and Robert H. Waterman, Jr.
Excellence. New York: Harper & Row, 1982.
Peters, Tom and Nancy Austin.
Random House, 1985.
A Strategic Approach.
In Search of
A Passion for Excellence.
Petrov, Boris. "The advent of the technology portfolio,"
of Business Strategy, 3.2, 1982.
New York:
The Journal
Philips, Lynn W., Dae R. Chang and Robert D. Buzzell. "Product
quality, cost position and business performance: A test of some key
hypotheses," Journal of Marketing, 47, Spring, 1983.
Porter, Michael. "The contributions of industrial organization to
strategic management," Academy of Management Review, 6, 1981.
Porter, Michael. "The structure within industries and companies'
performance," The Review of Economics and Statistics, LXI.2, 1979.
Porter, Michael. "The technological dimension of competitive
strategy," Research on Technological Innovation, Management and
Policy, ed. Richard Rosenbloom. Greenwich: JAI Press, 1983.
Porter, Michael.
Competitive Strategy.
New York:
The Free Press,
1980.
Porter, Michael.
Competitive Advantage.
New York:
The Free Press,
1985.
Powis, Raymond L. and Wendy J. Powis.
"Mastering the subtleties of
real time ultrasound," A Thinker's Guide to Ultrasound Imaging.
Baltimore: Urban & Schwarzenberg, 1984.
Prahalad, C.K. "Developing strategic capability: An agenda for top
management," Human Resource Management, 22.3, 1983.
Pratt, William K.
Digital Image Processing.
New York:
John Wiley
&
Prahalad, C.K. "Technology and corporate strategy," Harvard Business
School working paper, April, 1974.
Sons, 1978.
Quinn, James Brian.
"Strategic change:
Sloan Management Review, 20.1, 1978.
'Logical incrementalism,"'
Quinn, James Brian. "Technological innovation, entrepreneurship, and
strategy," Sloan Management Review, 20.3, 1979.
-190-
Rao, Vithalo R. "Pricing research in marketing:
art," Cornell University working paper, 1982.
The state of the
Roberts, Edward B. "Is licensing an effective alternative for
achieving commercial benefits of R&D results?," M.I.T. working paper
WP1182-81, January, 1981.
Roberts, Edward B. and Charles A. Berry. "Entering new businesses:
Selecting strategies for success," Sloan Management Review, 26.3, 1985.
Roberts, Edward, et al., eds.
MIT Press, 1981.
Biomedical Innovation.
Cambridge:
The
Rogers, Everett and F. Floyd Shoemaker.
Communication of
Innovations. New York: The Free Press, 1971.
Rosenberg, Nathan. Technology and American Economic Growth.
York: Harper and Row, 1972.
New
Rosenbloom, R.S. and William Abernathy. "The climate for innovation
in industry," Research Policy, 11, 1982.
Rosenbloom, Richard S. "Technological innovation in firms and
industries: An assessment of the state of the art," Technological
Innovation, eds. P. Kelly and M. Kranzberg.
San Francisco: San
Francisco Press, 1978.
Rothschild, William E.
Strategic Alternatives.
New York:
Amacom,
1979.
Rothwell, R., et al.
"SAPPHO updated - project SAPPHO Phase II,"
Research Policy, 3, 1974.
Rumelt, Richard P.
"Evaluation of strategy: Theory and models,"
Strategic Management, eds. Schendel and Hofer. Boston: Little, Brown
and Company, 1979.
Ryan, T.A., Jr. and B.L. Joiner. "Normal probability plots and tests
for normality," Technical Report, Pennsylvania State University, 1982.
Schendel, D.E. and C.W. Hofer. "Theory building and theory testing in
strategic management," Strategic Management: A New View of Business
Policy and Planning. Boston: Little, Brown and Company, 1979a.
Schendel, Dan E. and Charles W. Hofer. Strategic Management.
Boston: Little, Brown and Company, 1979.
Scherer, F.M. Industrial Market Structure and Economic Performance.
Chicago: Rand McNally, 1970.
Scherer, F.M.
Innovation and Growth.
Cambridge:
The MIT Press, 1984.
-191-
Schumpeter, Joseph. The Theory of Economic Development.
Harvard University Press, 1934.
Cambridge:
"The role of the interaction between the user and the
Shaw, Brian.
manufacturer in medical equipment innovation," R&D Management, 15.4,
1985.
Siegel, Sidney.
Nonparametric Statistics.
New York:
McGraw-Hill,
1956.
"Strategy, distinctive
Snow, Charles C. and Lawrence G. Hrebiniak.
competence, and organizational performance," Administrative Science
Quarterly, 25, 1980.
Stobaugh, Robert and Piero Telesio, "Match manufacturing policies and
product strategy," Harvard Business Review, March-April, 1983.
Tabachnick, Barbara G. and Linda S. Fidell.
Statistics. New York: Harper & Row, 1983.
Using Multivariate
Tassey, Gregory. "Competitive strategies and performance in
technology-based industries," Journal of Economics and Business, 35,
1983.
Taylor, James W. Planning Profitable New Product Strategies.
Chilton Book Company, 1984.
Radnor:
Technology Marketing Group.
Medicine Magazine, 1981.
"Ultrasound Forum '81," Radiology/Nuclear
Theta Technology Corporation. Ultrasound Equipment.
Theta Technology Corporation, 1981.
Wethersfield:
Thomas, Howard and N. Venkatraman.
"Research on strategic groups:
Progress and prognosis," unpublished manuscript, 1985.
Urban, Glen and John Hauser.
Design and Marketing of New Products.
Englewood Cliffs:
Prentice-Hall, Inc., 1980.
Urban, Glenn, et al. "Market share rewards to pioneering brands: An
empirical analysis and strategic implications," Sloan Working Paper,
1984.
Utterback, James M. "Innovation in industry and the diffusion of
technology," Science, 183, 1974.
"A dynamic model of
Utterback, James M. and William J. Abernathy.
process and product innovation," Omega, 3.6, 1975.
"The dominant role of users in the scientific
von Hippel, Eric.
instrument innovation process," Research Policy, 5, 1976.
-192-
"Technology push and need pull:
Voss, C.A.
Management, 14.3, 1984.
A new perspective," R&D
"Reflecting corporate strategy in
Wheelwright, Steven C.
manufacturing decisions," Business Horizons, February, 1978.
White, Roderick E. and Richard G. Hamermesh. "Toward a model of
business unit performance: An integrative approach," Academy of
Management Review, 6.2, 1981.
Wilkinson, Alan. "Corporate strategy and the buying and selling of
know-how," R&D Management, 15.4, 1985.
Willerman, Ben. "The adaptation and use of Kendall's coefficient of
concordance (w) to sociometric-type rankings," Psychological
Bulletin, LII.2, 1955.
Wind, Yoram and Thomas S. Robertson.
"Marketing strategy: New
directions for theory and research," Journal of Marketing, 47, 1983.
Zeithaml, Carl P. and Louis W. Fry. "Contextual and strategic
differences among mature businesses in four dynamic performance
situations," Academy of Management Journal, 27.4, 1984.
-193-
APPENDICES
-194-
APPENDIX
1
COMPANY PROFILES
-195-
Acuson
Acuson is a privately held, dedicated ultrasound company.
Acuson first began delivering products at the end of 1983 when it
introduced electronically focused linear array and sector scanner
systems.
Acuson developed computed sonography, which is the
formation of an image under software control.
Acuson is considered to be one of the fastest growing
ultrasound companies and is the market leader in sales of linear
arrays to hospitals.
Acuson is about a $30 million company.
-196-
Diasonics
Diasonics was founded in 1977 and shipped its first product in
1979.
The founder of Diasonics had headed Searle Ultrasound and left
to start his own company.
Diasonics started as a dedicated
ultrasound imaging company but soon expanded into other imaging
modalities:
digital radiography, NMR, and mobile x-ray.
Diasonics quickly became one of the leading companies in the
radiology segment of ultrasound.
The company developed mechanical
sector scanners with a novel transducer and with a computer base.
To expand into other ultrasound technologies, Diasonics
acquired Varian Ultrasound in 1981.
Varian had entered the
ultrasound market in 1975 with its introduction of a cardiac phased
array.
Diasonics also sold Hitachi linear arrays under its own
label.
Diasonics had acquired Fischer Imaging in 1983, but the
acquisition was rescinded in 1984.
To expand its marketing presence, Diasonics has acquired
several companies including Sonotron and Sonics.
Prior to the
Sonotron purchase, Siemens distributed Diasonics equipment in Europe.
Diasonics went public in 1983 and then followed the offering
with five consecutive quarters of losses.
In 1984 it was near
bankruptcy but has since restructured.
Diasonics was a $136 million company in 1984, about half of
which was ultrasound.
-197-
General Electric
General Electric is the leading seller of medical diagnostic
equipment in the U.S.
GE entered the ultrasound segment in 1979 by
buying Electro Physics Laboratory, which had developed a B-scan for
Litton and later Xonics.
GE upgraded the unit and introduced it in
1980.
Much of GE's ultrasound development has been through joint
Although GE's first linear array
ventures and joint developments.
was sourced from Aloka, it
developed a second linear array with
Yokogawa Medical Systems in a joint venture.
GE's cardiac phased
array, as well as its second generation phased array, was developed
with Analogic in a joint development.
Second Foundation designed a
mechanical sector and GE bought the design.
manufactured the transducers for GE's B-scan.
K.B. Aerotech, moreover,
GE's most recent
product, an abdominal phased array, was also developed by Yokogawa in
the joint venture.
GE has competed in ultrasound, then, by using outside
technology to enter the product markets and by using internal
development to create some further advances.
GE is a leader in the
phased array abdominal segment, which it entered in 1983.
GE is a $28 billion company.
-198-
GEC (U.K.)
GEC is a British manufacturer of electrical and electronic
components.
GEC's medical division was primarily concerned with
radiographic equipment until it purchased Picker International, a
full-line medical imaging company, in 1981.
Picker has changed hands several times over its recent
history.
Picker was founded in 1915 as a medical instrument
company.
Picker was first acquired in 1930 by C.I.T., which in turn
was acquired by RCA in 1980.
RCA sold Picker as soon as it could to
GEC, in 1981.
Prior to the recent. changes in ownership, Picker was the
leading ultrasound company in the U.S.
Picker was one of the first
companies in the ultrasound field, entering in 1964 and later buying
Physionics in 1968.
Physionics had introduced the first
two-dimensional scanner in the early 1960s.
Picker developed internally its product line until 1980.
Picker has used Hitachi linear arrays and Ausonics sector scanners
since then.
GEC also acquired Cambridge Instruments, who did
manufacture ultrasound equipment, in 1981 but divested them a couple
years later.
Cambridge Instruments no longer makes ultrasound
equipment.
GEC is a $7.0 billion company.
-199-
Hewlett-Packard
Hewlett-Packard is the largest producer of electronic
instruments in the world.
Hewlett-Packard's Medical Products Group
is a worldwide distributor and manufacturer of electronic medical
instrumentation.
H-P is not known for its diagnostic equipment as
much as its patient monitoring products.
H-P entered the diagnostic ultrasound field in 1980 with its
H-P quickly became
introduction of a cardiac phased array system.
the leader in that segment.
In order to acquire mechanical sector technology, H-P purchased
the Ekoline product line from Xonics in 1984.
Ekoline is an
amalgamation of several previously separate lines.
Litton Ultrasound in 1977 and then SKI in 1981.
Xonics purchased
Xonics went bankrupt
in 1984 and sold the ultrasound business to H-P.
SKI was one of the pioneers in the ultrasound field, starting
in 1962.
It formed a joint-venture with General Precision group and
developed the first commercial echoencephalograph, which it
introduced in 1963.
By working closely with researchers at the
University of Indiana, SKI was able to develop products for and then
virtually create the cardiac ultrasound segment.
In 1980,
SKI
purchased Mediscan, a small abdominal annular array company who sold
through Xerox.
H-P's product line now consists of its original cardiac phased
array and the mechanical sector from Ekoline.
H-P is a S6.5 bi~lion company.
-200-
Johnson & Johnson
J&J, a major supplier of health-care products, entered the
diagnostic ultrasound market when it acquired Technicare in 1979.
Technicare, a full-line medical imaging company, had entered the
ultrasound field by acquiring Unirad in 1976.
Unirad had started in
1971 and was one of the leading B-scan companies.
Technicare expanded its technology base by acquiring both Irex
and Echo Labs in 1982.
Irex had been a startup in 1970 and moved
into ultrasound with M-mode equipment in 1975.
Irex later introduced
phased array equipment in 1979 and was performing joint research with
Hoffrel on cardiac equipment.
Kontron.
Irex had been partially owned by
Echo Labs was a major producer of transducers, who supplied
(and still does) products to many of Technicare's competitors.
In
1984, J&J combined the three companies under the name J&J Ultrasound.
More recently, J&J Ultrasound has been sourcing its products
from Aloka.
J&J is a $6.4 billion company.
sold to General Electric.
In mid 1986, J&J Ultrasound was
-201-
Philips Ultrasound
N. V. Philips is one of the world's leading electronics firms
and is based in the Netherlands.
The shareholders of N. V. Philips
own in a trust about 60 percent of the stock of North American
Philips.
Although N. A. Philips technically is independent, it has a
very close working relationship with N. V. Philips.
We will use the
Philips name interchangeably between the two.
Philips entered the U.S. ultrasound market by acquiring Rohe
Scientific in 1976.
Rohe was a dedicated ultrasound company.
In
1967, Rohe distributed Kretz products, but in the early 1970s it
developed its own B-scan.
Rohe introduced the first stored video
gray scale in 1973.
Philips sourced its mechanical sector transducer from Hoffrel
when it developed its own scanner.
Philips also sources its phased
array unit from Matsushita, its doppler from Medasonics, and its
linear array from Hitachi.
Philips Medical Systems is a full-line imaging company and is
considered to be the second largest medical equipment company in the
U.S.
N. V. Philips is an $18.1 billion company.
-202-
Siemens
Siemens is one of the world's leading electrical and electronic
products companies.
Siemens medical imaging group is considered to
be the second largest in the world, and Siemens patient monitoring
group is considered the largest in the U.S.
Siemens is headquartered
in Germany.
Siemens was an early player in the ultrasound industry when it
developed piezoelectric crystals for ultrasound use in the 1950s.
Siemens also introduced one of the earliest real time scanners, but
it never really established itself in the U.S. until it acquired
Searle Ultrasound in 1980.
Searle Ultrasound had entered the market in 1976 with the first
microprocessor-controlled B-scan unit.
In 1977 Searle introduced the
first digita] scan converter and the next year a linear array. Searle
was the sixth largest ultrasound company in the U.S. when it was
acquired.
Since the Searle acquisition, Siemens has sold Diasonics
equipment in Europe, and sourced the rest of its product line from
Aloka, Matsushita, and Kretz.
Siemens is a $17.8 billion company.
-203-
Squibb
Squibb is a major U.S. corporation in health-care and
personal-care products.
its pharmaceutical lines.
Sixty percent of Squibb's sales come from
Squibb entered the ultrasound market by
acquiring ATL in 1979 and ADR in 1982.
For European sales, Squibb
also acquired Kranzbuhler in 1982.
ATL was a startup in 1969 and entered the ultrasound market in
1976.
By 1980, ATL was the leading company in the U.S.
ATL
originally licensed its technology from the University of Washington
in 1974.
ATL also licensed technology from SRl in 1980, but performs
most of its own development work.
ADR was also a startup that incorporated in 1973.
introduced its first product in 1974 -products.
It
one of the first linear array
ADR has been the leader in linear array sales ever since.
Although the combined ATL/ADR ultrasound group has been the
U.S. market leader for the last several years, for the last two years
rumors have abounded that Squibb has been trying to sell it.
losing money in 1985.
Squibb is a $2 billion company.
ATL was
-204-
Toshiba
Toshiba is based in Japan and is one of the world's largest
manufacturers of electric and electronic products.
Systems is a full-line medical equipment company.
making medical ultrasound products in 1968.
Toshiba Medical
Toshiba started
Toshiba at first sold
through Litton in the U.S. and then entered the U.S. market directly
in 1977.
Toshiba has always developed its own products and was one of
the first to introduce linear arrays and phased arrays into the U.S.
More recently Toshiba has introduced curvilinear arrays and
trapezoidal arrays.
Toshiba has the broadest product line and was
considered at one time to be the world leader in ultrasound
equipment.
Toshiba Corporation is a $15.3 billion company.
-205-
APPENDIX 2
COMPANIES PARTICIPATING IN STUDY
[
(
Key:
] acquiring companies
) acquired companies or supplier
-206-
Method of Participation
Company
Acuson
Survey
ADR [Squibb]
Interview
Aloka
Supplier
American Edwards
Survey
[American Hospital Supply]
American Electromedics
Survey
ATL [Squibb]
Survey
Ausonics
Survey
Bio-Dynamics [Biosound]
Subsidiary
Bion
Defunct
Biosound (Bio-Dynamics. Honeywell)
Survey
Bruel & Kjaer
Interview
Cambridge [G.E.C.]
Survey
Carolina Medica]
Survey
CGR [Thomson]
Survey
Cone (Kretz)
Survey
Cooper
Interview
(Xenotec)
Dapco
Survey
Diagnostic Sonar
Survey
Diasonics (Varian)
Survey
Echomed [Multiscan]
Survey
E for M [Honeywell]
Subsidiary
Elscint
Interview
Fischer (EMI) [GL]
Interview
GE
(Electro Physics, J&J)
GL (Fischer)
Survey
Survey
-207-
Method of Participation
Company
Hewlett-Packard
(Eko]ine)
Survey
High Stoy (Grumman, Interspec)
Defunct
Hitachi
Supplier
Hokanson
Survey
Honeywell (E for M) [Biosound]
Interview
Hoffrel
Survey
Imex
Survey
International Imaging
Survey
(Illinois Imaging)
Interspec
Survey
1pco
Defunct
Irex [J&JI
Interview
Ithaco
Survey
J&J (Technicare, Irex, Echo)
Interview
KB Aerotech [SmithKline]
Survey
Kontron
(Roche, Grumman)
Interview
Kretz
Supplier
Labsonics
Survey
Life
Defunct
Matsushita
Supplier
Medasonics [Kendall Hospital]
Survey
Narco
Defunct
National
Defunct
Organon Teknika
Defunct
Parks Medical
Survey
Pfizer (Aloka)
Survey
-208-
Method of Participation
Company
Philips (Rohe)
Survey
Picker
Survey
[G.E.C.]
Pie Medical
Survey
Roche [Kontroni
Subsidiary
Rohe [Philips]
Subsidiary
Searle [Siemens]
Subsidiary
Second Foundation
Defunct
Shimadzu
Interview
Siemens (Searle)
Survey
SKI (Mediscan) [Xonics]
Interview
Sonicaid
Interview
Spectrascan
Interview
Storz
Survey
Technicare (Unirad, Ohio Nuclear)
Subsidiary
[J&JJ
Toshiba
Interview
Thomson (CGR)
Survey
Ultrasonix
Survey
Unirad [Ohio Nuclear, Technicare]
Subsidiary
Unigon [Ultrasonix]
Subsidiary
Varian [Diasonicsj
Interview
Visidyne
Survey
Xenotec [Cooper]
Subsidiary
Xerox
Survey
Xonics (SKI, Litton)
[H-P]
Interview
-209-
APPENDIX 3
INDEPTH SURVEY
Questionnaire on the U.S. Ultrasound Industry
This study will cover the five years from 1979 through 1983 by
quarters.
Please indicate in which quarters any changes occurred. I am
only interested in the U.S. geographic market for medical diagnostic
ultrasound. Complete confidentiality will be maintained.
If another format is easier for you, please use it. I have
enclosed a return envelope for your convenience. If you have any
questions, please contact me. Thank you for your help.
John Friar
Sloan School of Management
E52-534
Massachusetts Institute of Technology
50 Memorial Drive
Cambridge, MA
02139
617-253-6651
Company Name
Your Name
Phone
#
Position
I.
Products
A.
Diagnostic ultrasound products on the market as of 1/1/79 and
original dates of introduction, e.g. Model X, sector,
cardiology, June, '75.
Product Name, Type, Application
B.
Products introduced between 1979 and 1983 and dates of
introduction.
Product Name, Type, Application
C.
Date of Introduction
Dates of any major improvements to existing products.
Product & Improvement
D.
Date of Introduction
Date Introduced
Dates any products may have been dropped from product lines.
Product
Last Month in Which Sold
II.
Pricing
A.
Average realized price for each product.
price changes.
Product
Year
1979
1980
1981
1982
1983
Qi
Q2
Q3
Q4
Qi
Q2
Q3
Q4
Qi
Q2
Q3
Q4
Qi
Q2
Q3
Q4
Qi
Q2
Q3
Q4
B.
1979
1980
1981
1982
1983
Service contract terms for each year.
Indicate time of
III.
Advertising and Promotion
A.
Dollars spent on advertising and promotion in each quarter.
Qi
Q2
Q3
Q4
1979
1980
1981
1982
1983
IV.
Personnel
A.
Numbers of sales people (or equivalents if distributors were used)
by quarter.
Qi
1979
1980
1981
1982
1983
Q2
Q3
Q4
B.
Numbers of service technicians (or equivalents if outside agencies
used) by quarter.
Ql
Q2
1979
1980
1981
1982
1983
V.
Sales
A.
Unit sales by product by quarter.
Product
Year
1979
1980
1981
1982
Qi
Q2
Q3
Q4
Qi
Q2
Q3
Q4
Qi
Q2
Q3
Q4
Qi
Q2
Q3
1983
Q4
Q1
Q2
Q3
Q4
Q3
Q4
VI.
VII.
Product Development
A.
Which products were developed internally and which ones sourced
from outside?
B.
Do you license your products to other companies?
Which ones?
Strategy
A.
What do you feel is the key to success in this industry?
B.
What have you emphasized to create a competitive advantage?
C.
As a percentage of sales, what has been your level of R&D
expenditures?
1979
1980
1981
1982
1983
VIII.
D.
Have you either acquired or been acquired by other ultrasound firms?
E.
Are you part of a larger medical equipment company or
freestanding? Of total medical equipment sales, what proportion is
in diagnostic ultrasound?
F.
Are patents important to your company? If so, approximately how
many patents in ultrsound equipment does your company hold?
G.
What proportion of your overall sales is in the U.S.?
Personal thoughts
A.
To understand the ultrasound industry, what does one need to
consider?
-217-
APPENDIX 4
GENERAL SURVEY
August 9, 1985
Dear
:
I am a Ph.D. student at the Sloan School of Management at the
Massachusetts Institute of Technology and am studying the management of
innovation and competitive strategy. An important question to be addressed is
the actual relationship between a technological advantage and success in the
market place. Having for several years both worked in and studied the medical
diagnostic ultrasound industry, I know that technological competition is
intense and market shares have been volatile. I am undertaking a large
industry study to try to explain the relationship between technological
competition and changing market shares. I would hope that the results of such
a study would be of interest to you so that you and your firm will participate
by providing some information about yourselves.
I have developed ratings of ultrasound equipment as to its functional
capability or ability to produce diagnostic information (study enclosed). I
would like to use these ratings with the other information I am requesting for
the forthcoming larger study. Naturally the quality of this analysis depends
on participation by all the companies in the industry.
I also realize that the information I am asking for is sensitive and
proprietary. Other than myself, no one will have access to this information.
The information will be encoded into a large data base and only the results of
the multivariate analyses will be available. This study is strictly for
academic purposes only. The presentation format will be similar to that of
the enclosed study, from which it is easy to see that all proprietary
information is protected.
The information I need is listed on the following form, but please use
any format that is easier for you. I will also contact you in a couple of
weeks to see if I can arrange a follow-up interview with you or someone in
your organization. If you have any questions, please contact me at the above
address or phone number. A copy of any papers stemming from this research
will be sent to you. I thank you in advance for your help and cooperation.
Sincerely,
John Friar
SURVEY OF U.S. MEDICAL ULTRASOUND COMPANIES
1.
Do you presently manufacture medical ultrasound equipment in the United
No
Yes
States?
2.
Are you the sole distributor of another company's or your foreign parent's
medical ultrasound equipment in the United States?
Yes
No
3.
If you answered no to both and you were in the market at one time, when did
you leave the market?
If you answered no to both questions 1 and 2, please go to the final question.
If you answered yes to either of the two questions, please continue with the
next question.
4.
In what year did you enter the U.S. market?
5.
What type of equipment and to which medical specialities do you sell?
(e.g., phased array-cardiology; test phantom-radiology)
6.
What is the approximate size of your firm?
Less than $5,000,000 in sales
Less than $10,000,000 in sales
More than $10,000,000 in sales
7.
Are you either a licensee or a licensor of ultrasound equipment?
Yes
No
Are you a free-standing company or a subsidiary of a larger company?
Sub
Free
9.
Do you sell in markets outside of the U.S.?
10.
If you had to allocate a budget among the following areas to improve your
company's sales, how would you do it?
Marketing
%
100
%
R&D
Price Promotions
% %
8.
Yes
No
11.
Why did you enter the U.S. ultrasound market?
12.
Company name (optional)
Your position
Thank you very much for your help and please return the questionnaire in the
envelope that was provided.
-221-
APPENDIX 5
SOURCES USED IN DETERMINING SEGMENT AND INDIVIDUAL COMPANY SALES
-222-
1979-1983:
Survey and Interview Data
Beyond the survey data and as a check to the survey data, published
estimates by:
Wall Street Analysts
Sanford C. Bernstein
F. Eberstadt & Co.
Hambrecht & Quist
Market Research Consultants
Frost & Sullivan
J. Lloyd Johnson
Klein Biomedical Consultants
Medical Products Marketing Services
Other
Diagnostic Imaging, Reader Surveys
Hamilton (1982)
-223-
APPENDIX 6
ADDITIONAL CLUSTERING INFORMATION
-224-
a)
b)
Attributes' means and standard deviations
Standard Deviation
Attribute
Mean
Marketing
41.52
25.216
R&D
48.92
22.217
Price
10.48
12.1626
Calinski & Harabasz test of the number of clusters
C
trace (B)jtrace (W)
k-1
N-k
where:
k = number of clusters
N = number of observations
If C rises to a maximum, clusters are present.
If C increases or decreases monotonically, no cluster structure
or a hierarchical structure are suggested.
-225-
APPENDIX 7
COMPUTATION AND RELIABILITY
OF INDEX SCALES
-226-
Computation of R&D Index
n- rank + 1*
Rating of image quality:
n - rank + 1
Number of updates:
raw number (0 to 4)
Type of development:
*
Entry into segment:
5 point scale described below:
5. Internal development
4. Technology acquisition with major
internal
3. Evenly split between external and internal
2. Mostly external
1. Completely external
*
n equals at least 5 companies
Statistics on R&D index:
mean:
9.484
standard deviation: 4.458
-227-
Computation of Marketing Index
n -rank + 1*
Sales force:
Rank determined using numbers from the
following formula if the number of salesmen was
not broken out:
(sm X ss +ts X te.)
n
i=1
where:
sm = number of salesmen
ss = sales to the segment (units)
ts = total sales (units)
te = technique weig hts
Technique weights:
5. Dedicated to the segment
4. Dedicated to ultrasound
3. Shared with other product lines
2. Manufacturers' reps
1. Mail order
*
n equals at least 5 companies
-228-
Service force:
n-rank + 1*
Rank determined using numbers from the
following formula if the number of service
personnel was not broken out:
(sm Xss +ts Xteg)
-~
n=1
where:
sm = number of service personnel
ss = sales to the segment (units)
ts = total sales (units)
te = technique weights
Technique weights:
5. Dedicated ultrasound
4. Shared with other product lines
3. Inhouse third party
2. Outside service companies
1. Mail
*n equals at least 5 companies
-229-
Advertising and promotion:
5 point scale determined by the first two
digits from the following formula:
(adv X ss.+ts.)
i=1
where:
adv = amount spent on advertising and promotion
ss = sales to the segment (units)
ts = total sales (units)
Product assortment:
raw number (1 to 5)
Statistics on the marketing index:
mean:
8.290
standard deviation: 3.653
-230-
Reliability of R&D Index
Correlation of R&D index measures
Inter - Item Correlation
(1)
(2)
(
Measure
3)
Entry
(1)
1.0000
Rating
(2)
.3440
1.0000
Numupdat
(3)
.6255
.5763
1.0C 00
Devel
(4)
.3557
.2645
.5(23
(4)
1.0000
N=31
Cronbach Alpha
Measure
Alpha if deleted
Entry
.6909
Rating
.7323
Numupdat
.5760
Devel
.7376
Alpha =.7453
Standardized Item Alpha =.7621
-231-
Reliability of Marketing Index
Correlation of marketing index measures
Measure
(1)
(2)
(
Inter - Item Correlation
3)
Sales
(1)
1.0000
Service
(2)
.8319
1.0000
Adver
(3)
.7119
.7495
1.0 000
Assort
(4)
.3778
.1249
.3 127
(4)
1.0000
N=31
Cronbach Alpha
Measure
Alpha if deleted
Sales
.6732
Service
.7305
Adver
.7136
Assort
.8883
Alpha = .8159
Rerun Alpha =.8883
Standardized item Alpha =.9068
-232-
Correlation Between Indices
Pearson product moment correlations
Index
(1)
(2)
R&D
(1)
1.000
Marketing
(2)
0.318
1.000
Price
(3)
-0.180
0.306
N=31
(
Inter -item Correlation
3)
1.0 00
-233-
APPENDIX 8
TESTS FOR NORMALITY OF VARIABLES AND
RANDOMNESS OF RESIDUALS
-234-
Test for Normality
Correlation of the data to the normal scores of the data (Ryan and Joiner)
Variable
Correlation
R&D
0.963*
Marketing
0.987**
Price
0.999**
Performance (log e units)
0.979**
Performance (ranks)
0.998**
*Significant at p = 0.05
**Significant at p = 0.01
Plots of Standardized Residuals Versus Predicted Scores
Equation 1
-235-
1 ..
a
*
-
,
Equation 2
--------
4-------------------------
-236-
APPENDIX 9
SQUARED SEMIPARTIAL CORRELATIONS
-237-
)
Squared Semipartial Correlations (sri 2
(Tabachnick and Fidell)
sr2
i
1(1-
dfre
r2)
Variable
R&D
Marketing
loge units
.15
.21
sri 2
rank
.21
.20