Problem Analysis
In 1993, Dell Computers experienced “Annus Horribilus,” which included a series of events that almost thrust them
into bankruptcy. Dell had previously followed a “fly by the seat of their pants” style of operations, which was
adequate during the beginnings of the business; however, by 1993, Dell found themselves in dire need of more
experienced senior management to give the organization more structure and to guide the company into the next stage
of maturity. Aligned with the company philosophy of “spartanism” and the company strategy of commoditization,
Dell prided itself on minimal and modest R&D spending, which subsequently meant product development was an
informal process.
This informal product development process consisted of varied teams, with different agendas, different personalities
and ultimately, different ways of executing on projects. As Dell developed as a company, the variety increased,
which inevitably gave their products a reputation for being inconsistent and unpredictable, as well as slow and
expensive, unlike its former reputation of being fast and cheap.
Dell attempted to resolve issues of inconsistency by implementing a formalized product development process with
six defined stages. This unsettled the developers and engineers, and many of them outright rejected it. Externally,
Dell was facing the heat of competition from other players and the overall pace of the industry was moving faster
than ever before.
Dell was no longer efficient enough to keep up and was forced to cancel a line of portables as the technology had
already become obsolete, resulting in a $20MM loss and causing a downward spiral that left Dell suffering. This
was a result of an appalling underinvestment in R&D and a “shrunken desktop mentality,” or a myopic vision
blinded by success in other areas. Just when everyone thought they had hit rock bottom, an engineering nightmare
occurred and Dell was forced to recall 17,000 notebooks. Ironically, because of the Dell’s direct customer
relationship, Dell had the customer intelligence to track down and grovel to the victimized customers.
By the end of 1993, Dell was in a crunch. Competition was surpassing the once Spartan and almighty Dell. Does
Dell have a chance to bounce back into the game? They are going to need a serious restructure of their internal
operations… as well as something to give them the upper hand above the competition. Batteries, anyone?
State of the Industry: A Growth Opportunity
In the early 1990s, the market for laptop PCs was growing much faster than the desktops market (see fig 1), and
gross margins for laptops were 3-5% higher than gross margins for desktops. In contrast to the overall growing
laptop market, Dell’s share of sales in laptops was headed in the wrong direction. Original forecasts projected that
laptops would account for 20-25% of Dell’s sales in 1993; in reality, laptops fell from 12% of sales in 1992 to only
2% of Dell’s sales in 1993. Dell was also dwarfed by rivals IBM, Compaq and Toshiba in terms of market share for
laptops (see figure 2). Therefore it was imperative that Dell repair its reputation from its laptop fiasco and re-enter
the laptop market with a quality product.
A New Approach to Product Development at Dell
In order not to repeat Dell’s previous mistakes, senior management rolled out a new 18-month process for product
development beginning in 1993, replacing the older, more informal process. The new process organized new
product development around cross-functional teams that were responsible for the project from the beginning to the
end. Project reviews were conducted every 3 months. There were six distinct phases:
Profile Phase – project scope, schedule and cost were defined.
Planning Phase – detailed business plan developed and scrutinized by senior executives
Implementation Phase – functional prototypes built and tested. Major financial commitment to Dell occurs at this
phase
Qualification Phase – production prototypes built and distributed to key customers. Sales force training begins
during this phase.
Launch Phase – extensive product testing; production ramped up; shipments made to customers
Acceptance Phase – post-launch customer feedback obtained.
Re-entering the Portable PC Market: A Critical Decision
Dell’s new product development process was put to test immediately as the decision was being made on how to
develop a new laptop PC. Market research identified battery life as the third most important feature to customers
when purchasing a laptop (see figure 3). Due to a chance meeting between Michael Dell and Sony executives, Dell
had the opportunity to have exclusive access to the new Lithium Ion (LiOn) battery technology which greatly
extended battery life, and offered superior overall performance to the standard Nickel Hydride (NiHi) technology.
The new technology would add value to Dell laptops as they sought to recapture market share. The technology was
not fully developed however, and there was a risk that it would not work. Dell thus faced a critical decision about
how to allocate resources for the development of the laptop. Four options were identified by the product
development team:
Option 1- commit to the old technology (NiHi)
Option 2 – commit to the new technology (LiOn)
Option 3a – over-design the computer so that it could accommodate either type of battery, thus deferring the battery
commitment until later
Option 3b – dual design (in parallel) of laptops that would use either NiHi or LiOn technology
Which option should Dell chose?
Option 1: Continue with a proven battery technology (NiHi)
According to estimates made by project manager and product marketer, Henry McCarty, Dell’s market share will be
2.5% if Dell stays with the status quo battery configuration of NiHi. This equates to 825,000 units sold over the
estimated 3 year product life. Given an average gross margin per unit over life of product of $600 and expected
$10M expected cost of development effort, the expected profit margin [E(PM)] is $485M under option 1. There is
100% confidence that the NiHi battery product will work (since it is the current used product) therefore, there is no
requirement to weigh the calculated E(PM) against others.
Pros: NiHi batteries are known, proven products that will not fail in concept. With such mitigated uncertainty, this
makes the financial evaluation metrics more reliable. Additionally, the lack of a new product development process
will remove any possible resulting delays. Traditional NiHi batteries also take less space than LiOn batteries. This
allows for packing in other features, such as communications control or memory management accessories. In the
end, Dell will merely produce a “good” product.
Cons: Option 1 severely limits Dell’s ability to differentiate. In an increasingly competitive market, a firm’s ability
to differentiate is critical to success. By utilizing a NiHi battery, Dell will be unable to advertise their battery as
adding value above that of any competitor. Most importantly, the E(PM) is not the highest out of all options.
Option 2: Go with the new battery technology (LiOn)
Under this scenario, McCarty predicts Dell’s market share to jump to 3.0% (~990,000 units over 3 years) if LiOn
technology works. However, if LiOn technology fails, Dell’s market share will fall to about 1.25% (~413,000 units
over 3 years). This drop in market share would be attributed to competitors having an established product on the
market, while Dell undergoes substantial rework (70% of original schedule, 30% of cost) switching back to NiHi.
There is only a 60% confidence that the LiOn battery product won’t fail so the E(PM) for option 2 is a weighted
average of E(PM, if LiOn works) and E(PM, if LiOn fails). Again, given an average gross margin per unit over life
of product of $600 and expected $10M expected cost of development effort, the expected profit margin [E(PM)] is
$444M under option 2.
Pros: Viewing battery life as a value-added feature is insightful given the fact that battery life is the #3 (27%
incidence) notebook evaluation criteria for actual business customer purchases (figure 3). If the LiOn technology
succeeds, Dell would obtain a distinct differentiating factor in an increasingly competitive market. It would give
Dell the ability to deliver a “superb” product versus a “good” product. Customers would now have a laptop that
“lasts all day” and “doesn’t run out [of battery life] on the plane.” In addition, Sony has a limited supply of the
LiOn technology. A successful Dell LiOn product would erect a strong barrier to entry. By using all of Sony’s new
batteries they would lock out any competitors from LiOn technology.
Cons: The profitability of option 2 is entirely predicated on the success of an unproven technology that is only 60%
likely to work! Wholly committing to LiOn technology could be a grave error. Even if the technology does succeed,
the development process would cause a delay to market. If the technology fails, significant rework time would be
spent reestablishing the NiHi technology, thereby compounding the delay to market even more. Failure would taint
Dell’s current brand image of quality and spillover into all its products: laptops, desktops, servers. This would
instantly cause irreparable damage to an image built over many years. Likewise, it would take many years to regain
that image. Also, LiOn batteries occupy more space than NiHi, thereby allowing less room for packing in other
features. Ultimately, option 2 has an E(PM) of $444M… the lowest of any option.
Option 3a: Dual Development – Defer commitment until qualification phase review
To calculate E(PM) under dual development, we estimate the $10M expected cost of development and “additional”
fixed cost at $2.5M because Dell would have to develop two technologies at the same time. These are the actual
project costs incurred. They include additional designers and engineers, material and tooling costs, etc. These costs
do not include the product opportunities Dell would forego if they had to pull people away from other projects.
Given the new fixed costs and calculating a weighted average of expected profit margins based on the success rate
of the technology, option 3a has an E(PM) of $542M… the highest of all options.
Pros:
This option combines the pros of new technology—LiOn, while avoiding the cons. By opting for dual
development, Dell can earn a “real option” and additional time to make the final decision until they receive more
LiOn information. Since Dell can develop these two systems at the same time, the product development team can
minimize the independence of NiHi and LiOn technologies and make sure there are not any delay problems when
LiOn fails. Superior financial results also support option 3a.
Cons: Dell needs to pay an additional fixed cost of $2.5 million to get this “delay option”. Senior management
should also take into consideration the potential demoralization and financial impact of discarding half the
craftsmanship in the final product. It is important to note, however, that with careful management skill, this is
something they could at least mitigate. The opportunity cost of pulling people from other projects also needs to be
considered.
Option 3b: Over-design – Defer commitment until qualification phase review
Similar as under dual development, to calculate E(PM) of pursuing an over-design strategy, we estimate the $10M
expected cost of development and “additional” variable cost of 0.5% of revenue (2.0% of margin) since Dell would
have to develop two technologies in the same time. Due to the LiOn battery’s different dimensions and properties,
Dell would have to over-design the computer case, charging circuitry, and battery management software to
accommodate either battery technology. This variable cost would add about $12 per unit. Given the new variable
costs and calculating a weighted average of expected profit margins based on the success rate of the technology,
option 3b has an E(PM) of $533M. This is the second highest of all options.
Pros: This option also combines the pros of new technology, while avoiding the cons. Dell can obtain a “real option”
by pursuing over-design. It will buy time such that they would not have to make a final decision until they receive
more LiOn information. By making piecewise commitments instead of binary choices, Dell can increase its
development flexibility. In the end, the quantitative results -- 2nd highest among options, proves option 3b is also
acceptable.
Cons: In pursuing this option, Dell would need to pay an estimated $11 million. However, over-design would
occupy more space, thereby pushing more hardware out of the overall design. Also, the bulkier design would be
less attractive to potential customers. In addition, the over-design style is not transferable to the next product
generation. If LiOn is successful, Dell eventually will need to develop a new product prototype for LiOn technology.
Our team also considered the Sensitivity Analysis of E(PM) if the confidence probability of LiOn technology
changes. Based on the assumption McCarty provided, it shows clearly that Option 3a is the best option as long the
Confidence of LiOn Technology is between 10%~90% (figure 4). It is also obvious that if Dell knows 100% that
LiOn will be successful Dell should choose Option 2. If the chance of success will be 0%, Dell should choose
Option 1.
The main reasons for this result are because
1) The estimated sales and margin is strong when the LiOn technology is workable and Dell adopts this
technology early to take the first mover advantage;
2) The expected cost, however, is also high if LiOn tech fails and Dell needs to switch back to NiHi. It is a big
disaster.
3) On the other hand, the option cost to delay this decision is relatively low. By pursuing the delay option, or
adopting a more flexible product development process, Dell can enjoy the benefits of late commitment
(deciding upon a technology after is validated) while gaining the advantages of early commitment (first
mover advantage). This is further proven by looking at the high means of options 3a and 3b complemented
by lowered risk in the form of a low standard deviation.
Recommendations:
A. Short Term (Latitude Series Product Launch)
From a quantitative analysis it is easy to see which development path should be chosen by Dell in choosing which
battery should be placed into its upcoming personal computer product launch. In looking at the sensitivity analysis
and expected profit margins it is clear that a more flexible approach should be instituted in developing the Latitude
series PC (figure 4). According to the data, the nickel hydride (NiHi) and lithium ion (LiOn) technology will each
yield an expected profit margin of $485MM and $444MM respectively, while choosing a more flexible approach,
whether it be dual development or over design, has the potential to yield $542MM and $533MM respectively. From
a pure numerical analysis it is clear that the development approach of choice will be the dual development option,
and it is this path that we recommend Dell pursue.
The reasons behind recommending this option stem from more than just a numerical analysis. While there are
obviously some drawbacks to each of the development processes under consideration, we believe the benefits of the
dual development process greatly exceed the risk associated with the other options by the greatest margin. For
example, if Sony’s final results indicate that the LiOn technology is stable and ready for mass production, Dell will
have an architectural platform ready to incorporate this new technology allowing for the company to achieve a first
mover advantage within the PC market. Additionally, the company will also be able to strategically outpace the
competition by locking in a contract for exclusive rights to the limited supply of LiOn technology, thereby excluding
their competition from competing along this parameter. With a growing demand for extended battery life due to
changes in PC market conditions, this technology could result in a huge economic payoff. Also, by leaving both
battery options available longer into the product development cycle, Dell can eliminate the risk and uncertainty
associated with trying to predict future market preferences and demands.
Typically, most company’s product development lifecycles become unnecessarily extended while market analysts
and product designers try and gather information as to what exactly the market conditions will look like at product
launch. Because commitments are made so early on in the product development lifecycle, and become costly to
change later in the process, most companies feel this step is necessary to reduce product launch risk. This risk can
be mitigated however by delaying important decisions, like which battery to install, until later in the development
process when designers can accurately know what the current market conditions are.
While this is a fundamental shift in how to approach the development process, both empirical and historical
evidence has already suggested that companies who have the ability to make changes close to product launch show a
positive correlation to increased customer satisfaction and firm performance. IBM’s product launch of their first
personal computer is a perfectly relevant industry example as to why product flexibility is critical towards
establishing market dominance. At the time of the launch of the first PC, IBM had to choose whether to utilize a
cassette based method of data storage, or move forward with floppy disk technology. Instead of making the choice
for the customers, IBM let the customers make the choice by designing their system to be compatible with both
options. As a result, IBM’s product launch was an enormous success and they were able to utilize real market data
to determine that the future technical direction of the company would be to pursue floppy disk technology. Since
Dell is in essentially the same position with its battery dilemma, they could borrow the lessons learned at IBM and
apply it to their current situation by choosing the more flexible dual development process.
B. Long Term (Rethinking the Product Development Process at Dell)
While it is natural for a company that has experienced a significant economic setback to react by rethinking its
processes and methodologies to gain more structure and control, it is critical that Dell does not overcompensate by
making its product development process too rigid. Therefore, in order to mitigate further future setbacks and costly
periods of indecision, Dell must avoid the temptation to try and control the process by making early economic
commitments in the product development process, and champion a process based on controlled flexibility.
In looking at a best-in-class benchmarking study of the product development process within the integrated circuit
development process it is clear to see why flexibility is so critical to obtaining a competitive advantage within the
high technology industry. Since customer requirements and technological advances occur so rapidly within this
segment of business, the company that is able to reduce the product development lifecycle to a minimum will be
able to stay ahead of the competition and better satisfy customer needs. In looking at a process with low flexibility
compared to a process with high flexibility within integrated circuit development you can see the greatly decreased
timeline from product conception to product launch associated with a highly flexible product development process
(figure 5). The majority of the gains enjoyed by the highly flexible process occur in the first three stages of
development: design specification, design development, and design verification.
While most companies are slowly gathering information to try and perfectly predict a product and process
configuration that will minimize product launch risk, the highly flexible process is able to bypass this information
gathering stage by designing a process that defers these costly decisions to later in the process when it is not so
costly to make changes. Therefore, they are able take the product development timeline from an average span of
eighteen months and reduce it to just eight months. Currently Dell’s product development timeline mirrors that of
the slower and less flexible product development timeline. Considering the challenges associated with the rapid
changes that face the laptop product segment, it is critical that Dell re-engineer its product development process to
become more flexible to face these challenges head on to realistically approach market dominance in this industry.
Fortunately the proposals we are recommending for Dell will not require a heavy capital expenditure to institute.
All that is required is buy in from top management to champion a product development culture focused on flexibility
by instituting four general management steps. First, project managers and product development teams must be
trained to understand that the product development process is an exercise in fluidity, and must abandon older
schools of thought that are focused on design freezes. By incrementally freezing design specifications throughout
the development process, you can avoid the time and money lost by having to go back and re-engineer specifications
that are no longer lucrative. Next, in order to achieve a cost savings from money spent on unnecessarily frequent
product development meetings or periods of indecision, it is critical that the company provide decision support tools
that aid designers and managers in making decisions correctly and efficiently so as not to extend the product
development timeline any longer than it needs to be.
Having a common framework and language under which decisions are made will empower development teams to
make the best choices possible while maintaining managerial confidence that decisions are being made in the best
interests of the company. Third, in order to speed the time to beginning design work, managers must approve a
process by which manufacturing parts can be approved on a piece by piece basis as opposed to waiting for a single
“go-no-go” management checkpoint. By allowing individual portions of the product specification to be approved
progressively, management can avoid creating the unnecessary barriers to change associated with making an early
commitment to all parts used in building prototypes. And one last simple measure that can take place is to begin
instituting metrics associated with the product development process that can act as warning indicators when the
process is starting to get out of control. This will allow the company to be proactive in responding to emergencies
before they become too late address.