1 Technological progress is a major driver of organizational change. Organizations may flourish of fail depending on their capability to recognize what technological innovations matter to them and how they should respond. This lessons examines the digital revolution, or fourth industrial revolution, a major technological progress, which in recent decades had major disruptive effects on organizations. 1. In the First part we will explore the roots, nature and size of the digital revolution, addressing questions such as what is the digital revolution? why it is happening now? 2. The Second part explores change that the digital revolution is producing in organizations. 3. The Third part explores the main challenges in change in response to disruptive innovation, by using lessons from past technological shifts. 2 3 Typically, you would point out to your wrist, as if there was a clock…. Yet how many of you have a clock at their wrist now? 4 We are so used to the steady flow of new technologies that it may seem a normal condition in human history But is this truly the case? 5 Lets’ take a moment to contemplate the recent pace of human development, as described in the book “Sapiens. A brief history of humankind” by Juval Noah Harari. He wrote: “A Spanish peasant who had fallen asleep in the year 1000 and woken up 500 years later, to the din of Columbus’ sailors boarding the Niña, Pinta and Santa Maria, the world would have seemed to him quite familiar. (….) But had one of Columbus’ sailors fallen into a similar slumber and woken up to the ringtone of a twenty-first-century iPhone, he would have found himself in a world strange beyond comprehension.” 6 The last 500 years have witnessed a phenomenal and unprecedented growth in human power. In the year 1500: • there were about 500 million Homo sapiens in the entire world. Today, there are 7 billion • the total value of goods and services produced by humankind estimated at $250 billion, in today’s dollars. Nowadays is close to $60 trillion • humanity consumed about 13 trillion calories of energy per day. Today, we consume 1,500 trillion calories a day • Five modern freighters could have taken onboard all the cargo borne by the whole world’s merchant Fleets • A modern computer could easily store every word and number in all the codex books and scrolls in every single medieval library with room to spare 7 For many thousands of years, humanity was on a very slow upward trajectory. Progress was almost invisible. Animals and farms, wars and empires, philosophies and religions all failed to exert much influence. But just over two hundred years ago, something sudden and profound arrived and bent the curve of human history—of population and social development— almost ninety degrees. It was the steam engine or, to be more precise, one developed and improved by by James Watt and colleagues in the second half of the eighteenth century to have unleashed an unprecedented growth in human power. 8 Cotton gin - cotton engine" – is a machine that quickly and easily separates cotton fibers from their seeds, enabling much greater productivity than manual cotton separation The cotton gin, for example, was a pivotal innovation in the textile sector in early 1800, but not outside of it. Instead, the steam engine and electrical power impacted all sectors, generating many more innovations 9 Prior to Watt, steam engines were highly inefficient, harnessing 1% of the energy released by burning coal. With Watt, steam engines become much more efficient, it massively increased the amount of power available and had a cascade effect. Watt’s increased this more than threefold. The steam engine massively increased the amount of power available to factories and free them from the need to be located near a stream or river to power the water wheel. The Industrial Revolution is not only steam power, but this innovation had a cascade effect, overcoming the limitations of human and animal power, enabling factories and mass production, railways and mass transportation. 10 “The First Industrial Revolution used water and steam power to mechanize production. The Second used electric power to create mass production. The Third used electronics and information technology to automate production. Now a Fourth Industrial Revolution is building on the Third, the digital revolution that has been occurring since the middle of the last century. It is characterized by a fusion of technologies that is blurring the lines between the physical, digital, and biological spheres.” (source: world economic forum) 11 There are three reasons why today’s transformations represent not merely a prolongation of the Third Industrial Revolution but rather the arrival of a Fourth and distinct one: speed, scope, and systems impact. The speed of current breakthroughs has no historical precedent. When compared with previous industrial revolutions, the Fourth is evolving at an exponential rather than a linear pace. Moreover, it is disrupting almost every industry in every country. And the breadth and depth of these changes herald the transformation of entire systems of production, management, and governance.” 12 Digital progress is doing for mental power, what the steam engine and its descendants did for muscle power. It allows to overcome its traditional limitations. While organizations have been using computers for decades, yet - just as it took generations to improve the steam engine to the point that it could power the Industrial Revolution, it’s also taking time to refine our digital engines. In the last decades we achieved a turning point, and the pace of technological impact is now surprising even experts 13 Several technological breakthrough went from being the stuff of science fiction to on-the-road reality in a few short years For example: In 2002 the US Defense Advanced Research Projects Agency (DARPA), launched a Grand Challenge to build a completely autonomous vehicle that could complete a 150-mile course. The race took place on 2004, with 15 participants. The “winning” car- from Carnegie Mellon university covered only 7.4 miles before getting stuck. In October 2010 Google announces a completely autonomous cars had been driving successfully, in traffic, on American roads and highways. TESLA cars on the market have auto pilot. 14 In 2004, experts admitted that “Human-level speech recognition is an elusive goal,” but .. in October of 2016, a team from Microsoft Research announced that a neural network they had built had achieved “human parity in conversational speech recognition,” … being more accurate than professional human transcriptionists. Since 2012 Forbes corporate earnings previews that appear on the website are generated by algorithms without human involvement. 15 To understand why it’s unfolding now, we need to understand the nature of technological progress in the era of digital hardware, software, and networks, namely three key characteristics: exponential, digital, and combinatorial. 16 17 The game of chess originated in present-day India during the sixth century BC. The inventor traveled to the capital city to presented the game to the emperor. The ruler was so impressed by the beautiful game that he invited the inventor to set his reward. The inventor proposed to place one single grain of rice on the first square of the board, two on the second, four on the third, and so on, so that each square receives twice as many grains as the previous.” The emperor agreed, impressed by the inventor’s apparent modesty. At half of the chessboard (thirty-two squares), the emperor had given the inventor about 4 billion grains of rice. That’s about one large field’s worth—and the emperor did start to take notice. However, in the second half of the chessboard numbers mount into trillions, quadrillions, and quintillions—we also lose sense of how quickly numbers like these appear as exponential growth continues. 64 instances of doubling yields more than eighteen quintillion grains, which is more rice than has been produced in the history of the world. The anecdote illustrates the power of sustained exponential growth and highlights the point at which the numbers start to become so big they are inconceivable. Source: Brynjolfsson, E., & McAfee, A. 2014 citing Kurzweil 2000 The Age of Spiritual Machines: When Computers Exceed Human Intelligence 18 Moore’s law and the second half of the chessboard In a 1965, Intel cofounder Gordon Moore observed that over the 10-years history of integrated circuits industry the amount of computing power you could buy for one dollar doubled each year. He predicted this would continue for at least another ten years- which meant a 500 more powerful potential in 1975 than in 1965. After five decades the law is still true, as well as for digital progress in other areas as well. Moore’s Law is not a laws of physics, of course, but it signals the constant successful efforts of the computer industry’s engineers and scientists. 19 Exponential growth leads to staggeringly big numbers. It explains why progress with digital technologies has been so much faster these days and why we’ve seen so many recent technological breakthroughs. “It’s because we’re now in a different regime of computing: we’re now in the second half of the chessboard. Cars that drive themselves, auto-generated news stories; cheap, flexible factory robots—have all appeared since 2006. Things get weird in the second half of the chessboard. And like the emperor, most of us have trouble keeping up.” (Brynjolfsonn and McAfee 2012) The BEA first noted “information technology” as a distinct corporate investment category in 1958. If we take that year as the starting point for when Moore’s Law entered the business world, and used eighteen months as the doubling period. After thirty-two of these doublings, U.S. businesses entered the second half of the chessboard when it comes to the use of digital gear. That was in 2006. 20 In 1996 the Accelerated Strategic Computing Initiative (ASCI Red) funded with 55 million $ by the US government was the world’s fastest supercomputer. It occupied nearly a tennis court of space. ASCI Red was the first computer to score above one teraflop—one trillion floating point operations* per second, using the energy that would power 800 homes. In 2006 the same performance was met by Sony PlayStation 3, for 500 $ dollars and 200 watts. In 10 years, exponential digital progress brought teraflop calculating power from a single government lab to all around the world. 21 22 23 Digitalization: turning information—text, sounds, photos, video, data from instruments and sensors, into the ones and zeroes that are the language of computers. Digital information has two key properties: - it is non-rival, namely it is not consumed or extinguished when it gets used by someone. For instance, an apple is a rival good, a news article is a website is not - it has close to zero marginal cost of reproduction, namely it is extremely cheap to make another copy These traits of digital information open new opportunities and technological developments. Let’s consider a few examples of new devices and products which are possible thanks to digitalization. 24 GPS-based app that provide driving directions tell you not only what route to your destination is best in general, but what route is best in that specific moment– considering traffic jams, accidents, road closures, etc. They make use of subscribers’ smartphones, which upload constantly to the company’s servers their location and speed information. 25 When translating from language X to language Y, Google Translate scans all the documents it available in both languages, looking for a close match, then returns the corresponding text. Namely what is does is statistical pattern matching over huge pools of digital content. 26 Internet is a tremendous source of user-generated information. A team of US used information on housing search to predict house price evolution, and the results outcompeted the predictions published by the experts at the National Association of Realtors. 27 Sensors like microphones, cameras, and accelerometers have moved from the analog world to the digital one and became subject to the exponential improvement of Moore’s Law. A Google autonomous car incorporates several sensors - the most important ‘eye’ being a LIDAR (a combination of “LIght” and “raDAR”) mounted on the roof. The first commercial LIDAR systems costed around $35.000.000 in year 2000; in year 2019 they cost 4.000$ * 28 29 Most innovation are not completely new but derive from recombining things that already exist, namely they result from combinatorial innovation. For example: • The sewing machine was enabled by the push for interchangeable parts in the late XVIII century munitions industry • Johannes Gutenberg applied the technology of the screw press that was designed for making wine and reconfigured it with metal type to create the printing press 30 “To create is to combine existing bits of insight, knowledge, ideas, and memories into new material and new interpretations of the world, to connect the seemingly dissociated, to see patterns where others see chaos.”* The more we have access to a pool of information and ideas that are diverse, cross-disciplinary, and wide-ranging, the more we can put them together in new combinations for new insights and ideas. Using a metaphor, “if you only have a few LEGO pieces of the same shape and color, what you can create is limited, while if you have many LEGOs made up of various sizes, shapes, and colors, you can create an almost infinite number of interesting combinations.”* *https://www.smithsonianmag.com/innovation/combinatorial-creativity-and-the-myth-oforiginality-114843098/?no-ist 31 Today, we have Internet components–software, protocols, languages, etc.— that are being combined to create totally new innovations. Such component parts are all bits. That means you can reproduce them, duplicate them, spread them around the world, and you can have thousands and tens of thousands of innovators combining or recombining the same component parts to create new innovation, and so we get a tremendous burst of innovation. The self-driving car is a current example of a whole lot of different technologies—digital mapping, GPS, machine learning, developments in laser and infrared sensor technology–coming together to create something truly innovative. The speed of combinatorial innovation may continue to accelerate in the digital age where new pieces of software can be sent around the world in a matter of seconds innovators everywhere can combine and recombine this software with other components to make new innovation. Innovators everywhere can combine and recombine then software with other components to make new innovations On top of that, through the internet and digitalization, we can access an unprecedented amount and variety of information 32 33 Digitalization is producing three major organizational transformations, by changing the balance between machine and minds, product and platform, core and crowd. Organization need to reflect on whether and how to implement such transformations. 34 35 Artificial intelligence is the “science and engineering of making intelligent machines.” The first conference on the topic was held in 1956 and a few years later the field divided into two approaches. An example is useful to understand the two approaches’ basic premises. Children learn a language by listening, their brain operates on statistical principles to discern the patterns in language, and they do not need to learn rules to learn a language. Instead, adults’ brain are different. When learning a second language they rely much more on learning explicitly rules. AI community split into two similarly differentiated camps. One pursued socalled rule-based, or “symbolic,” artificial intelligence, while the other built statistical pattern recognition systems. At first, it looked like the symbolic approach would dominate. Examples include programs to play chess or to solve theorems. However, other challenges proved impossible, such as in speech recognition, image classification, language translation, by the late 1980s major corporate and governmental sources of research funding dried up. 36 Two major obstacles explains the failure of symbolic approaches. First, there are a too many rules for certain tasks. Attempts to codify all of them have failed (e.g., in complex things like languages or image recognition). The worlds we inhabit follows several and inconsistent set of rules. For instance, in English two negatives can make a positive (“she is never not cheerful”), but two positives can never make a negative. Our human common sense allows us to live in the world’s complexity and inconsistency, but this knowledge cannot be codified. This leads to the second main obstacle, namely that commonsense reasoning processes are largely unavailable to introspection, we don’t and can’t know what rules we ourselves are using. This is known as the Polany’s paradox, namely that we know more than we can tell. In such cases, a rule-based system cannot be developed. 37 The other camp of artificial intelligence has tried to overcome Polanyi’s Paradox by building systems that learn tasks the way a young child learns language: by experience,d repetition, and feedback. They’ve created the field of “machine learning”. For a long time, progress in this camp also remained very slow, until recent incredible achievements – e.g., in language recognition, image analysis, etc. These breakthroughs have become possible primarily because of the Moore’s law and related astonishing increase in computer power and the phenomenon of “big data”, namely the explosion of digital text, pictures, sounds, videos, etc. Like young children, who hear a lot of words and sentences to learn language, machine “learns” exposed to many examples and provided by computing power to classify and make order in them. 38 The balance between machine and humans in work has changed drastically over the last two centuries. Steam, electric power, combustion engines freed humans and animals from most tasks requiring pure strength. Since the third industrial revolution (ICT), a shift of intellectual tasks has occurred too. Algorithms take decisions and make forecast – for example judging credit worthiness. Many decisions are nowadays made by algorithms, and we do not even notice. What tasks will be left for humans in the future? This has clearly huge consequences on how work will be organized in the future. 39 In 2004, Frank Levy and Richard Murnane wrote The New Division of Labor, focusing on the labor division between humans and computers, assuming that people should focus on tasks where they have a comparative advantage over computers. The put information processing tasks on a spectrum. On the one side are tasks that require only the application of well-understood rules, like arithmetic. Since computers excel at following rules, they should do arithmetic and similar tasks. For example, a person’s credit score is a good general Predictor of whether they’ll pay back their mortgage as promised. So, the decision about whether to give someone a mortgage can be boiled down to a rule and computers can take up this task. In turn, machines would take care of basic math, record keeping, and data transmission. This would free up people to make decisions, exercise judgment, use their creativity and intuition, and interact with each other to solve problems and take care of customers. At the other side lie information processing tasks that cannot be boiled down to rules or algorithms. For example, while driving a person faces a constant flow of diverse and complex information and stimuli, produced by other cars, drivers, pedestrians, 40 they might have to choose between formal and informal rules according to different contexts, and so forth. The same occurs with speech recognition or walking in the street. These tasks rely on human capability to take information via our multiple senses and recognizing patterns. Humans are very good at this, although we are often not able to describe and explain how are doing such task (which is know as Michael Polanyi’s paradox “We know more than we can tell”). According to Levy and Murnane, these tasks cannot be computerized and will remain in the domain of human workers. 40 This is counterintuitive. We usually think of sophisticated tasks as those related to human rationality as a unique skills compared to animals. Instead, “The main lesson of thirty-five years of AI research is that the hard problems are easy and the easy problems are hard”*. Or, as in the Moravec paradox that “contrary to traditional assumptions, high-level reasoning requires very little computation, but (…) it is very difficult or impossible to give them the skills of a one-year-old when it comes to perception and mobility”. * quoting Stephen Pinker Several “easy” tasks will remain for long in human’s realm, namely “As the new generation of intelligent devices appears, it will be the stock analysts and petrochemical engineers and parole board members who are in danger of being replaced by machines. The gardeners, receptionists, and cooks are secure in their jobs for decades to come.”* quoting Stephen Pinker 41 Actually not even the “easy” are easy.. 42 Another common belief is that as computers take care of routine tasks, people will be freed to make the more sophisticated decisions, to exercise their judgement and make decisions. This is clearly a core aspect of organizations, namely what is the role of machines and minds in taking decisions? From very simple to more complex ones? However, decades of research in cognition have revealed several limitations of humans’ decision making process. Most prominently, 1. We are easily overloaded by information, and in these situation with filter out useful and important information 2. We need meaning so much that if information if lacking or inconsistent, we see patterns and meanings fed by our own assumptions and beliefs, or we fill the gap base on our memory – which is also itself selective and limited 3. When pushed to act fact we tend to we jump to often flawed conclusions 43 Therefore, human judgment can often be replaced by machines with better results. For example, decisions on whether someone should receive a loan or not may be affected by prejudices, as well incomplete or inconsistent use of the information available to the decision makers. As a matter of fact, this is one of the first case of automated decision making in the era of corporate computing, through the development of score of creditworthiness. Progress in digital data and computing make more and more decisions managed by computers alone. For example, companies like Google run some of the world’s largest data centers, which are extremely energy consuming. Humans managed the equipment that keep data centers at the right temperature. Then, Google took years of historical data on data centers’ computing load, sensor readings, and environmental factors like temperature and humidity and used this information to train their algorithms to achieve better energy efficiency – which led to 40% less energy consumption. Automated decisions nowadays include giving recommendations for each client online, and all sorts of transactions that happens with no human involvement. 44 In some cases, instead of having machines provide data as an input to human judgment, human judgment serve as an input to an algorithm. For example, job interviews often confirm what someone already think of a candidate rather than truly assessing the candidate. Google, now rely heavily on structured interviews, a set of predefined questions where the interviewer indicates how did the candidate perform, but they’re quantified and used to assign a numeric score to job applicants. In other cases, humans have an important role of making the judgement effective. For example, in the field of medical diagnosis, computers outperform humans’ assessment in most specialties—radiology, pathology, oncology, and so on— thanks to the new availability of digitalized data (e.g. images, screening, etc.). In such an areas the computer make better assessments. Most patients, however, want to get their diagnosis not from a computed but from a person that helps them understand and accept it. In some cases Humans complement computers. For example, humans still hold an important role in decisions thanks to their openness to a diversity of information that we don’t preselect. While this creates confusion as we have difficulties in selecting the most useful signals, computers have the opposite problem, namely they only gather and analyze the data their programmers allowed to see. 45 This can create severe failures. For example, UBER automatically adapt prices based on the demand in a local area. In 2014, following a terrorist attach in Sydney, the prices in the area of the incident automatically and suddenly spiked and UBER was heavily criticized for being cynical. Hence, the decision-making process was changed, enabling human supervision to prevent similar problems. 45 46 The iPhone was launched on the market in 2007. It was a groundbreaking product in terms of design and features, like touch screen, accelerometer, GPS, etc. and many attributes of computers. For example, like computers have programs, iPhone had applications, called “apps”. The then leader of Apple, Steve Jobs, was known to hold a tight control on companies’ products and intended to maintain control over the apps as well– which could not be introduced from external developers. After a year, Jobs changed his mind. Nowadays we could not conceive a smartphone without thousands available apps created by independent developers. The apps act as a so-called “complements” to a smartphone: when I buy an iPhone, I also benefit from the related apps, which increase the value of the product, and hence how many people will buy it. In turn, iPhone also works as a platform for developing apps, and by turning to an open platform it greatly increased the value of the product. Of course, open platforms need to be curated, controls and boundaries are set by the owner of the platform, but in the current digital economy the potential of open platform is often much bigger than closed ones. 47 Platforms can be defined as online environments that take advantage of the characteristics of information in a digital environment. Namely, once information is produced, it is free, perfect, and instant to replicate – hence, a platform is digital environment characterized by near-zero marginal cost* of access, reproduction, and distribution. Digital platforms facilitate economic activity, they represent a new way to match product and client. They can create a new market or affect existing markets to a different extent. * Marginal cost means the cost of an additional unit 48 Digital platforms facilitate economic activity, they represent a new way to match product and client. They can create a new market or affect existing markets to a different extent. For example: In the case of the iPhone, the platform not only acts as a complement to the product, but also connects users to producers of apps Facebook is a product - platform and connects users and companies through ads Blockbuster was a store where to rent movies in VHS/DVD and it has been gradually replaced by illegal platforms (e.g. Torrent, etc.) and legal ones (e.g. Netflix) Amazon disrupted shops and malls Uber enables car rides, increasing the competition to traditional taxis Airbnb enables many more suppliers, and mostly affect the lower end of the market BlaBlaCar matches people driving from one place to another with passengers who want to make the same trip. It has small effect on other suppliers (e.g. train) 49 Platforms are particularly good at affecting existing organizations/providers when there is little difference among their offerings. For example, getting a ride from place A to B is a largely undifferentiated experience, as a traveler basically wants to go somewhere quickly, safely, and cheaply. Travelers' accommodations are not undifferentiated, there are much differences and that is the reason why hotels have not been displaced by Airbnb- their offerings are somehow different one from another. This is important to consider for activities that do not want to be displaced by platforms. question 1. Can platforms affect the Higher Education sector? How? 2. How can a university tackle the competition of platforms? Or use platforms? For example, some universities that have created their own platforms of courses for distant education, potentially threatening universities in distant places. However, Higher education is not an undifferentiated good, so the impact of online courses have been so far limited. At the same time, if 50 universities do not want to be displaced in the future, they must preserve the distinctiveness of what they offer. 50 Platforms that connects users and suppliers strongly benefit from network effects, namely the more the users and suppliers join the platform the more the advantages on both sides. Such network effects are so strong that in a competition between platforms, the bigger may outcompete the others largely based on such effect, in a selfreinforcing cycle that leads the winner to take the whole market. As a matter of fact, the most renown platforms have quasi-monopolistic positions in their respective markets – e.g., Airbnb, Uber, Amazon, Alibaba, Booking. 51 52 In late 1999 Jimmy Wales and Larry Sanger started to build a free universally accessible online encyclopedia: Nupedia. In order to contribute writing an entry, one had to be a true expert in the field and possess a Ph.D. After eighteen months and $250,000 of spending, Nupedia had only twelve completed articles. 53 In mid-January 2001 a new website was launched. It was based on a different approach in which any user could make a contribution, edit someone else’s contribution, or undo any previous edit. By the end of January, it contained 617 articles. By the end of 2001 there were 19,000. By 2016 there were 36 million articles across 291 languages. 54 Wikipedia, unlike Nupedia, activated contribution not from a closed group of experts (i.e. the Core) but potentially from any external person, i.e. the crowd. It did so by adopting the principles of openness and noncredentialism, meaning that everyone was invited to contribute and did not have to possess specific credentials, and of self-organization, namely everyone decided what to contribute to. 55 Modern organizations face complex environments and new problems. In order solve a specific problem, every organization has a set of competences, knowledge, expertise among their members: i.e., the Core. When missing specific expertise and knowledge, organizations can also rely on external consultants or opinion polls, which still represent a kind of semifixed set of expertise. Digitalization enables access to a much wider pool of knowledge, external to the organizational boundaries and often for free, the so-called ‘Crowd’. 56 The crowd has an immense problem-solving potential. For example, innovation scholars Karim Lakhani, Kevin Boudreau organized more than 700 challenges involving non-experts to solve problems of organization such as NASA, medical school, companies, and the solutions such groups found was always similar in quality or better that the internal solutions. WHY IS THE CROWD MORE EFFECTIVE THAN THE CORE IN ADDRESSING MANY PROBLEMS? 57 There are two main reasons explaining why the crowd so-often outperform the Core in finding a solution to a problem. 1. Expertise in the Core is necessarily limited and suffers from obsolescence 2. In solving problems, it is important to have different perspectives, multiple dissimilar backgrounds, educations, approaches, etc. First, an organization can face so many different and changing problems that the core will never have the expertise to address all of them. Moreover, its expertise tend to become gradually out of date. Instead, the crowd is so big and constantly renewing that won’t have such limit. Second, in order to solve problems, it is important to have different perspectives, people with multiple dissimilar backgrounds, educations, approaches, etc. This is a trait of the crowd, whereas organizations can only afford to hire experts whose skills are close to the problems that they typically met. For example, innovation scholars Lakhani and Jeppesen studied 166 scientific challenges posted on InnoCentive and found that the ones most likely to be successfully solved were those that attracted people who were technically or socially “far away” from the organization that posted the challenge. 58 The key organizational consequence of this process is to rethink the role of Core and the Crowd, to reflect upon the opportunities of Crowd and on ways to access the knowledge of the Crowd 58 59 60 Until the twentieth century, factories were powered by a single big steam engine that was typically located in the basements of factories. At the time, electricity appeared as a viable alternative. Once they were made electric, power sources could spread throughout a building and there could be several power sources instead of one huge one. Today it is unconceivable to do differently than this, but the concept of several powers sources was first met with skepticism. Most successful incumbent companies did not survive the transition from steam to electric power which occurred one century ago. Electrification was one of the most disruptive technologies ever, it caused something close to a mass extinction in US manufacturing industries. As a matter of fact, many organizations fail to adapt technological change. Why does this happen? There are three main challenges to adapt to a disruptive change, namely to 1. recognize that a relevant change is occurring, 2. decide that it is necessary to act, and 3. in implementing technological transformation. Let’ analyze these challenges more in detail. 61 62 The first step to change an organization in response to technological change is arguably being able to recognize that a relevant change is happening. Recognizing that a relevant change is happening is more difficult that it might appear at first glance. Disruptive innovations typically starts in marginal or new markets segments and grow from the ‘periphery’ to the core of the market and change the way customers perceive the product. Because they happen initially in a marginal market segment, these innovations only produce ‘weak signals’ that an incumbent organization (i.e., an organization which is among the market leaders) may be unable to detect or to believe in its importance. Part of the problem is that organizations tend to focus on their current customers’ needs. However, focusing too heavily on current customers’ needs can blind the company to a change which is happening in less important market segments. 63 Disruptive innovations make something so much simpler and more affordable, so that a new population of customers get pulled into the market. At the beginning, however, the performance of the disruptive technology is not as good as the performance of the established products that are sold by the incumbents. For example, in mid 80s offices typically had photocopy centers for printing. Xerox was the market leader producing high-speed, fully featured machines in such photocopy centers. Then, Canon introduced in the market smaller table copiers which could only do three or four copies a minute, with no possibility to collate, enlarge or reduce or do grayscale. Xerox customers were fully satisfied so the company did not get a clear signal that the Canon product could become important. However, Canon had its niche because for the simple things, the little Canon around the corner was much better. This granted Canon the time and resources to gradually improve until an entirely new market was created. Xerox missed this opportunity, paradoxically, because they listened to their customers. How could Xerox have picked up the weak signal and understand that the small copier was going to be important? 64 How could Xerox have picked up the weak signal and understand that the small copier was going to be important? Clayton Christensen – a major expert on the topic of disruptive technological innovations – argues that the attention should rather be on the job and in watching customers working. In this case, they would have seen that in many cases the workers did not have the time to walk to the corporate copy center and preferred to do the little additional work with the (initially) slow and lowquality printer. The questions driving organizations’ attention should be whether there is someone who can’t do something because the solution is just too expensive and too inconvenient. Outstanding companies can do everything "right" and yet still lose their market leadership – or even fail – as new, unexpected competitors rise and take over the market. There are two key parts to this dilemma. Value to innovation is an S-curve: Improving a product takes time and many iterations. The first of these iterations provide minimal value to the customer but once the base is created then each iteration is drastically better. At some point the most valuable improvements are complete and the value per iteration is minimal again. So, in the middle is the most value, at the beginning and end the value is minimal. By the time the new product becomes interesting to the incumbent's customers it is too late. 65 Even when recognizing the importance of an ongoing technological change, organizations may not be willing to adapt. Because incumbents are so proficient in what they do, they are victim of a status quo bias, or knowledge curse, as they are reluctant to abandon their business model or technology for what is initially only a marginally superior and less profitable option. Moreover… 66 67 Consider this excerpt from an interview to Clayton Christensen: “You could imagine that, in a doctor’s office somewhere, a nurse observed the gross dissatisfaction among the customers and saw this great opportunity. (..) But the doctor’s office is structured to make money (..) What they need is a way to be reimbursed $175 per visit (..) and the nurse comes at them with this $40 per visit idea. It just doesn’t make sense in the business the way it’s structured. That’s why disruptions are very hard to deploy within an established business.” What kind of trap is the incumbent organization in? 68 Intel recognized the change, but it did not feel necessary to adapt because it was focused on profitability and their most important clients. However, they later recognized that such a niche could become a serious threat in a near future. What was happening at Intel? Why they decide to act? Intel recognized the change, but it did not feel necessary to adapt because it was focused on profitability and their most important clients. However, they later recognized that such a niche could become a serious threat in a near future. 69 70 Some general lessons can be drawn from the case regarding organizational change and more specifically in response to technological innovations. Resistance due to different implicit interpretations of the organization’s priorities reasons for change need to be fully understood throughout the top and middle management in the short term, the dominant logic of profitability often clashes with the need to invest in a different course of action a strong leadership need to invest time and effort in spreading a vision, a common understanding and language – beyond a simple slogan Resistance or lack of commitment is not necessarily driven by self-interest, but rather by a different interpretation of the events and of the organization’s priorities Hence, implementing change requires in the first place that the reasons underpinning the organizational need to change are fully understood throughout the middle management Responding to technological change is particularly tricky for incumbent organizations because the dominant logic of profitability clashes in the short term with the need to invest in a different course of action It is therefore essential a strong leadership, which invests time and effort in spreading a vision, a common the understanding and language – which go beyond a simple slogan 71 72 As for Intel, the case of Kodak highlights the importance of a shared vision on the appropriate course of change. The new leadership had a traditional focus on profitability, they did not understand the necessity for change, they reversed the course of action and Kodak failed. Moreover, it provides further insights. First, a major challenge is that the organization must excel in two different technologies at the same time, which is very expensive. Think, as a further example, to car companies shifting to electric; for a long period of time, they will have to go on investing in people and infrastructures dedicated to traditional internal combustion engines, and at the same time open a completely new development path towards electric vehicles. Second, the new and old products or the traditional way of doing things are often underpinned by different logics, which hardly co-exist. In the case of Kodak, film was profitable and in the high end of the market, while digital was low profit and in the low end of the market. From an organizational perspective, the only solution was to create distinct divisions to enable different visions to succeed. 73 74 75