DEC is DEAD, Long Live DEC: The Lasting Legacy of Digital Equipment Corporation is by Edgar Schein, with Peter DeLisi, Paul Kampas and Michael Sonduck. Schein charts the creation of DEC, or Digital Equipment Corporation, which was created in 1957 and rose to become the second largest computer company in America before its decline during the 1990s, which ultimately ended with its acquisition by Compaq (although as clear in the title, DEC has left an enduring cultural legacy).
Schein’s argument is that DEC’s cultural values and practices resulted in it failing to adapt to the changing market. A recurring idea in Schein’s analysis is that DEC was missing the ‘money gene;’ the ‘cultural DNA’ of DEC encoded individualistic, technical and family values, which initially attracted extraordinary talent but remained non-negotiable even when they ceased to be compatible with profitability. Schein views the downfall of DEC as ultimately stemming from the difficulty of changing organizational culture, and an evolving company and changing market that left the previously effective culture of DEC toxic and inefficient.
I found the book useful for its insights into how the ‘cultural DNA’ of an organization has enduring impacts on organizational growth, and the resistance to changing core values that may arise even if change is necessary for the company to survive.
But in focusing on the culture of innovation at DEC as the primary factor that contributed to both its rise and decline, the book neglects the basic economics that underpinned this. DEC was innovative, but it was arguably economic factors that allowed the company to establish a monopoly position in the market. It is also possible to view DEC’s later decisions as stemming from economic concerns; specifically, a desire to hold on to the hope that they would continue to dominate the market with proprietary products, rather than an intrinsic commitment to innovation, led to them resisting change.
Table of contents
Chapter 2. Three Developmental Streams
Organizations do not fully control their fate — it is constrained by culture and technology.
(p.18)
The illusion that organizations can control their own fate stems from the failure to understand how technology and culture limit what is possible. We will see that as DEC pushed into new areas of technology, it had to make major trade-offs between developing innovations that pushed the technological limits and concentrating on commodities that were technically feasible. The cultural constraints entered the picture through the taken-for-granted assumptions and mental models of the founders and early leaders of that organization. Founders, investors, and leaders are not autonomous rational actors. Their own family, educational, and occupational backgrounds influence their values and assumptions.
Culture is an organization’s accumulated learning that has become taken for granted.
(p.20)
Culture in an organization can be thought of as the organization’s accumulated learning that becomes so taken for granted that it drops out of awareness (Schein 1992, 1999a). That learning covers both how the organization deals with its various external environments in accomplishing its primary tasks and how it manages its internal integration. If an organization is not successful in its early years, it will not develop a strong culture; on the other hand, if certain ways of thinking, feeling, and behaving continue to work, they become taken for granted and eventually drop out of awareness except when they are taught to newcomers as the way to get along in that organization.
At the core of culture are certain non-negotiable values and beliefs — that form its right to exist.
(p.23)
The most basic genes in the cultural DNA are the non-negotiable values and beliefs that creators of organizations claim as the basis for the right of that organization to exist. In the case of a technical entrepreneur these genes can be thought of as his or her technical vision that is sold to investors and ultimately to consumers. In the case of a religious movement it is the humanistic and spiritual values and beliefs of the founder that are initially attractive to followers. In the case of financial entrepreneurs it is the rationale of the deal they are trying to put together. However, these initial values and beliefs do not become shared and thereby become part of the cultural DNA until the organization succeeds and builds a shared history. The “organization” can be thought of at this stage as the “dominant coalition,” the network of executives, managers, and employees who share the basic assumptions and who mutually reinforce one another as the organization evolves. To decipher these key genes one must keep asking questions: What in the eyes of this dominant coalition keeps the organization afloat? By what right, in their view, does it exist? What is its primary task in the larger sociocultural context? What functions does it fulfill for society?
These beliefs are hard if not impossible to change — even when they prevent the organization adapting to a new environment.
(p.23)
The dominant coalition may recognize the need for certain new behaviors to adapt to changes in the technological environment, but if the learning involved challenges some of these non-negotiable beliefs and values, the leaders and members of the dominant coalition will not make the trade-offs necessary to acquire them. Insight and recognition are not enough to produce new skill sets if the gene demanding the outcomes of those skill sets is missing in the cultural DNA.
The money gene was missing from DEC’s cultural DNA, but introducing it would have required trade-offs incompatible with the values of DEC’s pre-existing culture.
(pp.24-6)
In the case of economic organizations in a capitalist society, their primary task and basic function is to provide a reasonable return to investors in the production of goods and services needed by the society and, in that process, to provide employment and technical and social innovations that help the larger society to adapt to changing environmental circumstances. For an organization to survive under these conditions it must have a gene that is concerned with making money, with economic growth and survival. The organization may have been founded on product, process, or service concepts that made it easy to make money initially, but sooner or later money per se becomes an issue as competition and technological evolution make the original idea economically less and less viable. The ultimate survival of the organization will then depend on the degree to which the commercial gene, or money gene, creates processes of innovation and adaptation that are geared to economic survival, even if that means abandonment of some of the original ideas, products, and services on which the organization was founded.
If we take this analogy into the DEC story, we will see that most of the genes in the DEC DNA were the technical and family values embedded strongly in an American individualistic tradition. DEC became a viable business because the basic individualistic, technical, and family values that Ken Olsen felt so strongly about created a management system that attracted extraordinary technical talent and produced a series of highly successful products that virtually sold themselves. In a sense, Olsen’s vision put DEC in the right place at the right time to “catch a major wave.” Ken Olsen anticipated a major societal value shift that henceforth would ascribe greater value to the person and would firmly place the individual instead of the monolithic mainframe computer in the center of the computing universe.
What was missing in this cultural DNA, however, was a set of genes for creating and sustaining a viable business, a commercial gene, a money gene, a set of shared values that would override the engineering and family values if those founding values became dysfunctional. This is not to say that DEC managers, including Ken Olsen, were indifferent to the values of making a profit, of giving a return to their shareholders, or of growing and stabilizing a business for the long haul. Ken Olsen cared deeply about profits and was proud to have produced profits in his very first year in business and every year thereafter until the late 1980s. At an espoused level, commercial values and the desire to run an effective profitable business were highly visible, and DEC’s management was continually reorganizing and developing new processes to improve “the business,” to increase the value of the stock, to enhance the company’s own economic well-being, and to protect the business for their stockholders and especially their employees. Over the years many professional managers were brought in to the organization who clearly had the commercial, or money, gene. But as we will see, the evidence that the money gene was missing in the basic cultural DNA was the unwillingness to honor those business values above the technical and family values. That would have required trade-offs that were never made.
For example, the presence of the money gene would have required earlier layoffs, pruning out some deadwood, setting clear priorities among development projects, killing some of their own obsolete products to free up resources for new development, designing products for new kinds of customers that were not seen to be glamorous, and giving more prestige to both marketing and finance as essential business functions.
Chapter 4. Ken Olsen, the Leader and Manager
Olsen believed in internal competition and “letting the market decide.”
(p.64)
Olsen clearly believed that competition among projects was a good motivator and that the market would ultimately decide which decisions were the correct ones. His faith in people, his belief in giving freedom, was a tremendous stimulus to creativity in all areas of company performance. At the same time, the early success and rapid growth of the company created enough cash flow to support a wide range of projects; there was no need to set priorities.
[Ed: crucial point we forget the structural stuff in the search for cultural reasons. Basically they had unlimited money and growth — they’d “hit the jackpot” technologically and economically. Money covers a multitude of sins. Reminds of a story told to me about IBM by a senior guy at a business school who’d consulted with them: basically they weren’t that good, at least by 70s, they just had a dominant monopoly position and reputation.]
He believed in loyalty and lifetime employment.
(p.67)
In 1994, Olsen expressed his feelings about loyalty:
It is common knowledge, today, and in all the literature that loyalty has disappeared from American companies…where [once] employees were loyal to the company and often stayed for their full working life, and companies were loyal to employees to the point where the relationship was solid often from generation to generation.
We did suffer from the loss of people who continuously were offered more money to go to other companies, but when there was a surplus of technical people [in the economy] and indeed, to some degree, a surplus of technical people within Digital, we showed loyalty to employees which meant after the economy recovered the employees had a sincere and deep loyalty to Digital which made them very effective employees.
[Ed: yes but you could be so loyal because you had so much money — you were literally swimming in it!]
Olsen’s neutrality and strategy of “letting the market decide” created an innovative culture that lacked efficiency.
(p.69-70)
The beliefs and values discussed in this chapter were interconnected and created a managerial climate that was, in my experience, unique. Olsen’s neutrality and willingness to go along with whatever a proponent could sell to his or her colleagues (“Doing the right thing,” “Getting buy-in,” and “Pushing back”) were among the most powerful forces for innovation that I have ever seen. It was pointed out over and over again in interviews that many of DEC’s innovations were not Olsen’s ideas but that Olsen created a climate of support for new ideas so that subordinates felt empowered to try new and different things. People learned that Olsen would argue against a position or a proposal but that he expected the final decision to come from the proposer. It was the proposer who would have to implement the proposal, so he or she should have the responsibility (“He who proposes does”).
This managerial climate created in Olsen’s subordinates a level of self-confidence, maybe even arrogance, that made it hard to reach efficient decisions in later years. Supporting everything and letting the market decide was a slow and often erratic process, but it became so much a way of working during the successful years that it was virtually impossible to convert to a more efficient hierarchical process when more management discipline and speed were needed.
Chapter 5. Ken Olsen, the Salesman-Marketer
Olsen’s sales philosophy was to sell only what would benefit the customer.
(p.74)
This philosophy worked well in building up a strong and loyal group of customers. The customers were eventually organized into DECUS (the Digital Equipment Corporation Users Society), a group that met regularly to exchange insights, learn about new applications, and provide feedback to the company on how things were going. With hindsight it is also evident that such a strong, loyal, and positive group of customers allowed Olsen and many other senior engineers to overlook the huge growth of another set of customers — individuals who wanted what PCs ultimately became.
Chapter 6. DEC’s Cultural Paradigm
Olsen created an environment where employees could grow.
(p.87)
Reflection from a former employee of DEC
Certainly, the timing of macro economic and technological factors played an important part in DEC’s success. However, it is much deeper. It was the employee culture that Ken created. It was an environment where people could grow, learn and make mistakes.
Chapter 7. DEC’s “Other” Legacy, Tracy C. Gibbons
Gordon Bell struggled to get buy-in for VAX and stress took a toll.
(p.104)
Bell was the architect of DEC’s VAX series.
It was hard to get buy-in for the VAX strategy, and it put him at odds with Ken Olsen. “I led from an understanding of the technology and what should be done from a technical strength point of view, and I was looking at it issue by issue. Ken and I disagreed. He was into options and playing them. . . . I made it happen, just kept on until the machines changed and the roles of the product lines changed.” But the stress took a toll; Bell had a heart attack and soon after left the company.
+worklife
Chapter 10. The Impact of Success, Growth, and Age
DEC’s cultural norms — such as unwillingness to impose on executives’ autonomy — impeded the enforcement of integrated systems.
(p.161)
The conflicts described above were observed at senior management levels, and as we will see in the next chapter, efforts were made to resolve them, but the cultural assumption that each executive was “Doing the right thing” to the best of his or her ability not only made it difficult to develop more integrated systems but also resulted in a clear reluctance to enforce the systems once developed. In these situations Ken Olsen became de facto the ultimate point of resolution and integration, but his response was usually to let the key parties sort it out for themselves. Olsen would not tell Ted Johnson or Andy Knowles what either of them should do, either because he believed that his “highly paid senior executives” should be able to reach a logical solution by themselves, or perhaps because deep down he would admit that he did not know what the right solution was. On the other hand, though he usually tried to remain neutral, his personal biases were known, and his subordinates learned to varying degrees how to calibrate him and how to get their own way.
More formal processes were needed, but Olsen resisted appointing a COO — issues of succession and the distribution of power became sensitive.
(p.164-5)
The Operations Committee continued to be the place where all of the fundamental issues described above were surfaced and debated. DEC was full of bright people who saw what was happening in the marketplace and the turmoil inside the organization. Ken Olsen’s commitment to openness led him to charge me, as the consultant to this committee, to regularly interview members, collect issues, and bring them back to the committee for discussion. After interviewing all the senior managers in 1972 and observing the rapid growth of DEC, it seemed clear to everyone that with continuing growth DEC would need an executive VP or chief operating officer. Olsen had far too many people reporting directly to him, and some formal processes were needed to manage the growing organization, yet it was clear that Olsen would not systematically enforce even the few formal processes that were in place.
Peter Kaufmann was the obvious choice for the COO role because he had shown himself to be a very effective and charismatic manager in building the worldwide manufacturing organization. He had the support of the other members of the Operations Committee and was willing to take a crack at the job. Win Hindle and I proposed this idea to Olsen. For reasons that were never entirely clear, Olsen perceived the suggested appointment as a threat to his own position, reacted very negatively to the idea, and misinterpreted it as a desire on Kaufmann’s part to usurp power. Why did Olsen react so negatively? One speculation was that Olsen had perceived Harlan Anderson’s efforts in the mid-1960s as being power seeking; another was that perhaps Olsen perceived any potential number-two person as a threat.
The problem of adding a COO remained unsolved. However, all of us had learned from this incident that Olsen was highly sensitive to perceived challenges to his power by any of his subordinates. He was himself conflicted about power in that he wanted to empower people, but those closest to him had to learn that in certain areas he wanted to retain complete control.
It became evident that distributing power widely below him also allowed Olsen to maintain control, something that he clearly needed. Kaufmann’s position in the group continued to be strong for a few more years, but he had lost credibility with Olsen and eventually left the company in 1977, when he felt that his managerial approach was better suited to a small-company environment and that DEC had grown too large.
Chapter 11. Learning Efforts Reveal Cultural Strengths and Rigidities
Remedies in conflict with cultural assumptions will not be adopted by an organization.
(p.173)
Too often we assume in our management literature that if we could just show people what is going on, they would act to fix things. We assume that insight leads to action. But just as therapists have learned that for various reasons patients often do not act on new insights, so one of the deep lessons of cultural dynamics is that if the remedy would require an organization to violate some of its deeply held cultural assumptions, that remedy will not be applied.
Shel Davis was unable to change Olsen’s style of management and Olsen eventually sent him away from Maynard.
(pp.179-180)
Shel Davis also developed a personal agenda based on his own value system. He deplored the way Ken Olsen vented his anger on people and decided that one of his personal projects would be to help Olsen to develop more constructive ways of dealing with his anxieties and angers. My own experiences along these lines had taught me that confronting Olsen directly on this matter would not work, so I counseled strongly against what Davis was trying to do. But Davis had strong values as well, and he could not abide what he saw as the sometimes brutal treatment of DEC managers and employees.
He tried many approaches to changing Olsen’s behavior, from direct counseling to various kinds of games, but none were successful. In fact, the main impact of Davis’s efforts to change Olsen’s style of emotional expression was that Olsen turned against him and, in characteristic fashion, mandated that Davis move to some other job well removed from Olsen. He would not fire Davis, but he would banish him from Maynard. Davis moved to Europe and lived out his career fruitfully in that environment, but we all had learned a lesson once again. If you fly too close to the sun, your wings melt and you crash. Davis had succumbed to the Icarus Complex and paid a personal price in terms of ultimate career accomplishments.
Chapter 12. The Turbulent 1980s: Peaking but Weakening
The introduction of measures of individual sales performance undermined cohesion at DEC.
(p.198)
The assignment of profit and loss responsibility to individual salespeople caused an inordinate amount of time to be spent fighting over who got credit for what and who got charged for what, rather than concentrating on the important aspects of growing the business. Perhaps more critically, the measurement of individual performance was strongly opposed to a culture that had always emphasized the good of the whole family, rather than that of its individual members. We were to learn how very difficult it is to get people to work together as team members when they are rewarded and held accountable for individual achievement.
The culture became one of intergroup competition and infighting.
(p.201)
Though Olsen gained prominence in subsequent years and continued to make strong efforts to control, or at least to guide, DEC, it was increasingly clear to me that DEC was more and more out of control and at the mercy of rampant intergroup competition and political infighting. As many of the departing managers and engineers said, it was harder and harder to tell the truth in this environment. For DEC as a total business, this climate proved too toxic.
By the 1980s the market had chosen the IBM PC, but Olsen was firm that DEC would not enter the market with a clone of the product.
(p.203)
Note that this comment reflects Paul Kampas’s analysis in chapter 9 of what happens when the market chooses a dominant design. Once that has happened, the door is opened for all kinds of category-killer companies to enter with faster and cheaper clones of the dominant design. Supnik continues:
The market was not going to take a proprietary personal computer from Digital. It had made up its mind. And I think where you can fault Ken is not for the first round of PCs but for the second round. Following the debacle of the Pro 350 and the Rainbow, the engineers had no illusions about the market, and the very next proposal from Engineering was a PC clone. Make a clone of the latest IBM machine, stay on their coat-tails, give it the Digital values of sturdiness and good industrial design, clean up the packaging and cabling, but fundamentally build a bit for bit, bug for bug copy of the IBM PC. It was called the DEC PC25 and 50 proposal. Now this is ’84. Compaq has not been founded or is just being founded, and the proposal was to do exactly what Compaq was going to do — a fast clone. But Ken killed it; DEC is not a copycat.
When this proposal was being debated, there was no one in Engineering who thought that making a clone was a bad idea or the wrong thing to do. Everyone said yes, this is obviously what you do, the market has gone here, and you do what the market says. . . Digital did have design, assembly and test, and distribution processes tailored to high volume projects, namely, terminal products. This proposal was effectively coming as an extension to the terminal group. We did terminals, okay so we’ll do PCs. The design and manufacturing processes were there, what wasn’t there was the willingness to have it be an IBM compatible PC from Digital, something that Ken would embrace fully in 1991. (Bob Supnik, interview by author, June 24, 2002)
Note that it was the deep assumption about product quality and DEC’s role as an innovator that was being challenged by the engineering groups proposing a clone. So even though all three PC entries failed to become successful products, the assumptions about what a product should be held firm. DEC would not compromise on quality or elegance, and DEC would not be a copycat. And this assumption was validated by the continuing success of the VAX strategy and networking with Ethernet.
Gordon Bell left the increasingly stressful environment following a heart attack.
(p.204)
As DEC grew and became differentiated into various functional and product empires, it became schizophrenic. Organizational health and toxic forces were running side by side, and the toxic forces were perceived but denied, rationalized, or absorbed by selected members who in turn became sick, notably Gordon Bell, whose heart attack in 1983 seemed clearly related to the stress levels that were building up within DEC.
Chapter 14. Obvious Lessons and Subtle Lessons
The strengths and weaknesses of a company are not visible from its public face.
(p.244)
LESSON 1. Don’t judge a company by its public face. Neither DEC’s strengths nor its weaknesses were visible from the outside. There is no way one could tell from its track record of growth to $14 billion when and where the causes lay for either its innovative strengths or its business weaknesses. Elements of the culture were certainly visible, and the enthusiasm of DEC’s employees sent a clear message of strength. But those cultural elements and that enthusiasm were still visible in 1992 when DEC was in deep economic trouble. In fact, during the Woods Meetings that summer, enthusiasm and optimism were still running high. It would have been difficult for an outsider to anticipate the rapidity of DEC’s decline as a business, and it would have been equally difficult to anticipate how strongly DEC alumni felt about the positive side of their culture.
A culture of innovation does not scale and becomes dysfunctional (though this interpretation of DEC overweights cultural effects).
(pp.244-5)
LESSON 2. A culture of innovation does not scale up; functional familiarity and “Truth through debate” are lost with size; “Do the right thing” becomes dysfunctional; managerial sense of responsibility changes with age and maturity; buy-in becomes superficial agreement.
Ed: Assumption that cultural tendencies which worked small did not scale up shows tendency to overweight cultural effects.
Instead you could simply argue that they could not switch tech paradigms. When you are successful in one paradigm it is extremely hard to switch to another paradigm — not for cultural reasons but because your success does not translate.
We don’t need complex explanations of why DEC’s innovation did not scale up. It’s simply that innovation is often of limited value — in a many horse race even if you are good you’ll mostly lose if there are other good horses. DEC one their first race (minicomputers) and in that area remained very dominant almost to the end. It was once the tech landscape changed and they had to re-compete that they lost.
Yes, it is true that Olsen missed the PC etc but even if they had been in there it would have been open competition …
A growing organisation must break away small units or cease to prioritize innovation.
(p.246)
LESSON3. If a culture of innovation works only at a certain small size, the organization must either find a way to break away small units that continue to innovate or abandon innovation as a strategic priority.
Growth based on technical vision can obscure business problems, which later require significant, often unappealing work to address.
(p.248-9)
LESSON 7. Successful growth based on a technical vision will hide business problems and inefficiencies until an economic crisis reveals them or until the business gene is switched on; recognition of those problems will not necessarily produce remedial action.
…
Furthermore, dealing with the business problems requires switching from the “fun” of innovation and growth to the “hard work” of creating a business strategy process, of becoming cost conscious, of changing organizational routines toward efficiency, of laying off deadwood or obsolete technical talent, of developing more precise measures of economic return that could lead to the killing of some products (eating your own children), of modifying the basic processes in engineering and manufacturing to respond to changing technology, and of allocating increasingly scarce resources. For the manager who has the business gene, these tasks can be fun, but for the manager who is the technical visionary/innovator the fun is in creation. Many observers of DEC pointed out that DEC management had neither the inclination nor the skill to do this hard transformational work.
### Members of a culture of innovation are unlikely to want to introduce the business gene or see the necessity of doing so — it must be introduced by the board.
(p.249)
LESSON 8. If a growing business lacks the business gene, the board must act to introduce that gene. The technology gene is clearly enough to start a company on the road to success, but at some point trade-offs will be required that will not be made unless the business gene is present. It is unlikely that the members of the innovative culture will perceive this necessity as long as they are having fun, or if they perceive it, that they will act on it. By the time they recognize the problem — for example, DEC in the early 1990s — it is probably too late to do anything. The board, acting on behalf of the shareholders, must then intervene to introduce the business gene into the organization’s cultural DNA. How this is done will depend on local circumstances, but the board may be the only entity that sees the problem early enough to take effective remedial action.
DEC’s culture of innovation meant it never gave up on having proprietary products (but this overweights culture — there were economic reasons for DEC’s attachment to the old model).
Why DEC did not pursue its own mission more aggressively and develop the Internet is a puzzle except in the context of the previous lesson, that DEC continued to try to do everything and never got sufficiently focused on networking. An alternative cultural explanation offered by Paul Kampas is that DEC never gave up on its need to have proprietary products.
[Ed: but it’s not a cultural reason!! It’s an economic reason. Getting out of proprietary products would have meant huge reduction in margins and a realisation you couldn’t do innovation anymore. You always hope you are wrong about having to make such a choice and keep hoping against hope your old model will keep working …]
(p.252)
Many observers felt that DEC not only lacked the business gene but also was obsessed with the vision of the minicomputer.
[Ed: more evidence for the previous point. Yes they were obsessed but getting the PC or the Internet would have meant abandoning what you already had and were the best at for something that you weren’t necessarily the best at …]
Strong customer focus may involve listening only to a segment of customers, and so mislead and fail to reveal important information.
(p.252)
LESSON 12. The value of“listening to your customers” depends upon which customers you choose to listen to. A company can start out to be very customer oriented yet fail to detect market changes because it continues to listen to and respond to only one customer segment. DEC’s philosophy from its very beginning was to be intensely customer focused, and it held this philosophy as a moral obligation. The sales function was there to help sophisticated customers figure out their problems and to involve customers in the solution. Anytime a customer had a problem, sales was ready to help, which led to a great many separate projects all over the organization. Customers were organized into a users’ group (DECUS), and this group continued to love DEC products. DEC managers paid constant attention to them and therefore believed sincerely that the company was highly customer oriented.
The governance system of an organization must evolve if it is to continue to be functional as the organization grows.
(p.253)
LESSON 13. The type of governance system an organization uses must evolve as the organization matures. Warren Bennis once wrote that “democracy is inevitable,” referring to the degree to which large bureaucratic organizations are increasingly finding it necessary to empower their employees. The DEC story suggests a reverse twist on this prediction: “hierarchy is inevitable.” DEC worked very well for a couple of decades with a minimum of hierarchy, but as it grew and aged, the negotiations in the matrix structure became more difficult as functional familiarity was lost among the players and as shrinking resources required more prioritizing.
Hierarchy may not be the only solution, or it may have to metamorphose into other forms, but some coordination or integration mechanism must evolve as the organization grows and differentiates — an effective strategy-building process is one such mechanism; charismatic leadership is another.
Debate between knowledge workers becomes an increasingly less effective means of decision-making as an organization grows.
(p.253)
LESSON 15. Knowledge workers cannot make efficient decisions together. One of the most striking phenomena of DEC’s later years is that key senior people, on both the technical and executive ends, were calling one another liars and were not trusting one another; each was totally convinced that he or she had the right answer for the future. At the same time, they blamed DEC’s demise on the “lack of any decision making at the top.” From my outsider’s perspective, and having seen decision making in academia that is anything but efficient in the short run (though it may be very effective in the long run), I cannot see what kind of decision making would have resolved the deep disagreements that existed among the senior technical and executive people in the late 1980s. Perhaps some powerful intellect such as Gordon Bell could have achieved some consensus, but wishing for such a person does not produce one. If a powerful executive had demanded more discipline and efficiency, many of these knowledge workers would have screamed foul and claimed that their culture was being destroyed.
If it is true that more and more companies are becoming complex networks of knowledge workers, and if those knowledge workers become representatives of groups that they lead, those companies will have major difficulties making efficient decisions and maintaining any kind of discipline in the implementation of decisions. What is possible among a small group of knowledge workers debating their different points of view to some clear decision that everyone buys into is less and less possible as each of these knowledge workers is successful, builds an empire under himself or herself, and enters the debate with a growing need to protect his or her own turf. The freedom that made the building of such a culture of innovation possible leads inevitably both to intergroup conflict and to a strong commitment to that culture.
Ultimately the culture and business entity of DEC could not both be preserved, raising the question of which was more valuable.
(p.254)
When all is said and done, the basic reason why DEC ended up where it did was that the evolution of the technology required transformations in the organization that the culture did not encourage or allow. The judgment of whether that was a desirable or undesirable outcome is a separate issue. Perhaps a culture such as DEC had should survive, as it has in its alumni. Is it worth changing such a culture to preserve the business entity? One aspect of the DEC legacy is to leave us with this tough question. What is, in the end, more valuable — a culture that is ennobling but economically unstable or a stable economic entity that changes its culture to whatever is needed to survive?
Ed: Very well put and an elegant summary. I would go further to suggest that tech change not only required transformation but meant that any transformed organisation would have been less pre-eminent and less economically successful because it would have been in a more competitive space with less advantages. Understanding that explains much of the deep reluctance to change — you hope your special position will continue.
Appendix A. Dec’s Technical Legacy
Ed: the focus on DEC’s innovation neglects the extent to which its success was arguably a product of economic circumstance and monopoly.
(p.269)
A slight comment on this section is that imitation is easy if that is your goal and you happen to have found a goldmine: get smart people together and give them freedom. (I’m being a bit harsh here — it is harder than this …)
The point is that, yes, DEC did all this stuff but partly because they did not care about profitability or were ahead of their time (e.g. Alpha). Just as Olson at the beginning was borrowing from state of the art stuff at MIT, so e.g. Intel would borrow from DEC in the 70s only to “own” them from the late 90s.
The point I’d make is that revering this technical legacy is a) part of the hagiography of innovation we have b) neglects the basic economics that underpinned this. It’s like looking at AT&T labs as extraordinary (and they were) but without seeing that it was directly made possible by the revenue from a regulated monopoly — it is the underlying economics that is worth remarking on, not so much the innovation (AT&T and DEC are similar to what a top-class university will produce … — perhaps less).
Put more bluntly: you can’t have your cake and eat it too. If DEC’s success is all this innovation and this came from a disregard for economic success (it was just at the beginning they happened to have found a goldmine) so their fall is attributable to the same sources …
Furthermore, one could argue that much of their success had little to do with their innovation — it was that they got a monopoly on a specific area. Just as microsoft were very successful without being very innovative (they were good and, importantly, fast at copying good stuff elsewhere and downsizing it to the massive new PC market).
Appendix B. DEC Manufacturing: Contributions Made and Lessons Learned, Michael Sonduck
How did manufacturing contribute to DEC’s culture?
(p.277)
Given this long, rich history, it is fair to ask what contribution manufacturing made to what we have described as the DEC culture. Interviews with many of those responsible for leading DEC manufacturing over this period as well as my own direct observation leads to several conclusions:
Although DEC experimented with a number of cutting-edge manufacturing concepts during its history, the company never made world-class contributions to manufacturing technology.
The foremost contribution of DEC manufacturing was to enable the products envisioned by its engineers to be reliably (in most cases) produced and delivered. DEC was never a low-cost producer. That was not a goal during most of its history. When it did become a goal during the final phase of its life cycle, it was met with only moderate success.
The manufacturing organization evolved over time in response to the company’s need for products and the underlying ethos of DEC. As an organization, it was often cumbersome, Balkanized, and inefficient when viewed from the larger corporate level.
The most significant contribution of DEC manufacturing was to the lives, learning, and development of the men and women who worked as part of it, and especially to those who attempted to lead it.
DEC manufacturing was designed for experimentation, not efficiency.
(p.279)
As long as customers wanted more and more, DEC manufacturing was bound and determined to build it and deliver it. It didn’t matter whether that translated into more plants than any other computer manufacturer had ever built; more square footage bought, planned, built, and opened per year than anyone else had ever done before; more résumés reviewed than anyone had ever done before; more air miles flown than any company had ever done before. These were necessities, not goals. They were what it took to get the job done.
In a more closely controlled, centrally planned and managed company these outcomes would have appeared to seasoned executives as ridiculous hyperbole. Kaufmann created a management culture that was anything but controlled, planned, and centrally managed. This was no accident. It was based on a belief, a fundamental belief that if you gathered smart people who were committed to the same thing you were committed to and gave them their head, even though mistakes would be made, the end result would be better than any other possible result.
There were significant positive and negative consequences — many of which would become harbingers of the ultimate decline of the company. Along with strong local control to ensure commitment came insularity to new ideas, especially when they were developed elsewhere. Along with the belief that brainpower would overcome brawn came the resulting overhiring (more is better). Along with experimentation came waste, which ultimately turned into excess cost. Along with a meritocracy came an oligarchy of smartness and “firstness,” which resulted in pushing aside, pushing out, and not listening to new ideas that came much later. Ultimately DEC manufacturing was an organization designed to experiment with ways to achieve what no other had ever done. Once that was accomplished, it was unable to refine itself into the finely tuned organization that was needed to become the low-cost producer.
Appendix D. Digital: The Strategic Failure, Peter DeLisi
DEC was in every segment of the market but failed to respond to the fact that it had ceased to be dominant in any of them.
(p.288)
As Digital’s lines of business proliferated, the company reached a point where it was confusing both its sales force and its customers about what it really was. I remember numerous discussions in those days with key customers who would ask, “What business is Digital in?” Unfortunately, Digital failed to take note of these repeated inputs, but most devastating by far, was its failure to note the erosion of its core identity as “the minicomputer company.” Unnoticed by Digital, the computer market had differentiated, and Digital was no longer dominant in any segment of this differentiated market. At the time, you couldn’t name one market segment in which Digital would be the first company to come to the minds of its customers. The company was in every segment you could imagine — mainframe, network, desktop, workstation, software, fault-tolerance, systems integration, management consulting, hardware repair, semiconductors — but sadly it was not dominant in any of them.
Appendix E. What Happened? A Postscript, Gordon Bell
Bell puts the failure of DEC down to incompetence on the part of leadership rather than a problem of culture, events or a lack of the money gene.
(p.293)
It really was simply ignorance and incompetence on the part of DEC’s top handful of leaders and, to some degree, its generally ineffective board of directors. Given the DEC culture of openness, honesty, letting the data decide, and taking personal responsibility, this straightforward explanation should suffice. The data clearly support the need to take individual responsibility for DEC’s problems rather than believing that it was the “events and the culture that made us do it.” (When former chairman and CEO Louis Gerstner arrived at IBM, the company was in the same relative position as when Olsen resigned from Digital; leaders can be responsible for the success or failure of a company.) These leaders lacked understanding of the nature of the computer industry in nearly every critical technology and product area:
- Moore’s Law. In 1989 Olsen demonstrated his lack of understanding that a \$300 CMOS NVAX microprocessor would equal and shortly exceed the \$300,000 ECL Aquarius performance. Figure E.2 from 1981 shows that ECL would have a short life when I had proposed the purchase of a part of Trilogy. My 1982 optimism was a costly mistake that required killing the project. Not building an ECL computer was a clear and easy decision when the technology failed to materialize in a timely fashion. The market rejection confirmed the decision. As Ed Schein shows in this book, Olsen loved having many options yet disliked killing projects implied by that — he was too much an engineer. (Although I refuse to believe that DEC lacked the money gene!…)