| [copied w/o permission]
The Rise and Fall of Digital
A 2004 Harvard Review Business Study
by
George Van Treeck
FORWARD:
I have never seen any company with so many people that are warm, honest,
dedicated to excellence, and hard working as at Digital. In the past, Digital
has used this great asset of wonderful people to overcome what seemed at the
time to be huge problems, while some of our competitors have a fallen aside into
insignificance.
There probably isn't anyone in this company that is not aware that there are
problems causing the current slowing of our business. And I'm confident that our
people can overcome the current problems and attain better than 20% annual
growth in the future.
However, being able to overcome the problems first requires knowing true
problems. Of course, our first response is to look at business issues like the
economy, net profit margins, organizational efficiency, etc. as the causes of
the problems. I don't believe that this company is yet aware of the true
problems that we need to solve.
Business people can make charts of past business, current business and make
projections into the future. I'm going to chart Digital's past, present, and
project a future based on something other than financial metrics. Like all
projections, it simply shows the likely future if nothing is done to change
business practice.
My hope is that this not be taken as executives and management neglegence or
incompetence. The purpose of this paper is to increase awareness of other things
that might be root causes of our problems. I'm certain that once management and
executives are aware of these things, they'll take appropriate remedial actions.
The Rise and Fall of Digital
A 2004 Harvard Review Business Study
The PROPHET:
The prophet has a dream that he or she thinks others can share. Ken Olsen and
his had a vision building their dream computer and thought that other engineers
would also like to have that dream computer. Their product was innovative and
demand was large. It was the beginning of the highly successful company,
Digital.
The company was founded on a vision. The vision was based on the intuition of
people thoroughly familiar with the technology and an good understanding of the
target customer. The founders of Digital were engineers thoroughly familiar
with the design of computers and understood the computer needs of the engineer.
There was a "consensus" of business people and venture capitalists that one
could *NOT* make money on such a venture. The founders moved forward without
that consensus to grow a multi-billion dollar company.
In the prophet stage, there is no clear distinction in levels of management,
little distinction between engineering and manufacturing, service was done by
people with engineering backgrounds, little distinction between marketing and
sales. Sales people were highly technical application engineers who showed the
customer how to use their dream products. Everybody worked together as a team.
All the 'marketing' messages and literature were highly technical manuals.
The BARBARIAN:
This is the next stage of business, where a successful company gains market
share from the competition. Most prophets that start new companies are also
conquering barbarians. The barbarian takes up the dream and drives to deliver
that dream to all territories. The barbarian takes no prisoners. The barbarian
wants it all.
The founders of Digital weren't barbarians. They didn't need to be. They were
nice guys fresh out of college. IBM had not entered the minicomputer market.
There was a large vacuum that Digital was gladly filling with their PDP line of
minicomputers. Digital didn't stay in this phase long because the competition
was small compared to market demand which was bigger than all the vendors could
supply. Digital grew very fast and quickly entering the next stage.
The BUILDER & EXPLORER:
The builder comes from the production side of the company. This person
understands the complete production process. The builder understands what is
required to make a product, get product out the door and provide support. The
builder knows how things should be done will tend to simply command how things
will be done. When the builder moves into management, he has difficulty
transitioning to the coordination and consensus process.
The explorer comes from the sales side of the organization. The explorer
achieves his goals through persuasion and building personal relationships. The
explorer thinks that business is all about getting the customer to buy. The
explorer wants purchase orders per second. If you work for an explorer, then you
better be out on the road meeting new customers and conquering new territories.
The builder and explorer have no tolerance for administrative systems that seem
to get in the way. Most builders and explorers have a lot barbarian in them.
This is also the stage where specialization occurs. It is during the builder/
explorer stage that design engineering becomes a separate organization from
manufacturing. Marketing is a separate function from sales. There are now
distinct levels of management.
The ADMINISTRATOR:
By 1980, Digital was a billion dollar company and still growing very rapidly. At
this stage the CEO becomes very concerned about wasteful duplication of
resources, internal competition, and control of how money is spent. "Control"
and "by the numbers" are the key descriptions of the administrative stage.
It was during this administrative stage that Digital started to focus on the
"process". The process provides consistent quality. There was a focus on
engineering design processes, statistical quality control processes, just in
time inventory process, MRP (material requirements planning) process, sales
training process, sales tracking processes, field service processes, etc. There
was even a management process, matrix management, a process to reduce redundant
organizations and achieve consensus.
Focusing on process was very important. Data General, Prime, Wang, Honeywell,
and others, all had minicomputer products. And having processes in place that
maximized efficiency and quality were key competitive advantages.
By the late 1980s, Digital had refined many processes, e.g., moved from MRP
to MRP II, from statistical quality control to zero defect manufacturing, moved
to a 5 phase engineering process, strategic task forces, better financial
tracking systems, etc.. When Digital finally started shipping follow-on VAXs to
the first VAX, revenues were strong and growing very quickly. On top of strong
revenues and growth, Digital was making a large effort to reduce the cost of
doing business and increase net margins.
Strong revenues, high growth, and an active campaign to control costs were music
to Wall Street's ears. In 1987, Digital stock, with a fair market value of $86
per share, soared to a record $199 per share on the New York stock exchange.
This seemed to justify current corporate philosophies and directions.
It never occurred to Digital's administrators in 1987, that their success was
due to momentum created by the earlier builder/explorer culture -- not by any of
new strategies of the 1980s. By 1989, Digital's momentum was stalling. And
Digital was well into the next stage of the corporate life cycle. The seeds of
Digital's decline started in the early 1980s. But the administrators claimed the
problem was a slowing computer market.
By the end of the 1980s: The PC market was now larger than either the mainframe
or minicomputer markets and still growing strong. Digital was late entering and
never made a sincere effort to capture market share of the single largest
segment of the computer market! The workstation market was a multi-billion
dollar market that was growing very fast. Late entry resulted in workstation
revenues being a much smaller contributor to Digital's overall revenues than it
could have been. Fault tolerant systems sales were growing extremely fast, but
Digital didn't have a product for that market. The most profitable and high
growth segment of computer business was becoming the software, yet Digital's
business model was still one of software as a loss leader to leverage hardware
sales. As a consequence Digital failed to enter the high growth, high profit
software business on multiple vendors' computers.
It's ironic, that with all this high growth, the administrators overlooked those
numbers and blamed a slow computer market. Yet, this is what one should expect
of an administrator. The administrator looks at financial reports of the
company's existing business instead understanding the end user -- the people who
uses computer technology every working day. The administrator calculates past
revenues, market share, and makes projections. The administrator sees only what
existed or exists. There is no vision. The administrator thinks Digital's
problems can be fixed with better processes and tracking systems.
There were two cultural problems from the administrative stage that started the
beginning of Digital's decline: 1) The administrative culture that focused on
numbers. 2) The administrative culture that focused too heavily on how to do
things -- the process.
1) The administrators had to wait for the "numbers" to prove the markets existed
or numbers that proved they were losing business before taking action. Numbers
were the only means of achieving a consensus that action was important. By the
time, Digital could respond to the "numbers" with products, it was too late. It
missed the window of opportunity. An ocean surfer uses intuition to catch the
wave at the right time. Digital waited for the numbers to prove there was a
wave. By then, it was being crushed by the wave.
2) Digital was so focused on the process -- "how" to do things, that little
time and resources were spent on "what" new things should be done. The
focus was on how to make our minicomputers even better, instead of re-evaluating
who the customer was, what the customer needed, and what innovative alternatives
might satisfy those customers. Focusing on the "how" with little attention
to "who" and "what", resulted in an inward looking company culture.
Digital's culture was an inward looking culture of engineers building the
dream minicomputers for the engineer. But the majority of the computer market,
even Digital's installed base, were no longer engineers. The non-engineer
dreamed of PCs and fault-tolerant computers, etc.. And even the engineers no
longer dreamed of minicomputers. They wanted workstations. Digital didn't look
outward at end-user (customer) needs and trends until the market numbers or
lost sales proved that action was necessary.
Waiting for the numbers to prove a market existed and too busy thinking about
"how" resulted in late entry into important segments of the computer market. And
even when it did enter those markets, the attempts were half-hearted. For
example, engineers dream of workstations with megapixel displays, megaMIPS,
megaFLOPS, etc. -- not PCs. And administrators didn't complain because PCs are
low margin business ("bad" numbers) with stiff competition. Thus, PCs were never
a major part of Digital's business even though that market was now bigger than
either the mainframe or minicomputer business and growing faster than Digital's
business.
In the 1990s, Digital eventually failed completely in the workstation market,
because it failed to perceive in the 1980s the need for the lowest-end
workstation as it's first and primary focus for the workstation market. This
wasn't the ECAD or MCAD engineer's dream computer. Administrators didn't
complain because of the low margin "numbers" and stiff competition in the
low-end workstation business. And the numbers didn't exist to prove the market
was important -- until it was too late.
Other vendors had captured so much market share with low-end PC-like
workstations, that they created defacto standards for big endian data formats
and defacto binary compatibility. This in turn impacted the chosen computer
architectures and vendors for the high-end engineering workstations. Digital's
high-end workstations were neither data compatable nor binary compatable with
dominant desktop computers. Digital was eventually locked out of the market with
what were perceived as incompatible and closed computers.
Likewise, Digital's software development tools were designed by engineers for
engineers. The engineers built ever more powerful tools to help the engineer
develop software. However, demand for application software outpaced what the
engineers with those tools could produce. Digital failed to recognize the market
trend away from engineers developing the major portion of software to masses of
end users developing their own software as the means of keeping up with demand.
There was never a major focus on code generators and application generators for
the mass market. And in particular, because most end users had PCs, this made it
even more difficult for Digital's engineers to perceive the need for end user
development tools. And of course, the administrators didn't complain because
there weren't any numbers to prove the market would ever be large.
When staff complained about the lack of innovation. Digital executives pointed
to all the research and advanced development they were funding. Unfortunately,
there was a lack of numbers to prove the market would be large for the dream
products of the researchers. And unfortunately, the administrators were focused
on how to make the current products better and didn't have time to understand
who the customer was and how to apply that research for those customers. Thus,
the research never turned into a major part of Digital's business -- until some
start-up companies generated the numbers to prove the market existed.
Digital was in many businesses but organized and managed as one large business.
Each business had it's own vision of how to be the best. But this was often at
cross-purposes to other businesses. It became increasingly difficult to gain a
consensus on major issues. The focus on the process (matrix management) of
eliminating redundant resources by having resources centralized and managed as a
single company was not appropriate for a large company in many markets.
Also during the administrative age, the sales division had become unbalanced.
The senior vice president of sales was an explorer. He wanted the sales force
out with customer using persuasion and relationship building rather than
focusing on technical details. He was always "committed" to training the sales
force. But sales force was always on some urgent sales drive, where there was no
time to get the training.
This lopsided focus was not counter-balanced by a strong builder personality in
the field who would ensure adequate training of the sales force. The result was
a sales force that could buy dinners and talk about the industry needs to the to
a vice president, but was unable to articulate to end users why our products
were better than the competition's products and often were not even aware of
which products would meet their needs.
A half-hearted attempt to fix the problem was the creation of sales support
staff to compensate for lack of technical expertise of the sales reps. But sales
support staff increased the cost of each sale. The cost of sales ("bad" numbers)
resulted in administrators never staffing sales support to adequately address
the sales opportunities.
Digital could have relied on its once strong VAR (value added reseller)
channels. The VARs had the application expertise to sell our products. But
the VAR sales reduced the sales "numbers" for the direct sales administrators.
Thus, Digital's relationship with VARs deteriorated and never recovered.
The BUREAUCRAT:
Around 1986, Digital started phasing into the bureaucratic stage. By 1990,
Digital was fully into the bureaucratic stage. This compounded the problems
inherited from the administrative stage.
In Digital's bureaucratic stage there were several layers of vice presidents and
several layers of management. Senior management and executives were so far
removed from the company's technology and customers, that staff had to create
sound bites and slick slogans to get management and executive attention.
Oversimplified messages percolated up through layers of managers and executives
resulting in poor decisions. Each new slogan changed the corporate focus, caused
re-organizations, new programs, new policies, etc..
People at the top felt frustrated by the quagmire of processes, systems and
organizations that seem to resist their new direction. The people at the bottom
created the quagmire to resist what they felt were ignorant and vacillating
decisions from on high. Everyone began to doubt they could have a positive
impact on the company's business.
The bureaucrats were the ones that spent most of their time in meetings getting
reports from subordinates on what happened or should have happened. They were
the ones that hadn't been directly involved in development of a new product or
service in many years. They thought that tighter controls and re-organizations
would solve their problems. They spent more time with other managers than they
did with sales reps and production people.
The bureaucrats in administration and finance were more concerned with
perceptions on Wall Street than with product innovation and better customer
service. Rather than admit lack of product innovation, administration and
finance reported a slow computer market. They said more cost control was
required. They said the company needed to trim people. It went against the
conventional wisdom of bean counters that the "excess" people should be put into
new, independent business ventures, e.g., a software business or PC business or
small business networking. That would have meant less "control" and the overhead
("bad" numbers) of redundant organizations like separate marketing and sales for
each new venture.
In contrast, venture capitalists like the ones that funded Digital expect only
about one out of four to six ventures to succeed. The profits from that one
successful venture more than compensates for the failed ventures. Digital had
the capital, but failed to fund small, autonomous ventures with significant risk
in hope of that one innovation that would create a major revolution in way
people work. The bureaucrats wouldn't even consider such risks -- and certainly
not worth considering without the numbers to prove the market will exist.
Also, Digital had no appreciation of role of personalities in a company. A
president must be a barbarian. The vice president of engineering must be a
prophet (like Gordon Bell in the 1970s and 80s); vice president of manufacturing
must be a builder; vice president of sales must be a builder as well an
explorer; vice president of marketing must be a barbarian; vice president of
finance must be an administrator.
In the bureaucratic era, the words from Digital's lips were the "customer is
first" and "industry leading". But the real focus was internal. Bureaucrats
perceived the path to success to be elimination of internal discord and
increasing cooperation and efficiency -- rather than looking outward for new
technology, new ideas, and understanding the end user's needs. The perceived fix
was re-organization and trimming of fat. Re-organize to shift responsibility and
control. Trim fat to increase the net margins.
There was no serious effort at innovation to spur growth. That is, there were no
prophets allowed to pursue their intuitive visions as semi-autonomous ventures
without consensus or numbers to support the vision.
By 1990, re-organizations were frequent. Employees were increasingly ordered and
instructed. There was less initiative and reliance on people at the bottom.
People at the bottom were allowed to have responsibility, but not allowed to
make decisions. By 1993, the working staff became unimaginative soldiers
following management's orders.
Marketing executives ordered the creation of megaevents for sales training and
marketing. They didn't trust the competence of staff to define their own
marketing strategies and tactics for their specific markets. Thus, application
and industry specific tradeshow support and technical training of the field
targeting end users of the technology declined in favor of DECtops, DECtechs,
DECworlds, etc. targeting the "decision maker".
In Digital, bureaucrats are the decision makers. And they perceive people like
themselves to be the important customer to target. Yet, most executives allocate
funds for a purchase, they leave the decision of which vendor is selected to the
people closest to the problem. Digital's sales people were unable to communicate
effectively with those closest to the problem -- the end user.
Even in Digital's inward looking culture from the administrative stage, there
was some vague awareness of who the customers were (end users). The
bureaucratic culture was now completely blind to the end user. They were no more
important to Digital than it's own lowest level, "techie" staff.
Bureaucrats always think they are the indispensable people with the knowledge
and vision to guide the company. Of course, none of them use the products they
manufacture and sell themselves. Very few had PCs or workstations on their
desks. None did their own system management so that they could understand what
our customers experience? None were using the competition's products. None were
involved in the research to develop radically new technology that will obsolesce
the current technology. The closest they came to contact with a Digital product
is reading their mail on a terminal. Yet, they thought they had the vision of
what a customer needs and what new innovations might best address those needs.
One such "vision" of these bureaucrats was "enterprise integration". Digital
thought it was in the integration business. Digital thought the integration
business was high profit consulting services leveraging large volumes of systems
sales. The customer thought the best service was products that didn't require
any service. The customer thought the best integration service was vendors
developing low cost, interchangeable components, like hifi equipment, that
didn't require consulting services. Digital moved in one direction, while the
customer was moving another direction.
By 1995, Digital was experiencing extreme decline. The Boston Globe questioned
whether Digital would even exist in 1997. By this time, Ken Olsen had retired as
CEO and president. A bureaucrat from finance was brought in to turn the company
around. Bureaucrats from finance are the ones most often brought in to turn a
company around because they're the ones that put into numbers exactly what they
want to achieve. And numbers are what the board and large investors require.
The pace of reorganizations was even more fast and furious than the six years
prior to 1995. You can always tell the companies in the bureaucratic stage:
there are two or more reorganizations in three years, decisions on how to
reorganize come from the top, and there's a continual 'realignment' to fine tune
the control and focus.
Digital's bureaucrats didn't understand that a maturing market was quickly
obsolescing the high margin business. There were many 100 MIPS
PC/workstations networked without any connections to low volume, high profit
servers. The high-end minicomputer market slowed. Digital's high-end was
under pressure from mainframe and supercomputer vendors.
It's during this tough time of 1995, that the barbarian bureaucrat rose in the
ranks. These are people that focus on short term gain at the expense of long
term health. They rise in the ranks by showing the great cost savings and
increased profitability they've created. The barbarian bureaucrats decided to
focus on Digital's central business and sell off the other businesses. It was
decided the central business was (you guessed it) high-end, high-margin
computers. A couple of the sold off businesses were taken over by barbarians
that fired most of the executives and management and grew into major vendors in
their market niches.
The ARISTOCRAT:
By 1999, Digital was now a $2 billion dollar company selling high-end server
technology in a mature market. Control was centralized. The leadership was more
interested in their own comfort than in Digital's competitiveness. Digital
now had a posh central headquarters building for the top executives. Resources
were diverted from research and development to expensive offices, social
gatherings and travel.
The aristocrats spent most of their time on financial matters, strategic
planning and re-organizing. They spent no time with the people who design,
manufacture, service or sell.
In 2003, Digital went into bankruptcy. Digital never came out bankruptcy.
This was a 2004 Harvard business case study. It was a study of imbalance and
ignorance about people issues: Strong focus on how things were done with little
focus on what things were done; business strategies based on numbers with no use
of intuition of the technology developers and end users; strong administrative
control with no tolerance of the chaos, intuition and risk associated with the
independent ventures from which innovation is born; ownership without decision
making capability nor adequate resources (matrix management); strong emphasis on
selling via persuasion and high level relationship building with no emphasis on
product details and end user relationships; poor understanding of what business
it was in; poor understanding of the end user; no understanding of corporate
life cycle cultures and the conscious effort required to keep the early the
culture alive; no understanding of the importance of personality types required
for each job and how to balance those personalities.
|