Ken
Olsen, President of DEC
by Richard Seltzer, from DECWORLD, the company
newspaper, September 1982
From modest beginnings in an old mill in
Maynard, Massachusetts, DEC
(Digital Equipment Corporation) has grown, in just 25 years, to be
the second largest manufacturer of computers in the world
(according to published surveys of the worldwide industry). Its
products -- from tiny microprocessors to large mainframe computers
-- have become models
of excellence in diverse markets and applications, helping
individuals as well as schools, governments, research institutions
and large and small companies of all kinds to perform their work
more efficiently and effectively.
To respond rapidly to market needs, DEC divides
its business into manageable pieces and delegates responsibilities
to many individuals, rather than concentrate them in the hands of
a few individual initiative, integrity arid accountability are
encouraged at all levels. This work environment means that the
talents, inspirations and efforts of many different people can
quickly be brought into play to meet the shifting challenges of
the highly competitive computer business.
In many ways, the most recent year was one of
the most challenging and satisfying in the company's history,
demonstrating the vigor and flexibility of this large organization
and the resourcefulness of its 67,000 employees.
"Our twenty-fifth year was a great one." says
Ken Olsen, President. "We grew 20%, invested heavily in new
products, were able to offset strategic price decreases by
reducing our costs in than ten years. That's not bad in the middle
of a worldwide recession. We have reason to be proud.
"We have suffered from having good times for
too long." he explains. "People, countries, economies and
companies cannot tolerate good times for very long. It's not
healthy.
"We had several years when things seemed too
easy. Our sales people had to spend most of their time telling
customers how long they'd have to wait to get our products. Demand
for our existing products made it impossible for us to develop new
products. Throughout the company, we developed bad habits.
"Two years ago I was frustrated. It took three
years to develop a new product. It took four months to get a
printed circuit board. We had committees on top of committees to
check everything.
"But now," says Ken, "after a lot of effort and
the recession, we are good. Now you can get a printed circuit
board in a week; and the new products move so fast, it's hard to
keep up with them.
"We did quite well with our old ways of doing
things," notes Ken. "Now the way we've started turning things
around, we'll be so efficient I don't worry about anybody in the
world."
Win
Hindle joined DEC in 1962 as assistant to the president. He was
promoted to product line manager in 1964 and to vice president
and group manager in 1967. He became vice president, Corporate
Operations in 1976. (Photo by Peg Blanchet, U.S. Area News)
"When I started at DEC in September of 1962,
the company had about about 400 active employees. It was growing
well/ That year we did sales of $8 or 9 million. Of course, at
that time we were a privately held company, and were not
disclosing our sales and earnings information to the outside
world.
"My first Job was assistant to Ken Olsen. He
gave me a variety of assignments. For instance, I started a formal
engineering scheduling system to ensure that engineering got done
on time, and I also recruited some engineers and sales people.
"We had a small Personnel Department, but we
had very definite ideas about how to handle people and how people
should manage. We had a strong feeling for the individual and
wanted to be sure our Personnel policies enabled us to provide
Jobs that people would be excited about and could accomplish, that
had goals and measurements, Many of the same things we talk about
today, we were Just as interested in then,
"We recruited mostly in the Massachusetts area.
I think at the time I Joined we had only one sales office and that
was in Los Angeles, where Ted Johnson was the manager. Then we
opened a second one in New Jersey, just outside of New York. A
year or so later, when we needed to open a sales office in
Chicago, I was sent there to hire someone to open and manage it.
We had a Works Committee that was the chief
policy-making body, as the Operations Committee is today. Ken was
the chairman and, as his assistant, I was the secretary and had to
prepare the agenda and write the minutes. In that committee we
decided that we were spending too much time on current issues and
not enough on long range planning. So we decided to get away from
our usual settings in the Mill and hold a meeting somewhere else.
Later we decided we should really get off and away, not Just go to
a motel. So we had a meeting at somebody's cottage in the woods.
That's where the name "Woods Meeting" came from -- a meeting away
from the plant where you consider longer range issues.
"In those meetings, we would plan the size of
the company over the next five years. As it turned out, when we
compared our actual performance to the long range plans, we found
that we had usually grown faster than anticipated five years
before."
PDP-6 and the product
line structure
"We did, however, face a very critical problem
around the time of the introduction of the PDP-6 in 1964. That
machine was larger and more complex than we should have attempted
considering how small the company was at the time . We had
difficulty finishing the hardware engineering and the software,
and all the best engineers in the company were recruited for that
project to try to make it work. While they were concentrating on
the PDP-6, where we were losing money, no one was working on our
other machines -- the PDP-5, PDP-7 and our modules -- that were
making money.
"It was at that point that Ken realized that we
had to change the organizational structure so that we wouldn't
arbitrarily put all of our resources into one product and to make
sure that we spread our resources in the same proportion as we
were seeing success.
’So it was out of that experience with the
PDP-6 that the product line structure in the company came about.
Each product line budgeted resources that nobody else could take
from them, and those resources had to be planned by the beginning
of each fiscal year, and the plans had to show how we were going
to make a profit.
"We could have gone out of business because of
the PDP-6 if we hadn't made those changes. Fortunately, we were in
good shape with our other products and they more than made up for
our losses on the P'DP-6, thanks to that change in our
organization that protected their resources.
"Most of the first product lines were oriented
around hardware products rather than markets or applications, and
they each had their own engineering departments. In other words,
instead of a central engineering department, we had PDP-4
engineering, PDP-5 engineering, PDP-6 engineering and Modules
engineering, along with a few smaller projects.
"We still had Sales, Manufacturing and Field
Service as separate functions. The product lines managed the
marketing, the engineering and the planning. Each of the product
lines asked for a certain amount of resources from Sales,
Manufacturing and Service and negotiated their needs with the
functional managers.
"Each product line started with Just the
resources it had at the time the new structure went into effect.
Because the PDP-6 had already taken away many of the engineers,
these groups had to rebuild, either by hiring these engineers back
from the PDP-6 or by recruiting new ones.
"You could propose adding resources, but we
didn't add very many during the first year. It took us a while to
get into the new organization.
"In 1965, Ken asked me to become the manager of
the PDP-6. I was already product line manager for three small
product lines and was asked to add the PDP-6 to what I already
had.
"After looking at the situation and seeing how
bleak it was, I recommended to the Works Committee and they agreed
that we fill current orders, take our losses and get out of that
business. When I told the PDP-6 group that, they pushed back,
saying, “We think we can build a new computer based on the PDP-6
which will be really super." So I went back to the Works Committee
with a proposal to start a new computer to be called the PDP-10.
After several attempts to get the project approved, we finally
succeeded, and that product became the basis of our very
successful DECsystem-10 and DECSYSTEM-20 families of 36-bit
computers.
Evolution
of the product lines
"Originally, the product lines taught the sales
force about each new product as it came along, and every sales
person sold every product of the company. Then in the late 60s,
some product lines began to feel that their products were getting
so complex that they needed sales specialists. This need was
particularly true with the large computer, the PDP-10. People felt
that it was so big that you couldn't expect a sales person to
understand it technically and also understand all the other
products of the company. So the PDP-10 product line made a big
push to have a specialized sales force concentrating in large
computers, and to a large degree that was accomplished.
"Around the same time, the PDP-8 group
developed a wide variety of applications for their product -- such
as for laboratories, factories, medical departments and schools.
So we began to train the sales force on these applications, At
first, every sales person had to learn about every one of those
applications. Then there started to be some specialization in the
sales force by market.
"Not long after the PDP-11 came out in 1970,
there was some competition between product groups. The PDP-8 group
was pressuring sales people to sell 8s, and the PDP-11 group was
pushing them to sell 11s. We became concerned that our sales
people were confused as to what they should sell to a given
customer. We felt that it was time to specialize our small
computer sales force by market area because small computers were
such an important and growing part of DEC's business and the
number and complexity of applications was also growing. So we kept
the Large Computer Group and the Modules Group as separate
product-oriented product lines, and we broke down our minicomputer
business, the 8s and 11s, into a series of market-oriented product
lines. Many of those — Laboratory, Education, Medical, Industrial
(which we now call Manufacturing) are still in existence today.
"At first the OEM (Original Equipment
Manufacturer) business was a part of the PDP-8 group, of which
Bill Long was the manager. When we divided the PDP-8 group by
markets, Bill became the OEM product line manager. That group
handled both 8s and 11s, as did the other market-oriented product
lines.
'These market-oriented product lines determined
the marketing thrust for the PDP-8 and the PDP-11 in their area,
and would then educate the sales force and the customers as to
which machines were appropriate for which applications.
"That change in the product line structure
raised the Question of what we should do about engineering. People
who were working on PDP-8 and PDP-11 engineering were brought
together into a central engineering organization, under Gordon
Bell, and the Large Computer Group continued to do its own
engineering as a more traditional product-oriented product line.
"I think the whole idea of managing the
business through our product lines was a stroke of genius. I don't
believe we could have ever grown as a company nearly as rapidly as
we did if we hadn't formed product lines, if we'd tried to manage
it as one whole business.
I believe you just can't manage a fast growing,
fast moving organization in detail from the top. It limits the
growth if you try to do it that way. So we've continuously tried
to push decision-making functions down inside the organizations to
product lines, to engineers.
"One of the concepts that hasn't changed from
the beginning of the company is that people are responsible for
the success of the projects they propose. 'He who proposes does.'
When they propose projects and get acceptance, then they go ahead
and carry them out and are Judged on the results. That fundamental
philosophy hasn't changed. I hope it never does.
"But the complexity of things has changed. It's
very easy for a company of our size to get so locked up in bureaucratic
decision-making processes that people feel like they go from
committee to committee making proposals and never getting
anywhere. So we have to keep working to make sure engineers feel
they can propose things and can, once they get acceptance, go out
and do them, that they aren't powerless, that they can get
decisions made.
"One of the thi9ngs you do as a company grows
is try to beat down the hierarchy and the red tape so people can
get their jobs done easily. A lot of what we do, certainly a lot
of what Ken Olsen does, is to tear away the red tape and allow
people to do their jobs. We spend a lot of time trying to make it
fun to work here, make it challenging, make you feel as though you
can make important contributions.
"As for the future, I'd like to see a company
where each individual really feels that he or she has a role to
play and has the freedom to succeed or fail based
on their own ingenuity. One of the horrors of
modern society is 'group-think' or 'group-do,' where you're never
singled out as an individual and don't have an opportunity to show
what you can do all by yourself, based on sour own drive and
ingenuity. I hope 25 years from now we will have a company where
individuals feel challenged and feel that they can really live up to their full
potential.
"My vision of a beautiful company is one where
individuals when they go home at night feel that they really made
an impact, that they've been able to accomplish something, and
they feel proud of themselves and proud of the company they work
for."
Jack
Smith came to DEC in 1958 as a technician. He was involved in
building and testing the company's first logic modules and its
first computer, the PDP-1. Later he was responsible for the
development and growth of the Systems Manufacturing operation.
He was promoted to vice president, Systems Manufacturing, in
1976 and to vice president, manufacturing, in 1977. In 1982 he
took on additional responsibilities as associate head of
Engineering (Photo by Peg Blanchet, U.S. Area News)
"In 1958 when I started at DEC there were only
12 people in the company. I was hired as a technician. Sometimes I
did engineering work. sometimes manufacturing. I also spent a lot
of time painting system modules and sweeping the floor. There
wasn't any formal structure. You Just pitched in and did whatever
had to be done.
"As we grew to around a hundred people, about
I960, we started organizing departments.
"At first we produced one narrow product line —
systems modules, and served one market — the engineering lab area.
So our Manufacturing Department was small and uncomplicated.
Somebody supervised the people putting the product together, and
somebody else was in charge of buying material, and all the
manufacturing disciplines reported to the same manager.
"Then with our first computer, the PDP-1,
manufacturing became much more complex. We had several products
that we were selling to different markets.
"The first few PDP-ls were handcrafted. We had
non-technical people stuffing the modules, but then technicians
and engineers had to take those modules and the frames and plug
them together and wire them, one by one.
"At that time, there was a strong belief in the
industry that you had to understand how a computer operates to put
one were going to get our costs low enough to be successful at
this business, we couldn't afford to have highly paid technicians
and engineers doing assembly work.
"The most complex part of assembling a computer
was stringing the wires. We reasoned that to do a good job of
that, you just have to know which wire goes from where to where.
So we hired non-technical people with basic manufacturing skills.
"At first we had no automated equipment at all.
We simply gave a person a wiring diagram that said, "Take the wire
from point A and bring it over to point Z.“ The person who strung
the wires also had to solder them. Then someone else would go
through and check the work against the diagram, wire by wire, with
over five thousand wires in each computer.
"As it turned out, technical people were not
needed for that kind of work and, in fact, did a far worse job of
it than people with physical dexterity and the ability to pay
attention to detail and concentrate on repetitive tasks.
"Since the product was successful, we had to
hire lots of people to string wires. That also meant we had to
manage and supervise them.
I didn't have any significant role in the
company up until the PDP-l. Then my Job was to supervise the
people who put the machines together and to develop ways to make
sure that they went together right.
"Although my title changed a number of times
and I had various additional responsibilities over the years, up
until I took over as vice president of Manufacturing in 1977, I
was the person responsible for putting together all of the
company's computers, a function that came to be known as ‘systems
manufacturing'.
"We made about forty PDP-1s, which for our
size, was a lot of business. Then we went to the PDP-4, PDP-5,
PDP-6 and PDP-7, still making computers one at a time -- wiring
things together and connecting them.
"The PDP-8, introduced in 1965, was our first
volume computer. We were going to make thousands of them. So to
get the costs low enough to sell them for a low price and open up
new markets, we set up assembly lines. Instead of having three to
five people working on a single computer from start to finish, we
divided the job into a series of separate tasks that different
people could work on at the same time. A computer went through
various stages of assembly, and came out the other end complete.
"With the success of the PDP-8, growth became
the most important issue in manufacturing -- being able to expand our
capacity fast enough to keep up with demand. To speed up growth,
we formed manufacturing groups. Instead of having a single
manufacturing group with a single manager, there was a group
dealing with tape units and another group dealing with central
processors and another group building power supplies.
"To manage all those separate groups, when
there were about 300-1000 people in Manufacturing, we adopted what
is known as 'matrix management.' In other words, we built a strong
central staff, with high levels of expertise in all the various
manufacturing management specialties, such as materials,
production and inventory. These experts were available to help the
individual groups manage their business. For instance, the
materials manager was expected to oversee all the material
functions in all of the individual manufacturing groups. Even
though the materials manager didn't directly supervise the
materials managers in the various operations, that person (the
'matrix manager') was expected to know what was going on in each
one of groups and feel responsible for it.
"We set up this way because we had to grow very
rapidly, and there simply were not enough qualified people, and we
couldn't give people experience fast enough for each group to have
its own set of experts.
"While most manufacturing operations were in
the Mill in Maynard, a matrix manager could easily walk from floor
to floor and talk to people and get things done. Also, back in the
1960s and the early 1970s, the products were mostly based on
modules and central processors, so the different manufacturing
groups faced very much the same kinds of problems and it made
sense to have a strong central staff.
Plants
throughout the world
"In the late 1960s and early 1970s,
manufacturing spread throughout Massachusetts and then the world.
We opened plants in new areas because we couldn't hire enough
people in Maynard.
"When locking for another site in
Massachusetts, we first chose Westminster because we wanted to be
as close to Maynard as possible and still tap a significant new
labor supply. Westfield was the next step beyond that, and so on.
"When we started a new plant, we would send out
a team -- typically about five people -- from one of our existing
plants. Everyone else was hired locally.
"Often the plan was to get 50-75 people in the
first year, and by the end of the second year to have 400 people,
all used to the DEC way of doing things. After that was done, then
they could worry about questions of further growth. That was the
pattern not only in the U.S., but also in Canada, Puerto Rico and
Europe.
"But as we spread out geographically,
communication with matrix managers back in Maynard became
increasingly difficult. Also, as we got into new businesses, like
printers and terminals and disks, manufacturing operations became
diverse, requiring different kinds of expertise. At the same time
the need for increased volume intensified. Gradually, central
control became more of a hindrance than a help.
"A matrix organization gives you the ability to
grow rapidly with a limited number of experienced managers, but it
also tends to make responsibility fuzzy and to make decision paths
very long. Who is really responsible? Is it the line manager or
the matrix manager?
“By 1977 matrix management had outlived its
usefulness; so we reorganized Manufacturing into standalone groups
that were responsible for their own products.
"Up
until then we had one group producing all the modules for
everyone, another producing power supplies for everyone. We
divided and shifted those resources so, for instance, Storage
Systems could make its own modules and its own power supplies, do
its own assembly work, its own testing, its own systems work. It
took a few years to arrive at the point that each group had under
its own control the capacity it needed to get its job done.
"Meanwhile, during the rapid growth of the
1960s and 1970s, each of the company's organizations, such as
Engineering and Manufacturing were structured and chartered to
operate independently. There was very little interaction between
Manufacturing and the rest of the company when it came to choosing
plant sites or making other decisions, and there was no formal
process for deciding how much of what product we should build.
That was a good thing when the main goal was to grow as fast as
possible. But now there is more benefit to be gained from working
efficiently and doing what we need to do to win now. So the
emphasis has shifted toward close cooperation between the various
parts of the company.
"For the future, I believe that we have to
continue to reinforce our strong standalone manufacturing groups.
They have to build their own capacities to the point that they
feel they control and are responsible for their own destiny. We're
still shifting capacity to help reach that goal.
"Our basic strength has always been the
attitude and commitment of our people. I think the most important
thing Manufacturing can do is to continue to provide challenging
opportunities for personal and professional growth while we
reinforce our commitment to achieve manufacturing leadership in
our industry."
Jack
Shields joined DEC in 1961 and was named manager of Field
Service in 1964. He was promoted to vice president, Field
Service and Training in 1974 and vice president, Sales, Services
and International in 1981. (Photo by Peg Blanchet U.S. Area
News)
’If you think about what's happened in computer
technology in the last 25 years, it's like future shock. Each year
more and more events take place which bring us further into the
technological future. If you think of the millions of years man
has been on earth, 25 is so short a time it's almost not worth
talking about. But in that time incredible progress has been made
in the field of computers that has had far-reaching effects on
many aspects of life — from education and entertainment to
medicine and space travel. A lot of this has been captured in the
Digital Computer Museum. It's been exciting to live in the time of
these developments as well as participate in some of them.
"Twenty-one years ago when I started at DEC, I
was twenty-one years old. The company was young too; it had only
shipped two computer systems. There was a PDP-1 at Bolt, Beranek
and Newman and another at the Itek Corporation. Then in August or
September, 1961, we delivered the third to MIT.
"I worked in Engineering, but back then, with
less than a hundred employees in the whole company, the few dozen
of us in Engineering did a bit of everything. Maybe you'd work on
design one day and test the next, and then you might be called on
to do installation or servicing. People in Manufacturing helped
out in Engineering and vice versa. It was a closely knit group,
and everybody wore a lot of hats.
In fact, the only functional differentiation was based on
who you worked for.
"As it turned out, I and a few others ended up handling more service
calls than the rest of the people did. After a while our names
naturally came up when customers called needing help.
"So when someone was hired to set up a training
and field service facility, I was asked to So on loan for three
months to set up the service organization. At the end of the three
months, I was asked to stay in the Job. So Bill Newell and I
started and developed the DEC Field Service organization.
"We installed and maintained the equipment,
designed and built our own logistics system, and developed
techniques for handling calls. From time to time, we would have
our trainees work in Manufacturing so they could get on-the-Job
training while helping to test and build the products and get them
shipped on time.
Jack Smith used to run a piece of
Manufacturing. He and I sometimes had to get together to work out
shipment schedules because the service organization couldn't
handle all the products he could build and ship, or because he
needed some help from my people to get more systems tested. Once
the systems were tested, manufacturing people would go out with us
in the field and help install them. Then, having given the
manufacturing people a taste of the field life, it was easier to
recruit some of them. All in all, the reciprocity was great, and
our size made it much easier to work together on common company
goals.
"The organization became international around
1963. Ted Johnson transferred from California to Germany, and our
first Canadian sales office opened that year. I remember having
booth duty with Ted at a trade show in Basle, Switzerland, showing
the PDP-4. That was my first business trip to Europe.
"We were primarily selling modules and logic
kits, and then we started to sell our computers and memory
testers. Customers were more sophisticated then than they are
today, but they still required support.
"Originally, we hired local people for service
in Europe. Then we transferred Ken Senior to run our European
Field Service organization out of the United Kingdom. As I recall,
in 1964 we had about 11 field service people in Europe.
“Today we have nearly 20,000 people in Customer
Services worldwide, and everything is highly specialized to reduce
cost per repair arid increase productivity. We have specialists
for call handling. We do a lot of remote diagnosis. But in the
early days of DEC, the technician in the field had to be a
generalist, able to handle all aspects of diagnosis arid repair of
hardware and software. Many times you didn't have all the modules
with you, but you did have the components, so you would
trouble-shoot and repair down to capacitor and transistor level.
(Today those components are buried -- tens of thousands of them on
a single integrated circuit --
and you can't get at them.)
"Back then you had relatively few diagnostics
to work with, and they were relatively unsophisticated. Generally,
it was a customer's program that failed, and you had to figure out
what that program was doing and trace the problem from there. A
lot of the programming was done in machine language rather than in
higher level languages, and the programs you would use to exercise
the product were written by the customers. Our techniques for
checking the circuitry were marginal. We would vary the voltages
applied to the various gates to try to induce or reproduce the
failure. Also we had to cope with a lot of design problems that
couldn't be eliminated because testing was nowhere near as
sophisticated as today.
"To me the tremendous growth of DEC over the
last 25 years presented experiences that I don't think many people
will ever have again. You needed technical skills and had to keep
those finely honed, and at the same time you had to build an
organization and therefore develop the people and organizational
skills which were required, and then add to that the business
skills and the ability to create a new function and a new way of
providing a service. I'm extremely grateful for the opportunity to
face those challenges and to grow. Personally, it's been an
incredibly rewarding and rich experience.
"Quietly, without a lot of fanfare, DEC changed
the way companies view service. We took an activity that companies
had always thought of as a nuisance and a problem, a necessary
evil, and we made it into a profitable business. We started
showing a profit way back in the early 60s, and over the years we
were able not only to provide high quality service, but also to
develop new techniques which allowed us to become more productive
and cost effective and pass those savings on to our customers. We
created a new way of approaching service that today the rest of
the computer industry is trying to emulate.
We have done similar things in Educational
Services and Software Services, in parallel with the company
growing, the technology changing, and our people maturing.
"To me the most remarkable thing is the
stability of the services organization over the years. I recently
looked at a memo that was written in 1964. It listed the 35 or so
people who were in Field Service at that time. There are 20 of
them who are still with DEC today. I think that speaks for the
stability, the commitment and the performance of the organization.
"From a business performance point of view,
it's been an unprecedented success story. We continue to meet or
exceed our fiscal goals for the twentieth successive year, and in
the Just completed annual customer opinion survey -- the real test
of our performance -- the improvements were dramatic. On a scale
of 1 to 10, we're well above 8 on average. No other company that I
am aware of can point to that kind of balanced performance.
"The future looks even more promising. The
Software Services organization is a tremendous asset that will not
only provide a source of revenue but a tremendous competitive
advantage for DEC well into the 90s. Educational Services has
tremendous areas of untapped potential. For instance, the latest
computer-interactive video disk technology that they developed
with our Small Systems Group (IMIS) and the courseware that goes
with it will put us in a leadership position in computer-aided
instruction. The Field Service organization will continue to
innovate, continue to find new methods, new services and new
sources of revenue. I'm betting that they'll continue to be the
best in the industry, and they'll still grow at an unprecedented
rate."
seltzer@seltzerbooks.com
privacy
statement