Conference 7.286::digital

                      DIGITAL Internal Use Only

Bill Strecker's September 2, 1993 "Cafeteria Talk" at the Mill

Introduction by Ralph Christensen:

Good morning. My name is Ralph Christensen. I am the Engineering Human
Resource Manager, and I welcome you to to this morning's Engineering
Communication Forum. This is one of a series of five similar sessions
over the course of the next week and a half that will be presented
across Engineering. We are going to be taping this particular session
to make sure that people who aren't able to attend one of the live
forums will be able to have access to the communication. 

The purpose of these sessions is to have good open communications
throughout the Engineering community about the directions of the
corporation, the directions of Engineering as an organization and to
provide an informal Q&A and information exchange with Bill Strecker. The
format for today will be Bill's comments, followed by as much time as we
need for Q&A with the group. 

It is my privilege to introduce Bill Strecker, the Corporation's Chief
Technology Officer and Vice President of Engineering. 

Bill Strecker:

Thanks, Ralph. Ralph described my talk here as comments. The number of
slides I have probably goes somewhat beyond the definition of comments
-- I brought a fair amount of material along. I hope we will get through
this in perhaps the first hour or so, and that will give us a second
hour for Q&A. 

We have made so many changes in the corporation over the last eleven
months that there are an awful lot of things to talk about, and it is in
some sense hard to prune it down. I think there must be a lot of 
interest in this community to get an overview of all those changes, so
that is what I am going to try to do for you. 

First I would like to talk a bit about the transformation of Digital. The
transformation of Digital began in October of 1992 when Bob Palmer took
over as CEO of the Corporation, and I would like to give you a bit of an
overview of some of the things I believe Bob was thinking about when he
took over the organization, and some of the goals that he had for the
short term. 

I would then like to talk a bit about our Corporate Strategy and
organization. I think you know that I am going to tell you the strategy
is still in the making, but want to tell you where it is right now, and
a bit about the organization that we put in place to implement that
strategy. 

I would then like to move on to the engineering organization. As you
know over the past eleven months we have made a lot of changes in
engineering. A lot of them are very positive. Some of them obviously
have had some negative overtones to them. Significant downsizing has
been a very painful experience for all of us. But hopefully we have
emerged with an engineering organization and structure which will enable
us to do the job that we need to do. 

I would then like to move on to the area of product strategy to give you
some background around some of the processes that we used in the Spring
to think about product strategy (which like our Corporate Strategy is
still in the making). In fact, I think a fact of today's existence in
the information technology industry is that strategy is always in the
making. I'll talk to you about the process that we use and then what the
strategy is at this particular point in time. 

I had the opportunity in July to give a presentation to the Worldwide
Sales and Marketing meeting which was hosted by Ed Lucente and included
many members of the senior management at Digital. During that meeting I
gave a presentation on product strategy. I thought initially I might
give that product strategy presentation it here, but I decided against
it, and in fact I am giving a somewhat messier version. The reason that
I did that is that the one that I gave at the sales and marketing
meeting really looked nice and tidy and neat and I suspect that
engineers probably couldn't stand anything like that. You really like to
see the nitty gritty. 

So, this strategy version I am going to talk about really will show you
that there are some things that we still are working, things that we
need to do, but will also give you an overview of some substantial work
that has been accomplished and the direction that we have set and we
will be sticking with. 

As Ralph said, I am going to talk and then we will have Q&A. I would
much rather have a dialogue in which there was interaction continually
while I was talking, because I think I could probably direct the
comments to fit the size of the audience, but I think given the size of
the audience and the way the room is set up we can't do that. But
normally I feel a whole lot better when I can have an interactive form
of presentation rather than just a monologue. 

When Bob Palmer took over in October of 1992 he had a list of
goals for Digital Equipment Corporation. They were goals that he felt
had to be acted upon very quickly to ensure the viability of the
corporation. I've given a list here of the goals, some of the goals
anyway, or at least my version of them, that he set down in October of
1992. They are sort of in priority order, although I would not try to
argue about the relative ordering of any two adjacent items. 

The first one was to stop the losses. As Bob often remarks, we were 
losing three million dollars a day in October of 1992, and if one does
the appropriate arithmetic forth one can see that the corporation would
simply cease to exist as a viable entity in a year or two, if it
continued that rate. So, clearly there was a major job ahead to stop the
hemorrhaging that was going on then.

The second goal was to reestablish the culture and values of the
Corporation. I think Digital has a very strong culture and had or has a
very strong sense of values -- although I think there was great concern
in recent years that some aspects of this culture and some of these
values were being tested as to whether they really were parts of the
culture and values of the corporation. 

It is also true that the culture and values of this corporation were
established in a time when the industry in which we work was very
different, so it is possibly that some of the culture and some of the 
values that was established then are not completely and entirely
appropriate for the industry in which we participate today. 

So, there was some real concern about what the culture and values of
Digital should be, and how we can ensure that once we decide what the
values and culture are, we in fact live by those principles as opposed
to violating thenm.

The next thing was to rebuild a functioning organization. As you know,
Digital had gone through enormous turmoil in several years prior to
October, 1992 -- constant reorganizations, an awful lot of internal
warfare and so forth. Bob believed it was necessary to get a clean 
organizational structure and get the various parts of the organization
working together. 

The fourth point market and customer focus, which is something of a
"motherhood" statement at one level is. You could hardly find any CEO of
any company who would say, well, we are not market- and
customer-focused. Of course we are, how could you claim otherwise? But
it was very clear that Digital was not market- and not customer-
focussed as a company; and it is very clear that that was getting in
the way of our success as a company. 

So, this in fact formed a major challenge, which is is to try to figure
out how we will reflect much more of a market and customer focus -- in
terms of our organization, the way we do business, and indeed our
culture and values -- than we had recently, and perhaps more than ever
in the entire history of the Corporation. 

The next thing was to internalize the realities of the current
information technology marketplace. Digital was created and grew up in 
an industry very different than the industry we face right now. And I
think that most of us, particularly the engineers in the organization,
clearly know things have changed in a big way over the last four or five
years. 

The real question is not whether you know things have changed, but
whether you do anything about it. And I think that many of us who know
things have changed in a very big way still have not quite internalized
what we actually have to do about it, if we are going to succeed
personally and as an organization. 

So it is very important that this remain a very significant goal. Not
only is it a goal that has to some degree been acted upon over the past 
year, but it is one that I think ought to remain on everybody's list of
goals. Have we really internalized what is going on? Do we really know
what we are going to do about it? Or are we going to just continue to
wish it didn't happen quite the way it happened? 

The next thing I think particularly significant for the Engineering
community is that I think we are particularly privileged to have Bob
Palmer as a CEO of this organization. Bob Palmer is very committed to
technical excellence as a differentiator. Even as we do things like get
a market and customer focus, which are so important, Bob believes that
in the final analysis, technical excellence is going to be a major -- if
not the major -- differentiator for Digital Equipment Corporation. And
it is very important that this technical excellence be something that is
not perceived only by us. If it is perceived only by us it is not a
differentiator -- it has to be perceived by the marketplace for it to be
a differentiator for the Company. 

The next goal was to get a clear Corporate Strategy. I think a Corporate
Strategy of one form or another is absolutely essential for a large
diversified corporation -- which has a set of behaviors that of
necessity causes a large degree of interaction and collaboration across
that organization. There needs to be some sort of strategy in which you
can judge behaviors and judge actions and make decisions and so forth. We
clearly have not had one for some number of years. Bob placed it as a
goal to get the process started to get one, and I am going to suggest to
you we do not have one yet, but we do have a process working on it. 

And the last goal -- and I put it at the bottom probably because it was
the one which couldn't be achieved in the shortest period of time -- is
to return to profitable growth. Both profit and growth are important
here, we need to generate a profit, one which is acceptable to the
marketplace, and acceptable return on assets. Growing is also a goal
because I think we recognize it would be very difficult for us to
continue if we do not have some sort of growth in our overall plans.

It is going to take a while to get there. I think we have made some
significant progress over the last three quarters or so to first stop 
the losses and even get into a profit position, but we are by no means 
assured a position of profitable growth. I think that in particular the
first quarter of the year is a very difficult one for Digital
historically. It is an especially difficult one this quarter because of
so much of our business outside of Europe and in Europe being in some
difficulty. 

So, this road of returning to profitable growth will not be one that is 
absolutely linear and monotonic. We have some formidable challenges
ahead of us in order to get to a position in which profit and growth are
in some sense taken for granted, something we can expect routinely in
our company. But I think we have made some significant progress there. 

I want to say a few words about culture and values. This is obviously a
hard topic to talk about, and I think one that you might think about
bringing up in the Q&A session to see whether we are really talking
about the same thing here -- and whether what I say matches what you
feel and see in this company in terms of culture and values. 

In putting some slides together on this topic I've listed two categories
around culture and values, things that I think there are similarities in
the sense that I think Bob Palmer's goals around cultures and values are
very similar to what we historically thought we had at Digital. And then
there are other things which I think show a real desire to make some
changes in the culture and values of the organization. 

In terms of similarities, there are four here I think particularly
important. First is honestly, integrity: the expectation that
individuals will deal with one another with a high level of honesty and
integrity. And I think that honesty and integrity go beyond the literal
sense of the words, but include the full sense of the interaction. It is
not right to say something that is technically correct but in fact
misleading. It is not right to agree to do something and then not do it.
There is a lot which goes into the whole issue of honesty and integrity
which goes beyond the literal sense of the word. 

Second is respect for the individual, and that is a belief that
individuals can genuinely make a difference in the organization. And
that as much as we talk about strategies and organizations and 
structures and so forth, in the final analysis it is individuals who
make the difference, and it is how we deal with individuals and respect
them and treat them that will make the difference in the organization. 

The third is fiscal conservatism, and that is running our company is a 
fiscally conservative fashion. This fiscal conservatism is an extremely
important point. If this company had not been run in such a manner in
its prior history I believe it would not have survived. Given the depth
of the problems that we had faced over the last couple of years, we were
able to survive as a company because we had an extremely strong
financial position. If we were a highly leveraged company or a company
without the assets that we had, we could not have survived the kind of
losses that the corporation had over the last couple of years. So, there
is a strong belief that we need to retain this because it provides a
kind of buffer for hard times. 

And last is technical excellence. I mentioned that earlier, but it is
the belief that technical excellence is a differentiator for this
corporation. 

And now I'll move to some implications of changes around culture and 
values. And it is not to suggest that elements of these things were not
in our culture before, but they seem to be things that perhaps were not
stressed or reinforced or actually acted upon the way we would expect
over the last several years. 

The first is teamwork, and that is the belief that even as we talk about
individuals making the difference, for the corporation to succeed,
individuals have to work together in teams. Often as not these teams are
not respecters of organizational boundaries. As you will see as I show
the strategy and the organizational structure which has emerged from
that, there is an enormous need for teamwork, both within organizational
structures and across them. 

The second thing is rapid, formal decision making. This, I think, really
is a change. I think that Digital grew up with a decision-making 
process that was based upon consensus. Fundamentally, you get everybody
involved. Anybody who wanted to be involved in the decision could be
involved and the process took as long as necessary to get everybody to
agree to something. 

Now there may have been a time when that was a good way of doing things
and there may have been a time when the company was small enough and the
rapidity of decision-making was suitable for that style and it could
produce good results. But it is clear that in the recent years that
process very often resulted in not making any decisions at all. You
could take so long to make a decision that it made no difference 
whether you made it or not and in fact, no decision was the real end
result. 

I am not suggesting this necessarily happened at every level in the
organization, but certainly in a lot of the places where I looked at the
decision making process, I would say that was the outcome. Bob Palmer is
looking for a great deal more rapidity and formality in decision
making. You collect the data, you make a decision, you re-open in the
decision only if the data's changed. And the people who have to make a
decision are those people who are directly affected by it. Now, one of
the by-products of rapid, formal decision making is that not everybody
will get what they want. One of the reasons why consensus takes so long
is you are fundamentally trying to get something that everybody agrees
to. 

If you put rapidity in your decision-making and you have a lot of
different points of view, some of those points of view are not going to 
be honored in the final analysis. So, this represents a real change for
us. I think many of us are very often accustomed to saying, "Well, the
decision has not been made." What you find underneath that is that the 
decision was not made along the lines I would have like to have made it.
So, the culture accepts the idea the decision that is not made. 

I think you will see a lot of change in that and it is going to take a
while to do it. Like every cultural change, it does not happen
overnight. And one of the reactions it is going to have on all of us is
that we have to accept the fact that when a decision is not made along
the lines we like, then we accept it and move on. 

The third area is a particularly important one and that is around
accountability . And again, I do not want to suggest that there was not a
lot of accountability in the organization, but I think collectively
there was not very much accountability at all. In recent years, the
culture allowed a lot of people to make commitments around things -- 
seemingly make commitments around things saying, "I am going to do this
if you just give me that in return," and then not delivering against it.
And feeling it was alright to make a set of excuses as to why they
didn't deliver against it. I think that the whole notion of
accountability is that excuses are not acceptable. If you make a
commitment to do something, and it is not obvious how you do it, then
you find some way to do it. The whole concept around accountability for
Digital will be an enormously significant change in the way we do
things. 

The two points I talked about earlier -- the idea that we have a
Corporate Strategy so we can have a common point we can refer to as we
make decisions in the organization and then a focus on the customer -- I
include as a goal as well because I really believe that it requires an
enormous cultural change in the company to get focused on the customer.
In the past, our culture was not to be focused on the customer. Our
culture was to be focused internally, to be focused on technologies and
things like that -- as opposed to being focused on the customer. An
enormous change in behavior will be required to achieve that customer
focus. 

So let me move on to the second part of the presentation which is around
the corporate strategy and organization. 

In order to rebuild the organization, as I mentioned in the earlier
slide, it was necessary to have some sort of implicit strategy guiding
the definition and structure of that organization. Since there was not
time to go through a formal corporate strategic process, Bob Palmer set
a couple of goals down, or a couple of initial statements, I should
say, around strategy. I would call that the initial strategic intent.

A couple of thoughts about strategy: You can see how these particular
statements (below) actually drove the way that organization was put
together. These two statements are actually very simple. If you do not
look at them carefully, you might view them as the same. But, in fact,
they are very different:

 The first statement is to be number 1 or number 2 in delivering
 information technology-based solutions for selected industry markets. 

In fact, the markets are defined nominally in industries -- so to look
at the banking market or take the aerospace market (or whatever), we see
that customers exist in these markets and they have problems to solve in
these markets.

This statement says that we will select a set of industry segments and
we will identify the solutions that are needed in those segments and we
will aspire to have a number 1 or number 2 position in market share and
delivering those solutions to those industry segments. 

 The second statement is to be number 1 or number 2 in delivering 
 selected products, technologies, and services to the general 
 information technology market. 

And here the statement means we will pick a particular technology, like
network hardware or high performance microprocessors or whatever, and we
will aspire to achieve a number 1 or number 1 market share worldwide
across effectively all markets. 

This market is defined by the technology, not by a particular customer
or a particular industry. We would like to achieve a number 1 or number
2 position in the industry in terms of market share and selling to that
market as defined by technology or product. Clear examples of that are
Intel who has 80% of the microprocessor market of the type that they 
manufacture. 

These statements look superficially similar, but are in fact quite
different. Whether a company can actually do both things at the same
time is an interesting question. And, I think last October there was not
any information to suggest that we either could or we couldn't, or in
the long run, which one of these two things would be the better one to
be in, or whether you could do both at the same time. 

So, there is essentially a stake in the ground saying we are going to
pursue both of these right now. We believe we have enormous internal
capabilities to pursue both. We do not have enough data or insight or
understanding to make a decision. It ought to be just one or just the
other. And we probably do not yet know how to manage trying to do both 
of these at the same time. 

But this particular set of thoughts very much drove the organizational
structure that emerged. And that organizational structure looks like
this:

There are nine Business Units in Digital Equipment Corporation. All the
business of the company flows through one or all collectively of these
nine Business Units. And every dollar of revenue that is received from
the customer accrues to one of these nine Business Units. 

In addition, there are four major functional groups. There are some 
additional smaller functions as well, but the four major functional
groups of Engineering, Manufacturing, Consulting (which used to be
called Professional Services) and Sales and Marketing, all serve these
nine Business Units. These Business Units effectively contract with the
functional groups to provide resources to implement their businesses. 

In some cases, the Business Units also have resources of their own. How
it is done is pretty much at the discretion of the individual Business
Units and we will talk about that in a little bit more detail as we go
on. 

The other thing that is particularly important is that all of these
Business Units are world wide, so the geographic bias of the company
around business is gone. We are not looking at our company in any 
significant degree by geography. We are looking at all of our
activities as being world wide activities. Now, the Business Units
themselves -- there is one centered on services; there are three
centered on products; and there are five so-called customer segments --
the industry segments I talked about before. Let's take a closer look
at these. 

The first is the Multivendor Customer Service Business Unit. It 
provides remedial hardware, network, and software service for Digital
and non-Digital products. And that is particularly significant -- the
non-Digital entity. To some degree, we have been servicing non-Digital
products for a while. But the important understanding or recognition is
is that for remedial hardware/network service, that business around
Digital-only products would be declining. 

And it is declining for two reasons. The first is that we have not been
selling as much, proportionately, as we were in the past and secondly,
the products that we are selling typically require less remedial 
service than they did in the past. So, if you just focused on Digital
product in terms of providing remedial service, that business would be
declining. There is no desire to have this business decline, so in
fact it is being expanded to include non-Digital products.

It is also being expanded to include different classes of activities --
one of which is software distribution and update; second is network 
installation; and third is desktop integration. These are services which
are sufficiently routine that they fall short of what you would call
systems integration service or fully-customized services. These are more
standardized services and that can be gotten into from a more
standardized basis such as you find in other aspects of the multi-vendor
customer service business. 

Now, what are the responsibilities of this business? Well, first is
Profit and Loss. They are responsible for profit and loss and the P&L
of this, added up to the other eight P&Ls, represent the P&L of the
company. They are responsible for setting the strategy around service,
which is what they go after; what the services are; what they charge
for them, and so forth. 

They are responsible to engineering to a very limited degree. That is,
they do engineering around tools that they use in the service business.
And they do some sales and marketing, although it is fairly limited.
Most of the sales and marketing around services is done by our world
wide sales and marketing force. But in some cases, we get specific 
off-Digital-base business. They have some resources of their own for
sales and marketing. 

One of the things I might say about this particular Business Unit is
that it accounts for more than all the profit of Digital Equipment
Corporation. Effectively, most of the Business Units of the company
report a loss. So the MVCS Business Unit generates more than all the
profit of the company. 

Next there are three Product Business Units. The first is personal
computers; Intel architecture/personal computers specifically. Second, 
storage devices; and third components and peripherals, which from the
standpoint discussion you can think of as terminals and printers. 

The objective of these Business Units is to provide standardized
products at an industry-competitive cost, and over time to generate a
major portion of their business -- perhaps most of their business -- off
of the traditional Digital base. 

In fact, the belief here is that these kinds of products are
sufficiently standardized in the industry and, rather than focusing on
selling them just to the Digital system base, they should be sold into
the general information technology marketplace. 

These product businesses were set up to respond to the stated strategic
intent -- delivering these technologies, products and services to the
general information technology market. And to a large degree, the
direction of the Multivendor Customer Services Business Unit is also
fits in this category. 

Now, what are the responsibilities of these Product Business Units?
Again, P&L and the strategy for their respective businesses. Very often
they do most of the engineering. That is, most of the engineering done
around personal computers is done in the personal computer business
unit, most of the engineering around storage is done in the storage
business unit and so forth. 

The reason for putting the engineering in the Business Units is because
there is a belief there is an extremely strong connection between 
engineering in the market -- and a kind of a loop, if you like, that has
to be very short and very tight between the marketplace and the
engineering organization. And in the PC market it is very common to have
product cycles and pricing that changes on almost a monthly basis. So it
requires extremely close coupling. 

The fourth PBU responsibility is marketing. 

Fifth is setting up channels of distribution. It is generally believed
that in order to reach all of the non-Digital systems base, it is
necessary to set up extensive channels of distribution to get these
products to new customers. And the Business Units are responsible for 
setting up and building those channels of distribution. 

The remaining five Business Units are called Customer Business Units.
And these CBUs are largely set up to respond to the first strategic
intent bullet that I showed earlier, having to do with delivering
solutions to particular industry segments. 

The five Customer Business Units are: Communications, Education and
Media; Discrete Manufacturing and Defense; Financial and Professional
Public Services; Consumer Process and Transportation, and Health 
Industries.

A lot of time was spent defining the grouping of the industries
included in these business units. I should say initially there is no
desire to take any Digital customers and tell them that, in fact, they
were not welcome anymore in Digital Equipment Corporation. So
all of the customers that we currently have were mapped into one, and
in some cases, several of these Business Units (in the case of 
conglomerates0. 

Initially, there was what you might call a portfolio effect here of only
wanting a manageable number of units, but also trying to include all
existing customers. It could be that over time, there will be a "honing
down" that happens here. And some of these areas might not be quite as
important in the future as they are now. 

Another dimension of this is that there is a lot of emphasis on trying
to achieve synergy between the industry segments represented by these
Business Units. If you look at what is going on in the industries, you
can see that things are evolving and are interesting in new ways. For
example, we are all familiar with the fact that every other article in
the business press now talks about the coming convergence of
communications, entertainment, and so forth. 

The grouping of the Digital customers or accounts into one business unit
is really reflecting the fact that those customers are very much
concerned about that convergence. Therefore, putting all of the accounts
and all of the customers in those industries into one Business Unit
allows us to put together strategies to address that convergence much
more effectively than if those industries were in different Business
Units. 

And to a greater or lesser degree that is true of all of these Business
Units. For example, the Health Industries CBU, includes pharmaceutical
companies, health care providers and health insurance -- and health
insurance was not included in the financial Business Unit. You might
think of including it if you were focused around just financial
services, But instead it was included in the health business unit
because there is a belief that health insurance is really part of this
overall health value chain in the industry and there is a very strong
coupling that is going to be emerging out of all the changes that are
occurring in health care these days. So, there is a lot of emphasis put
in trying to get the right grouping.

What are the responsibilities of these Customer Business Units? First,
P&L; second, strategy for their particular market segments; third, doing
the marketing for those marketing segments; fourth, determination of the
sales, technical support, service and system integration resources that
are going to be applied to those particular market segments. This
represents probably the most significant effort with respect to changed
behavior. Prior to the putting in place of this organization, we did
have an emphasis in this company around industry marketing. There were a
number of industry marketing groups which I am sure you are familiar
with. 

The problem was that these industry marketing groups had no ability to
drive how the resources were deployed in support of those markets. In
fact, the resources were deployed on a geographic basis and driven
entirely by countries and geographies. So you had, at a very high
level, a substantial disconnect between your strategy, which is being 
set by the industry marketing units and your resource deployment, which
is being set by an entirely different entity -- the Geographies. And,
what typically happens when you do not apply your resources to line up
with your strategy is that you get poor results. And, indeed, we did. 

So what we have done here is replace that informal thing that didn't
work well with a very formal process in which the contracts between in
the Business Units and the Sales and Marketing organization and the
Consulting Services organization -- to get resources in place for the
right accounts and implementing the particular strategies and solutions
-- are determined by the Customer Business Unit. 

Two other things that are important here relative to Engineering are the
determination of industry-specific products and applications. That is,
the Customer Business Units are responsible for identifying which 
products and applications (that are unique to their industry) are needed
for those Business Units to deliver solutions to their customers. And
they can acquire those products externally; they can contract to Central
Engineering to do them; or they can develop those products themselves.
But, in any event, they are responsible for getting those 
industry-specific products. 

Last, and of particular importance to us, is that they are responsible
for providing the non-industry specific product requirements back to
Engineering. This represents another very major change. Previously in
Digital, Engineering really had no product requirements process.
Engineering basically decided what products it wanted to build and
basically tossed them over the wall and hoped somebody would sell them.
For quite a while that worked very well. Recently it has not worked too
well at all. 

What is signaled by this is that we are going to have a formal
requirements process. The Customer Business Units are going to place
requirements on our products and in return, they are going to give us a 
commitment. They are going to sell them, which of course, is the other
side of the contract. So, I think this will be a very interesting
experience as we put in place our product requirements process and
actually see if we can mutually engage in this contract -- where they
tell us what they want and, if we give it to them, they sell it. 

At this point I will move on to Engineering and talk about what we are
doing in the Engineering organization. 

At the same time we were trying to build a new corporate structure, 
starting last October, we were also trying to build a new Engineering
structure. When I was given the privilege of heading the Engineering
organization, my staff and I defined a set of goals for restructuring
the Engineering organization:

First was to respond to the customer and CBU product requirements. 
Because the CBUs didn't exist at the time, there was not a whole lot we
could do about that last October. But we know that the BUs were coming
and we knew we had to put some processes in place that would allow us
to interact with them. 

Second was to define and communicate a clear product strategy. As much
as we need a corporate strategy to make sure the corporation works
together, we need a product strategy to ensure that engineering works
together and so that we can tell our customers and our Business Units
what we are all about. 

Third was to identify and invest in key product development focus areas
and buy or partner for the rest. At one level this is another one of
those "motherhood" statements. Of course you focus on what you're good
at and buy or partner for the rest. 

But the truth is that, historically, we have not been able to focus. And
this is not due to any irresponsible behavior on our part. We grew up in
an industry that provided highly proprietary systems. In making those
highly proprietary systems, you had to do everything yourself. If you
needed a network, you had to do a network; if you needed a database, you
had to do a database; if you needed a transaction monitor, you had to do
a transaction monitor, and on and on and on. We were delivering complete
systems to our customers and in order to do that we had to to deliver
all of the pieces because there was no general market for those pieces.
So we had to do everything ourselves. 

Over the years, Digital Engineering has invested in an incredibly broad
range of technologies. We start at the silicon level; we do magnetic
storage devices; we do operating systems; we do hardware; we do network 
hardware and network software; we do all kinds of different software
products -- a very, very broad range of technologies. I want to expand
on this point because it explains how and why we have changed our
engineering strategy for the future. 

Let me go back a little bit in time. In the industry as it existed five
or ten years ago, there were some vertically integrated companies like
IBM, Digital, and HP. These vertically integrated companies defined the
competitive playing field. Over time, this changed as niche companies
emerged and found ways to work between these vertically integrated
companies -- by focusing on a single area of expertise such as discs.

The vertically integrated companies were not terrible excited about
these new entrants, but they were tolerated, and sometimes embraced. For
a long time, industry dynamics continued to be led by the vertically
integrated companies. Slowly, we have witnessed a reversal of role.

Today, the niche companies, so to speak, get most of the market
attention. So when you think of databases, you think about Oracle, not
Digital nor HP nor even IBM. In networking you think about Novell, not
about Digital or IBM. If you think about operating systems, you think
about Microsoft. So, there is almost a complete reversal of role. In
fact, some people even say the vertically integrated companies are
dinosaurs and are going to go away. 

Now, as part of our process for examining our Research & Engineering
spending this Spring, we made a list of the best-in-class niche 
companies who engaged in activities that were similar in scale and in
scope to what we do. For example, in the database area, we picked
Oracle. We and they make good relational database products. The products
are similar in scale and scope. In PC integration, we picked Novell.
Look at Novell's network product and compare it to our PATHWORKS
product. And the list goes on.

What you will find if you examine that list is that the R&E investment
by the niche companies exceeds the R&E investment by Digital. Oracle
spends roughly three times as many dollars on database R&E as Digital,
and we are trying to compete with them. The same is true for Novell and
the others. The niche companies now outspend the vertically integrated
companies in almost all cases. 

Now, this is a critical consideration in the redefinition of Digital's
product strategy. The niche companies have a great deal of advantage
over the vertically integrated companies because they are extremely
focused. The niche companies are now grown to a point where they can
both have focus and also out-invest us. What this suggests to me
is that we must focus our own R&E on doing a few things well. We need
to spend at least as much if not outspend the niche competitors in
these chosen areas, and then partner for the rest. 

The next goal was building a simple Engineering organization. By
simple, I mean that it is very clear who is doing what. This gets back
to the whole issue of accountability. We will invest in a few
technologies, and there will be one manager who is responsible for
success in that area. There will be no overlap or redundancy.

The next goal was to get R&E spending in line with industry realities.
Bob Palmer gave me the job of cutting our R&E spending rather
significantly. At the time we were entering our fiscal year, we were
spending almost 13% of our revenue on R&E. And the target that Bob set
down for us was around 9%. We set 9% of revenue as our target based on
the benchmarking of competition that I mentioned earlier.

Last, but not least, a year ago we set a goal to achieve "Engineering
Excellence". My basic notion about Engineering Excellence is that
engineering is a machine. What goes in one side are product requirements
and what emerges out the other side are products. Excellence has to do
with how efficient and how effective that machine is in turning
requirements into products. We acknowledged that Digital's "machine"
simply was not world class in either its efficiency or effectiveness.
You simply can't generate world class products with less than world
class processes. 

A lot of what goes into generating a world class product is speed. If
you have a good idea and you have good technology, success is
determined, large part, by how quickly you can turn that technology into
a product. If you have less than world class processes, you simply can't
get the product out quickly enough, and any advantages you had in 
technology are lost. 

Another aspect of excellence is quality. I do not think that quality is
something you should have your field organization insert once it gets to
the customer. You must have world class processes to insure that you
engineer components with quality built in. Our commitment to engineering
excellence will guarantee that we have world class processes to achieve
leadership products. 

Let me talk a little bit about our progress against these goals. We set
goals in three phases. The first phase was to put in place our new
organization. This is now complete. Second was to move aggressively
around what I would call product rationalization. This was work that was
necessary to get to a significant reduction in our R&E spending. We 
have achieved our objectives in this area faster than anyone expected--
in a little more than eight months. 

The second phase lasted from January to June 1993. We started building
linkages to the Customer Business Units. We identified those areas
where we really want to excel, and we formulated a strategy based on
benchmarked comparisons to our competitors. We achieved these three
goals.

The third phase, implementation, began in July. You will witness the
results today and over the coming months and years.

Let me move on now to strategy -- first with a little bit of 
background. We will have a very strong focus in what we do in
Engineering. We will identify very clearly the activities in which we
ourselves arre going to excel and those activities where we will buy or
partner.

We have chosen three areas of technology focus for Digital. And since
strategy work is always work in progress, this represents a snapshot of
where we are today. The first area was high-performance silicon-based 
computer systems. As you are all very well aware, we have put an
enormous investment in our microprocessor technology, in terms of both
the architecture and processes that enable us to implement this
architecture. We believe we have absolute leadership technology and a
sustainable advantage in this area. 

It is unfortunate for us that leadership technology does not guarantee
leadership sales or leadership profits. Right now, the technology, in
large measure, is not selling at anywhere near the level necessary to
sustain the investment. But the Corporation basically said we are going
to stay the course here. We are going to invest in this area and we are 
going to put together the plans necessary to get us into a market
position that does sustain this investment. 

The second area focuses around networks. This is a traditional strength
for Digital. Digital was the leading peer-to-peer network supplier in
the early- to mid-1980s. We built our business in large measure on
networks. I think it is arguable whether it was VAX and VMS or networks
that was most accountable for Digital's success in the middle eighties.
Networking is an essential part of how our customers think about the
role of Digital -- they think of us as a company that is good in
networks. And that is an important strength. 

However, it is also very clear that in recent years we have lost our
lead in networking. If you pick up any of the networking publications,
you will find that Digital is rarely mentioned. And one of the reasons
for that is that, like in so many other areas, we lost technology
leadership. We were not focused on where the growth was. The growth in 
networking over the last 4 or 5 years has been in PC networking and
Digital has not had a major presence there. In another dimension
networking growth was around TCP/IP, and we decided we were going to go
in a different direction with OSI. Our market share declined because we
made poor decisions and did not execute well. 

We are determined to regain our position in networking.

The third area of focus is what I call software architecture. On the one
hand, Digital has enormous capabilities in software architecture -- that
is, in designing complex software systems that work. On the other hand,
in recent years we have not been able to exploit this capability
effectively. We have designed components that have simply not succeeded
in the marketplace. 

Our intent is to take our software architecture and deploy it in a
fashion that adds value for our customers -- and enables us to enhance
our market position. We will do this by moving away from development of
software components and look at software in terms of integration
frameworks. And the purpose of those integration frameworks is not only
to serve as products, but also to serve as a vehicle to lower the cost
of delivering solutions to our customers. 

Now what is our product strategy? Here is our product strategy in one
phrase: "Leadership open client/server solutions to deliver customer
solutions." Every word is significant:

Leadership. Why would you aspire to do anything less? 

"Open" is a great motherhood word -- everybody is open today, everybody.
You can't read an ad, an article or a product description, that does
not have the word open in. To us, open is giving the customer what the
customer wants and that is, in my mind, the most compelling definition
of open that I've heard. 

I think in the past we often thought that open was essentially a set of
standards. And that if we could just convince the customers that if they
adopted this set of standards, then everything would be fine. Well, open
was not quite that. Open was something else altogether. It was giving
the customer what the customer wanted. PCs were never "open," according
to our definition but that didn't stop customers from buying them.
"Open" is giving the customer what the customer wants. 

Client/server. I think in previous strategy statements we would have
used "distributed" -- we like to use the word "distributed." But
client/server is a better term because when we talked about distributed
systems in the past, we tended not to focus on clients. Servers were
good enough, or maybe distributed time sharing systems were good enough.
"Client/server" more suggestive than "distributed systems" in terms of
what we need to understand about the marketplace. 

Systems. This gets back to understanding that what we can do as a
vertically integrated company is to make sure that all the pieces we 
deliver work together. If we do not do that, we have no advantages in
vertical integration whatsoever. 

And then finally, to deliver customer solutions. That gets back to what
I said earlier about software architecture. There are a lot of
components that will emerge from our work in building client/server
systems, that can be sold as components. But ultimately, if we are going
to succeed, we have to make sure the technology, when it is all put 
together, enables the delivery of solutions to customers. 

Let's look at client/server systems at the next level of detail. First
are computer systems. Second are networks that link them together. Third
is software that enables the deployment of applications. Fourth are the
applications themselves. Fifth are the support services that support an
actual running client/server system. And sixth are system integration
services, which are used to facilitate the design of complex activities,
those which are not routine. 

For us to be a leadership supplier of open client/server systems, we
have to be leaders in each one of these areas. That is not to say we
must necessarily leaders in all the pieces that make up these areas. But
we must ensure that when it all comes together, the result -- as the
customer sees it -- is in fact a leadership offering.

The product strategy applies to the first three components of the
client/server solution:

We have explicitly decided as a part of our strategy that we are not
going to develop end-user applications. So when the product strategy
mentions delivery of leadership applications, we mean that we must have
the products, the strategies, and the services that encourage those who
do create leadership applications to put them on our platforms. Our
progress in these last three areas is led by other functions in the 
company.

Our product development will focus on three areas: leadership computer
systems (silicon hardware and operating systems), leadership networks,
and leadership open client server software frameworks. So let's look at
our strategy in each one of these areas:

Computer systems strategy. The first point is that we have to ensure
continuing Alpha AXP performance and price/ performance leadership.
There is absolutely no purpose to having the Alpha architecture unless
we can keep leadership performance and price/performance. So it is
absolutely imperative that our basic microprocessor technology remains
on the leading edge. If it were to fall off, it is all a waste. And we
will build personal and server systems that take the chip performance
and actually translate it into system performance. 

The second thing is low cost. Hardware is getting cheap very quickly.
And the fundamental focus we have to have in everything we do, be it
personal systems or server systems, is focus on cost. 

Third, we want to support the operating systems our customers want,
namely, UNIX, OpenVMS, and Windows NT. Lots of people ask me, "How can
you afford to do three operating systems?" I have a simple answer to
that. There isn't any one of these three we could afford to drop. More
importantly, Alpha AXP is the only architecture in the industry that
can support multiple operating systems without bias. 

If I look at the next level of detail around hardware, our objective is
to produce two families of systems: a family of personal systems and a
family of server systems. The personal systems are differentiated by
processor and graphics performance. The server systems are
differentiated to a large degree by processor performance, because they
can be a both single and multi- processor, but they are also
differentiated by storage capacity. 

Both these systems are based on some common elements: Alpha AXP and
leadership balanced performance. This means that we need to look beyond
just the microprocessor performance, and look into the storage,
graphics, and networking areas to achieve total performance as the user
sees it. This also implies adoption of PC industry business practices
and use of the standard PCI I/O bus. 

In the personal systems area, cost and graphics performance are the key
differentiators, assuming you have a leading microprocessor, as we do.
We have got to become very good at controlling the cost side of the
equation, as well as the graphics side. Those are the only
differentiators which are available to us in this area. 

In server systems, storage capacity is clearly a differentiator, as is
processor performance. Server systems are going to play a different role
in a true client/server environment. They really are servers, and they
are not typically places where major applications are run. Applications
very often run on the client. 

The criteria about what constitutes a good server system is different
from that of the client, and we must change our mode of thinking in this
area. For servers, we will look at the total performance "loop:" a
request arrives on the network at the server; the server then does
something, such accessing a disk, getting the information back, and
putting it back on the network. 

You do not judge the performance of such a server system in terms of
SPECmarks. Instead you look at the instruction count between a request
arriving on the network and the time the satisfied request appears on
the network, and how many of those kind of activities can the server can
do at the same time. So as we think about our engineering in the server
area, we will not be using the kind of criteria we might have applied to
a personal system or an earlier time sharing system. 

One school of thought says everything is going to the desktop, and if
you want a server, you flip the desktop system on its side and declare
the job finished. I think there are a set of attributes around servers
that are different than personal systems. And if we do well at those, we
can have a bright future in the server area, as well. Such
differentiators in the server space include some characteristics in the
areas of modularity, reliability, the ability to add storage, and the
ability to repair things online. And these are reasons for me to believe
we have a good opportunity for in servers. 

You know we have three strategies around operating systems: UNIX,
openVMS, and NT. The tag line around UNIX is leadership unified UNIX
now. We believe that we are ahead of the pack around the support of
multiple standards in our UNIX system. We support Berkeley, System V and
OSF in one product. Everybody is moving to convergence. We believe we
have a great advantage in being there first. The other advantage we have
is a 64-bit native implementation. We bit the bullet and went to 64 bits
early. Everyone else is obviously going there. 

The most important new developments are in our commercial capabilities.
A surprising number of our customers want to deploy commercial
applications on UNIX. 

For OpenVMS, we have made major investments in terms of porting it to
Alpha AXP. This means that the major disadvantage OpenVMS had in the
past -- price/performance -- is gone. We have made other major OpenVMS
investments in the area of open systems in order to achieve portability
and interoperability. So another disadvantage of OpenVMS --
interoperability with the open environment -- is also gone. 

The third area is Windows NT. Our tag line here is that we provide the 
first full-line vendor support of Windows NT. Here we are trying to
achieve a differentiated position in the industry. The other major
computer vendors -- Sun, HP, IBM -- are not embracing Windows NT fully.
They are trying to do with Windows NT what we tried to do with UNIX a
couple of years ago -- they are trying corral it off to one side and say
it is only useful for a couple of things. In contrast, we are putting NT
in our mainline strategy. 

Digital has basically said that we are going to declare up front that
Windows NT is a strategic system for us. We believe over time it will be
a full scale system, an enterprise level system, and we are going to
support customers if they want to deploy Windows NT as an enterprise
level system. So we will have NT support in our full range of hardware
over time; Windows NT will be a full member of our networks; Windows NT
will fully support our client server enabling software; and Windows NT
will be backed by full Digital services. So we are trying to position
ourselves as the vendor that can sell and support Windows NT as a full
line, or enterprise level system. No one else is doing that. 

In terms of our network strategy, we are going to try to provide the
hardware and basic software infrastructure for open client/server
computing. The strategy is simple; we are focusing on three areas.
First, we want to play in the high-volume price/performance area of
networking hardware. Right now this is network hubs. As you know, we
announced our second generation hub product just a few weeks ago, and it
is a very competitive product. And we are hoping we can grow and take
some significant market share in the hub area with that. 

At the other extreme, the very high-performance level, we are investing
in switches for high performance ATM and FDDI backbones. Even though the
total size of this business is probably smaller, we believe we need to
be in this business because this is where the leading edge technology
happens, and we want to be seen as a full line supplier of networking. 

The third area is around wireless and mobile computing. This is not now
a very big market, but we believe by the end of the decade, a very
significant fraction of all network connects in the information system 
arena will will be wireless and mobile connects, not the hard connects
that you see around the desk today. And we are starting to invest in
this area so we can be there when this market opens up. 

Let's look next at network software. The basic strategy is to produce a
backbone. That backbone is based on open standards such as TCP/IP, and
to some extent OSI, DCE, CORBA, and so forth. We will plug our systems
-- Windows NT, OpenVMS, and UNIX -- directly into this backbone. Given
all of the standardization activity which is going on in the market 
today, we also expect that other vendor systems, particularly their UNIX
systems, would plug directly into this backbone as well. This is the
center of the networking strategy. 

But right now, there aren't many other systems in the world that can or
would plug directly into that backbone. And maybe not in the future,
either. They interface to different standards, either by virtue of 
legacy investment or volume, and they may not conform to those
standards. 

The first such set is PCs. Most PCs are on networks that do not comply
with these standards. They are on LANmanager, Appleshare, or Netware. So 
our basic need is, if you like, a gateway capability that allows us to
join the PCs that are on these networks to our backbone. And that is our
PATHWORKS capability. 

The other class of systems you want to connect with the backbone are
legacy systems like IBM S&A systems. And there is another set of gateway
products that connect to that. The fundamental objective of all this
work is to surround Novelle. Novell is, in my opinion, a formidable
competitor. We have tended not to see that, because they only deal with
PCs, and for the longest time we were not worried about PCs. But they
are moving up into enterprise networking. They want to establish Novelle
netware as the backbone; that is their objective in life. Our objective
in life is to keep from doing that. 

Now, let me talk about software. I said we really wanted to try to take
advantage of our software architectural skills, by redeploying them in a
way that would be more successful. We have been able to use those skills
in recent years, in software "above" the operating system. We have used
a lot of different words to describe this, and the one we happen to like
right now is "frameworks." But we might use yet another word, if we find
one that is more appealing, particularly in a marketing sense. 

The purpose of these software frameworks is to shift the emphasis from 
components to solutions. We want to lower the cost of selling and
delivering and solutions. We would like to develop these software
frameworks to provide architecture, methodology, facilities, and tools
to solve the basic problems in client server computing. 

So software frameworks are the building blocks on which you build
client/server solutions. They provide support for our systems
integration and system solution activities, and would in some cases be
seen as end products by certain users. Thus they have a spectrum of
targets, and very importantly, they are based on extensive use of third
party products. This is extremely important. 

Many customers come to us and basically say, "I've decided on a
particular technology." For example, "I've decided on Oracle databases.
That is my corporate strategy. Now, Digital, what are you going to do
about it? What kind of solutions can you provide, given that we have 
selected Oracle as our corporate strategy?" 

We have to have a software strategy that is very robust in the presence
of customers who make individual technology selections around
components. We cannot define software strategies that work only if they
use a selected set of Digital components. It has to be robust in that
way. 

We have talked about a number of different client/server frameworks. In 
the presentation to the sales and marketing meeting in July, I talked
about four frameworks or five, depending on how you count, and today
I'll talk about six software frameworks. Over time, the relative
importance of these frameworks may change, and some may be added or
dropped. And we may want to rename them or re-position them in some way. 
But I want to share with you the work in progress -- the finished
product will be a somewhat more stylized subset of these. 

What we want to do is think about some of the basic problems that exist
in client server computing. If you wanted to build enabling technology,
what would you do to solve those problems? The first such problem is
client server logical connectivity. Of course at a level below that
there is physical connectivity provided by the network hardware. But
what about logical connectivity? The fundamental problem with client
server computing is, how can you join any client to any server? 

That is the basic problem because in principle, that is what is needed
by the customer. The product we have that fits here is PATHWORKS, as
extended by some of our gateways. But we are now trying to look at this
in a very unified way. If you define the problem as: "How do we
logically join any client to any server?" and then define our product to
have that objective, something interesting happens.

The objective of the product is not PC integration; it is not print
services for PCs. It is fundamentally a product that allows us to join
any client to any server. If you think of it that way, you probably
drive your product development in a somewhat different way. It may not
look too terribly different from PATHWORKS, but it probably isn't quite
the way it looks right now. 

So the question here is how you define the problem, and how you define
the product that solves that problem. I suggest that what we want to do
in this particular area is define the problem as, "How do we join any
client to any server?" And as we do that, we may find there are some 
things we have to engineer; there may be some third party things we can
pick up to do this or that but we must put it all together and and make
it work. The Sales person could then walk in to the customer and say,
"Digital can join any client to any server." 

Rather than having the Sales person say, "Well, you can use the MAC with
the VMS server, and you can use the UNIX server for your DOS machines,
but you can't do this with that..." and so forth -- that sort of thing
simply does not help the Sales person. 

Once we have a package of pieces that can solve the basic problem of 
joining any client to any server, we can then begin to talk with the
customer about the next problem, which in this case is data integration:
"How can you take all the databases that are scattered around, and
provide a uniform interface to all of them?"

We have a product that comes reasonably close to doing that. It is
called Accessworks. So the second problem in client/server computing is:
"How do you take the data scattered around and provide a uniform
interface to it, for any client?" And after that, the next problem in
client/server computing is: "How do you link together what are
fundamentally end-user applications?" End-user applications like word
processing and spreadsheets and whatever.

And the problem is not only how you can link them together in the sense 
that they can exchange data, but how do you link them together as part 
of business processes? For example, to process expense vouchers (or
whatever) you need to bring various applications to bear in a certain
sequence. This kind of end-user application integration is what we call
workgroup integration. Our traditional product for doing that sort of
thing is ALL-IN-1 and Teamlinks. 

In a couple of weeks we will be announcing a new product that meets that
objective very nicely (now referred to as Object Works but it will have
different name at announcement). The difference between Object Works and
ALL-IN-1 is that ALL-IN-1 is time-sharing-based VMS-based solution to
work group integration, whereas Object Works is an open client/server
solution. 

The next problem is enterprise messaging. A fundamental need of many
customers is an enterprise-wide mail and messaging systems. The question
is: "What are the set of components such that, regardless of the client
or server or mail system, you can build this enterprise messaging
system?"

The fourth problem is production system integration. You need to bring
together a set of applications and a set of databases to solve
production systems applications, which are typically transactionally
oriented. The question is, "What is the framework that allows the
customer to pick a transaction model, a database, a forms package, and
so forth, such that we can all these things work together routinely?"

The fifth area we identified, which is a little more arcane and a little
more specialized, is in technical computing. Here, the idea is that we
have enormous capability with Alpha AXP and OSF/1 and to get back into
the technical computing market. The question here is, "How can we
routinely tie together all the things necessary to build a technical
computing environment?

And lastly, when you build one of these client/server environments,
there is a really big problem managing that environment. So what can we
say around a complete set of tools and capabilities that allows you to
manage the client server environment? We started on a path of something
we call we POLYCENTER. 

Workgroup computing implies the integration of various kinds of 
applications. The objective here is to surround Lotus. Lotus is 
setting the tenor in workgroup integration in the marketplace. They
have a very nice product, but it is a highly incomplete product. And I
believe the word "surround" is important. I do not want to stop Lotus 
Notes, but I certainly would like our work group software to be seen as
the framework, and Lotus Notes as just part of that framework.

Again, we have a traditional product here, but it clearly does not meet
the criteria of open client/server. The new product, Object Works, is
very representative of what a framework would be in an open environment.
It runs on the client of your choice (MAC's, PCs, UNIX work stations,
whatever). It runs on the server of your choice (UNIX, VMS, NT, DEC's
UNIX, IBM's UNIX, HP's UNIX, you name it). It allows you to connect the
clients to the servers with the networking technology you want (Netware,
PATHWORKS, whatever). That, in my mind, is an example of an open
client/server product. 

You can go to the customer with that and say, "What is your client, what
is your server, what is your application? We will integrate it all 
together." And we have a very powerful capability for doing that. In
enterprise messaging, the objective is to make an enterprise messaging
backbone that has built-on standards (the X400 and X500) and then has
gateways to all the kinds of mail systems that exist in a customer's
environment (legacy mail systems such as SNADS, PROFS, ALL-IN-1 Mail, 
whatever). 

Object Works also has gateways to all the PC LAN clients as well. In
terms of actual clients, the mail space now is increasingly dominated by
PC LAN clients like CC Mail. We have a gateway that takes that CC Mail
LAN and connects it to this enterprise-wide backbone. 

The objective here is to surround Softswitch, which is a company some
of you may not be familiar with, but is the leader in the marketplace 
around enterprise messaging today. It is another example of a niche
company that does not have anywhere near the technology we have, but
which does have the mind share of the customers. 

The production computing framework is the area where I think we are
probably the least far along in terms of formalizing what we are doing,
although there is actually some very good work going on, and we may be
very much closer to formalizing what we are doing here than those
comments might suggest. The framework must involve how we do these
various activities: transaction processing, reliable transaction routing
in a distributed environment, connecting the databases, and storage
management. 

The company I've identified as a competitor here, surprisingly enough,
is HP, not IBM. My experience is that in the new marketplace, HP has got
the mind share, again. Customers want to go from proprietary production
systems down to open systems. They view HP as having the strongest
product capability and the strongest market position to achieve that
downsizing. HP is the company to beat, for the time being. 

In the technical computing framework, we want to provide programming and
run-time environments that support high-performance parallel 
structures. These parallel structures can be multiprocessor systems, or
they could be workstation farms, or they could be massively parallel
machines. The idea is to provide a relatively uniform environment so
that people can develop applications and then put them on these parallel
structures with a minimum of effort. I've identified the competitor here
as Sun. 

And Sun is not at present the technical computing leader. But the reason
I identified them is that Digital owned the technical computing business
in the early 1980's with VAX and VMS. And we allowed the Sun to drive up
and carry all of it away. I would like to get it back. Those customers
are probably still with Sun, so that would be the place to get them back
to Digital. 

The last framework I want to mention is our system and network 
management framework. We have been working on system and network
management for a long time. There is an umbrella term that covers it,
called POLYCENTER. We made a decision about a month and a half ago that
we were not going to succeed with the work that we were doing around
system and network management. The product called DECmcc simply was not
going to succeed in the marketplace. 

The reason it was not going to succeed is not because it was not a good
product. In fact, people would typically say it was the most advanced of
the system and network products out there. The problem is that it had no
applications on it. None. We could not recruit any ISVs to write
applications for DECmcc. I am not going to go into why that is the case,
but the simple fact is that this was a case of good technology and no
applications. The basic decision was that we have to move to a system
and network management framework where it is possible to attract
applications. 

Most of the applications in network management do not come from Digital
or IBM or any vendor. They come from third parties and customers. We
have to create an environment that has a set of APIs that people
actually write applications to. We looked around and the one we found
the best combination of technology and ability to strike a good 
business and technology deal was IBM Net View 6000, which is the net
view product that runs on AIX. We struck a technology deal with IBM. I
am not going to discuss what the nature of the deal is, but basically,
we have access to that product. 

We are going to port it to NT and to our OSF/1 UNIX. And OpenVMS will
probably provide migration tools to get DECmcc applications to it.
Digital and IBM are jointly marketing this framework to third parties.
Digital and IBM are going to market system management to our respective
customers, with our own respective spin, but we jointly are going to
market to third parties. So the two of us will go out and try to recruit
third parties to write applications to this system and network
management environment. And new DEC applications that we write around
system and network management will be written to this API as well.

This is an example of where we made a decision to fundamentally get out
of a technology because we saw no way that we could make that technology
successful. This probably will not be the last thing that we do like
this. 

So with that, I'll finish up since I've gone beyond my target hour. 

We talked about how we are transforming Digital, and about some of the
goals that Bob Palmer had when he took over in October of 1992. We
talked about our corporate strategy, initial statements of stratagic
intent, and kind of organizational structure we are put in place to
respond to those initial intents. We talked about what we are doing in
terms of engineering restructuring and the goals for the organization,
including the organizational structure we put in place and the
background around the strategy -- which is fundamentally focus, focus,
focus. We then talked about the product strategy itself, which I
presented to you in a rather messy form that shows the work in progress
-- as opposed to the final form which looks pretty good. 

[end of presentation]


                       DIGITAL Internal Use Only
    

                         DIGITAL Internal Use Only

So with that, I am finished, and I would like to throw open the floor to
questions and answers. And I would really encourage you to ask questions
about anything. It does not have to be product strategy. You can be
asking about the organization. You can be asking about values. You can
ask about hiring freezes or salary freezes or whatever type of thing
that is on your mind. I am here to entertain as broad a range of
questions as you'd like to ask. 

Question: 

   I was wondering about our networks business. You said that we have a
   big networks business. We had an opportunity years ago. We didn't do
   very well in getting our PCs into the marketplace. And then one of
   the problems we had is that they were too expensive, because the
   quality we wanted to put into them to match the quality of our other 
   products, made them too expensive. 

   And it turned out, the customer wanted something cheap that he could
   throw away in a few years because it would not meet his needs. And it 
   turned out he was right, because if he bought that equipment, it
   would not be suitable to run Windows on. So at that time, if we had
   done something like bought the Commodore, which was cheap, and sold
   under that brand name, it would have allowed us to sell a cheaper,
   lower quality product without compromising the Digital name. We can
   make that product run very closely with our more expensive products.
   It would create a migration path. 

   I was wondering if we could do the same sort of thing with networks.
   If we wanted to be really big in the network business, if you keep
   selling stuff under the Digital name, people are going to think,
   well, I do not want to use this network stuff because I do not have
   Digital equipment. And if we were selling it -- say, if we took
   DECNET and all of that and rolled it up on the PATHWORKS instead,
   then people would see PATHWORKS in the store, even if they were not
   seeing other DEC equipment. And they would say, yes, this is the way
   you connect your computers together. And then our salesmen who are
   selling Digital equipment would say, yes, well, we use PATHWORKS
   because it is good and because we have a close working relationship
   with them. We own them. And I think that would allow us to get these
   separate business out into the marketplace, and not to get the
   customer stuck in a frame of mind that you only use that if you have
   other DEC equipment.

Answer: 

   There are a number of interesting thoughts in that, which we are
   pursuing in various ways. First, the observation that a company might
   want to create a separate line of products with a different name on
   them, to put in, say, a different quality standard in the
   marketplace, that has certainly been tried. In fact, most of the PC
   companies right now typically have two lines of products, where they 
   have their line more oriented towards the low end, and then the high
   end line. It allows them to maintain different pricing structures and
   so forth. So that is being tried out. In fact, we are probably doing
   some of the same thing in the most recent PC announcement, to really
   try to create two differentiated line of products. So some of that is 
   happening. 

   On the question about whether we get technology to the customer only
   through a direct sales force, typically, our model has been that we
   only sell our technology through a direct sales force. The product 
   business units have a very, very different model than that. In fact,
   their objective is not to sell it with a direct sales force. 

   One thing which I didn't mention about the Business Units is that in
   fact, in addition to selling terminals and printers, they actually
   sell some other Digital products. They sell things like DEChubs. What
   we are trying to do is to establish DEChubs in the marketplace quite
   independent of the DEC sales force. They would be sold through
   industrial distributors, much like Synoptics hubs and Chipcomm hubs
   are, and so forth. So they will not only be sold through our direct
   sales force as part of network integration projects, but they'll also
   be sold as components through industrial distributers. So, much as
   you suggested, we are trying to establish a customer base, a
   distribution channel, and a brand identity which is independent of
   the Digital sales force. 

   The third comment I might make, though, and this gives me an
   opportunity to talk about quality -- I do not particularly believe
   the story around Digital PCs being higher quality than IBM PCs. I
   believe we use processes which we thought produced higher quality,
   and therefore delayed the time to market and increased their cost. I
   could probably defy you to prove that our PCs were actually of higher
   quality. And I think that all gets back to the issue of what are our
   processes in total, from how we figure out what the customer
   requirements are, to how we actually implement and deliver products. 
   And I believe what happened in the PC space was a total failure of
   the process. 

   Individually, the engineering that we did was quite good but the
   product was a total failure because we didn't understand the market
   requirements, and we didn't deliver in a timely way. And that really
   comes back to is whether you have quality engineering processes. And
   we didn't have them, and we had a massive product failure, even
   though we didn't have an engineering failure. And I think that has to
   be applied to everything we do. 

Question: 

   You had emphasized a shift to a much stronger focus on customers.
   How are we going to go about doing that? How are we going to
   formalize that a little bit more?

Answer: 

   Well, the first thing is that, for the products that sell into that
   first segment (which are most of our system products) the way we want
   to get the customer focus is to have a very tight coupling between 
   Engineering and the Customer Business Units. The Customer Business
   Units are out on the front line, selling solutions. They have
   requirements on what products they need to sell those solutions. So
   we are trying to couple with them very strongly in our product
   planning process. They were coupled to a limited degree over the last
   planning cycle, primarily because they were not up and running fully
   then. But in the next cycle, I would imagine, will bring much, much
   closer connection between the CBUs and engineering. 

   The other half of our products -- not half by count, but just the
   other class of products we have -- are the ones that are sold by our 
   Product Business Units. And I would not be surprised if more 
   products fit that model in the coming years. Those Business Units are
   extremely close to the customers. They deal very directly with the
   customers. So products that have this really, really tight loop and 
   extremely short product cycles, I believe, have to be cast in more
   of a Product Business Unit model than the traditional model that
   Digital has, which has several layers between the engineer and the
   customer. And that is the other thing, a structural way of dealing
   with it. We are starting the experiments to do that. If those
   experiments are successful, I think we will extend them.


Question: 

   Would you share your plans to re-establish technical excellence?

Answer: 

   Share my plans for establishing technical excellence. That is a
   simple question, I guess, with a complicated answer. I think that
   technical excellence, in large measure, is obtained by focus. And
   what I mean by focus is that you probably can't be excellent in a
   huge number of areas. You probably have to pick. Because if you do
   not pick, you're going to end up with a mish mash of all sorts of
   things. 

   Secondly, you have to understand what the technology is, and you have
   to understand what the competition is in those areas. And you have to
   understand what the customers are going to do with the technology. So 
   all those things have to be put together. But the whole thing
   underlying it, is focus, and obviously good people, good plans, good
   strategies. I do not really, know how to answer your question more
   crisply. We do have areas in this company in which we are technically 
   excellent. There is no question about it. We are technically
   excellent, for example, our microprocessors. And it was a very, very
   long, hard process to get to that point. 

   Ten years ago people laughed at our microprocessors and thought we'd
   never be in that business. Now we are in a world leadership position.
   How do you do that? You do that by focus. Picking an area,
   understanding the competition, understanding the technology, and
   having a very clear goal to be the best at the end of that process.
   Maybe there is something more I could address. I am just not sure I
   can give a general prescription of how you become excellent. It seems
   like a hard task. 

Question: 

   Do you have revenue targets associated with the various areas and the
   product strategy that you outlined? And if so, what is the growth
   rate associated with those targets?

Answer: 

   Yes, we do have revenue targets. Again, I am talking about a process
   that we are putting in place. So again, if you ask about how is it
   all working and everything, I do not know. But what we are trying to
   do is to get a true product marketing focus that really has revenue
   objectives against the product marketing activities. 

   Now, for a variety of reasons, we are not going to be doing the
   product marketing on a per component basis. We have too many
   components -- about a thousand components we are engineering at any
   given point in time. And having a thousand product marketing
   activities is absolutely crazy. You can't do that, and that accounts
   for a lot of the problems we have had in the past. We have taken all
   of our product marketing activities and focused them into six
   "initiatives," as they are called.

   There are six activities in product marketing that have specific 
   revenue objectives against them. And those revenue objectives -- or
   the status of those plans -- are reviewed jointly by me and Ed
   Lucente, who is the head of Worldwide Sales and Marketing. We review
   those on a monthly basis. The actual magnitude of the revenue growth
   for products is not very high this year. I think it is in around the
   15% range, which is not high in absolute terms, although it is high
   by historic terms. So it will be a stretch to achieve that. But if
   you want overall for all of our products, it is in about the 15%
   range.

Question: 

   The worldwide statement you made earlier -- can you demonstrate how
   we are actually going to approach that scene, as Asia is the largest
   growing market area -- and Eastern Europe and what have you? Are we
   really going worldwide? From a product perspective, how we are making
   our products suitable for the worldwide market?

Answer: 

   OK. We are doing it slowly. However, I think we are going to put a
   whole lot more energy behind it. I basically sent out an edict which
   said, by December of 1994 we would not expect any products at all
   that were not internationalized. Now, I am sure people will appeal
   for exceptions, and there may be some case where we will give some 
   exceptions, but by and large, the operative plan is that we will have
   fully internationalized products everywhere, across the board, by
   December of 1994. 

   So that is one kind of concrete action, which is to put out this top
   level edict. We sent it out to engineering managers. And we just will
   not ship anything that isn't internationalized. I do not think that
   fully addresses your question. It turns out that Asia actually is a
   growth area for Digital. We are not growing in the US. We are not
   growing in Europe. We are growing in Asia. There is a growing area
   force there. It is not by volume or size the biggest area, but it is
   the one with most rapid growth for us as a company. And we do have 
   plans to participate in the Asian market.

Question: 

   I have a question about commodity operating systems. It is really a
   marketing question, obviously. You mentioned it very briefly. Could
   you elaborate a little bit more on what our plans might be to
   establish significant market penetration with operating systems,
   both network operating systems and, for instance, OpenVMS? Are we
   going to open VMS to our Alpha AXP licensees so that they will also
   be open to VMS licensees, and expand the base for some DEC 
   operating systems?

Answer: 

   The specific situation around OpenVMS is one where I would not be
   especially optimistic that we are going to be able to turn that into
   a commodity operating system. We have certainly had on the table for
   our various current and potential Alpha AXP partners, the fact that
   we are willing to license VMS as we are the Alpha AXP technology
   itself, but we have not had any takers. There is very little interest
   in the industry in anything other than UNIX and NT these days. So I
   would not expect that we would be particularly successful at
   commoditizing OpenVMS. 

   In terms of the operating system itself, we also do not have any
   plans, if you like, to commoditize UNIX. Now, the industry probably
   does. And we are participating in the various industry activities
   that might get UNIX to a much more commodity status. I think it was a
   very interesting dynamic going on there, and it is largely driven by
   economics rather than technology. So it will be very interesting to
   see how that plays out. 

   The third area is Windows NT, and that is a commodity. And Microsoft
   is driving that business very much in that direction. And obviously,
   if no one else figures out how to match them, they are going to walk
   away, I think, with most of the marbles. Now, fortunately, I think we
   are positioned to take advantage of that, because we are building a
   good piece of our strategy based on that success. But obviously, in a
   business which has Windows NT in the middle, we do not get much value
   added out of Windows NT. Microsoft gets that. We have to do the value
   adding in other areas around the platforms, the networks, and the 
   solutions. But it would not be around the operating system.

Question: 

   You talked about how Digital is unable to do everything. We are going
   to have to partner with people. I noticed, though, later on in your
   presentation, when you mentioned our competitors like Novelle and
   Softswitch, you said "surround them." Why do not you just say acquire
   them? Isn't that the fastest way to get a leap ahead?

Answer: 

   It is probably cynical to say the fastest way to lose a lot of money
   is to acquire a company. There are not a lot of positive experiences,
   I think, in the technology area, around one company acquiring
   another. It is not to say it is never successful. But the acquisition
   of a company, I do not think, is a panacea. 

   It is very difficult to acquire a company. It is costly, and there is
   a very elaborate integration process, which is not easy to do. The
   costly part is not insignificant these days. Many of the niche
   companies that I mentioned actually are worth more than Digital.
   Digital could not acquire Novelle, even though Novelle is a tiny
   company by Digital's standards. Its market value is larger than
   Digital's. 

   So you end up with a very funny situation. It is not a viable option
   either, even if you could. I think, when I use the word "surround," I
   mean that we are not going to displace Novelle from the networking
   business. They have a very big chunk of the PC integration business.
   And we should have a strategy that recognizes that, as indeed we do. 

   But we should position ourselves as the network backbone supplier,
   and we will provide an interface into existing Novelle networks, and 
   we will work with them to do that. But by and large, we want to own
   the backbone. That is what I mean by "surround." We own the
   backbone. We establish the architectural franchise. When you pick up
   the networking rags, you want to see Digital mentioned as the lead
   company. And of course, Novelle will be part of the equation, but
   they are mentioned second and we are mentioned as the lead. That is
   the objective we should have. 

   Any more questions? Does that mean everybody's happy? Seriously. What
   does it feel like being an engineer at Digital?

Question:

   In my role, I travel around a lot with the engineering organizations
   in the company, and I notice a bifurcation, where the members of the
   senior leadership team believe that Digital is on a fast path back to 
   profitability, but the engineers do not see that. I notice that when
   I need tasks to get done and I need cooperation from other groups,
   what used to be done in a twinkling, I get a lot of the DEC nod; I
   get a lot of excuses why it can't be done. When I ask for volunteers
   to help on a task force, no hands are raised. No one wants to be the
   nail that sticks up that is going to get hit by the hammer. 

   I would love to see the corporation get past this downsizing phase
   and now go going after revenue opportunities. Here's how we are going
   to do it, and we want all of the employees at all levels of the
   company to participate. We need to get active participation back in
   the company and a strong feeling that this is a company that we want
   to work for. It used to be, you had a career at DEC. And I think for
   many people, it is, you have a job at DEC. And that is not the
   company I would like to work for. 

Answer:

   What will it take to make it seem more like a career than a job?

Question:

   In part, talks like you're giving, showing where the company is
   going. And in part, giving people the freedom to take chances and not
   feel that they are going to lose their job if they fail. To encourage
   that entrepreneurship that we had when we first started. I do not
   know how you do that in a large company. But I think that is what
   made us great before. If we could encourage that feeling of
   entrepreneurship and taking a chance, with the full realization that
   sometimes you take a chance and you'll fail. Maybe one out of three
   times, or two out of three times, you'll fail. But you want to
   nurture that guy with a great idea.

Answer: 

   So your concern is that the current climate of downsizing and expense
   controls and everything are very inhibiting to taking individual
   initiative.

Question:

   Taken to a severe degree, yes. DEC has a reputation for slow
   management. And it has not gotten better, from my experience. It
   still takes forever to get product decisions made. And one of those
   slides you put up suggested that it was a high priority to get that
   fixed. To give you an example, I am working on Alpha AXP notebook
   products, and I've been trying since December to get a clear
   management direction as to what products we want to field, and
   whether we want to do them inside or outside. And there is still now,
   nine months later, no understanding of what is going to actually
   happen in that space. 

Strecker: Are there some more experiences?

Question:

   I think that one thing that will help a lot is if DEC were making a
   lot of money. And I think that, as this gentleman pointed out, a lot
   of DEC people, engineers and every other function, still do not have
   total confidence that we have got a total understanding of the
   marketplace, and that we can really grow again. I am included in
   that. And I think DEC still does not understand pricing structure.
   The question I asked about commodity operating systems -- I was
   frankly not all that happy with the answer. I think that we have to
   understand that in every business we are in, we have to be the price
   leader. Even look at Acura's Legend advertizing. They advertize that
   they are cheaper than Mercedes Benz, BMW, and even Lexus. And that is
   their push: price. And they are in the Cadillac, so to speak, type
   market. 

   And we are in every market from commodity stuff to high end product.
   But in every space, and especially operating systems and software, we
   do not emphasize being the price leader enough. And I think that is
   obvious. That is why a lot of our stuff failed, from PCs to the
   DECmcc -- it is just not viable. We could probably make OpenVMS a
   commodity operating system if it is priced correctly. We do not
   understand pricing. It is not an engineering issue. 

Answer:

   Let me first totally agree with you around the pricing issue. You
   have to understand what your product is worth in the marketplace,
   vis-a-vis the competition. And you have to price it appropriately, or
   you lose the business. So in that regard, I think you're absolutely 
   right. I think there is a lot of evidence that we are moving in that
   direction. You may not think quickly enough, but I think if you
   looked at the pricing situation that we have right now, compared to
   where it was, say, two years ago, there is been a dramatic
   improvement. 

   The other situation which you might contemplate a little bit is that,
   if we were to instantly move to a commodity pricing structure,
   probably about 80,000 more of us would be out the door. Because, in
   fact, there is not enough margin with commodity pricing across the
   board to support an organization of this size. And it is also
   probably true that a lot of what we do isn't commodity. 

   I do not think that the way OpenVMS is typically deployed, for
   instance, really is a commodity product. And the expectation of the
   customers around how we are going to sell it and support it and
   service it and everything, do not support commodity pricing. So
   therefore, saying instantly, if I put a $100 on OpenVMS, that would
   solve all the problems -- quite likely, all it would do is reduce the
   company's revenue by half a billion dollars, and a lot more of us
   would be looking for jobs elsewhere. So there is a very complex task
   here, navigating through this and recognizing in some cases, what we
   are doing is not commodity work. 

   The third thing I might say is that I think it is important that we
   do have some realism. It is clearly true that, for instance, we could
   put a major thrust behind trying to make OpenVMS a commodity
   operating system. It would probably require some substantial
   re-engineering, because it can't be used that way right now. And it
   would require some substantial marketing and a whole lot of other
   things. You might ask whether that is something we should be doing.
   Digital has a very finite amount of resources. Is that really where
   we want to put them? 

   I think the biggest single challenge for the company is one of
   choice. Finite resources, infinite opportunities. Which one do you
   focus on? And we are going to make a lot of touch choices, I think. A
   lot of choices around which you could have argued the other side.
   Should we have put a whole lot of money in trying to make DECmcc a
   standard? Maybe we could have. Should we spend a whole lot of money
   making OpenVMS a standard? Or should we put a put a whole lot of
   money trying to take a good position in networking right now, and get
   it into a better position? 

   These are the kinds of choices we have make. And I think the really
   difficult problem is that this is a massive, massive cultural change.
   For the first thirty years of Digital's existence, the problem was,
   how do you spend all the money the business is throwing off? We were
   growing at 40% a year. The basic thing is, how do you spend all this
   money? Now the problem is, how do you really, really focus in on just
   a few things you're going to do? That is the big cultural change that
   is really hard to get through. It is one we will get through. But it
   will seem very disappointing, I think, when we see all these
   opportunities we might pursue. But the reason we are not going to
   pursue them is that we do not have the resources to. So it is a
   tough, tough thing, but those are very appropriate comments you made. 

Question: 

   I was going to ask you, when you were talking about mind share, how
   do we win that? You just talked about it a little bit. But I sort of
   feel that it is more than price. And it is an issue of backing the
   things you're going to do. And I think spirits are kind of low in
   some parts of Engineering, because people do not feel like the
   management is really backing the thing they are working on, or
   anything that they can go and work on. A lot of engineers,
   especially in software, do not understand what they should be working
   on that is going to be valued by the company.

Strecker: 

   You're actually talking about internal mind share, I guess?

Question: 

   I was talking about internal and external. There is this issue, I
   think, of, are we going to really push and promote the things that we
   build? And I am concerned a little bit also about the CBUs and their
   role in this. Recently there was a note floating around the
   engineering network, somebody's summary of some input from CBU's, and
   it mentioned a whole list of things that we are currently working on
   that the CBUs didn't seem to be very interested in. That affects
   morale.

Answer: 

   Let me try to break that up into a couple of areas, because I think
   it was a compound question of sorts. One of the things that is true
   about the current structure is, in some sense it is perhaps more
   anticipatory of what is going to happen, as opposed to a complete
   reflection of what is happening. You might say the industry is
   bifurcating into systems integration and components. But it turns
   out, most of Digital's business is neither system integration nor 
   components. So you put in place a structure which you think is very
   forward looking, which is an interesting perspective, and yet you
   recognize it does not map especially well under the current business.

   So therefore, a lot of what we do, if you like, the traditional
   VAX/VMS systems business and all the things that go around it, do not
   seem to clearly map into that. Neither the Product Business Units
   have a big interest in marketing those products, nor do the CBUs, who
   tend to see themselves more as system integrators than as product
   marketers. So there is a problem in that. I think that we will get
   through that. The reason I think we will get through that is because
   I believe that there is an emergence, re-emergence if you like, at
   the senior management level, of the importance of products. 

   And you might say, "That is interesting. How could it possibly be
   that such interest was not there?" Well, it was not there in recent
   years. There was an enormous emphasis, particularly on the sales
   force, around services and not products. Basically, they had given
   up on selling products. They assumed they couldn't sell products, in
   part for good reasons, because Digital's product engineering was not
   really headed in the direction our customers actually wanted, in many
   cases. 

   So it was a convenient agreement, a handshake. We would not engineer 
   quite what they wanted, and they would not sell it. (laughter) That
   is not the kind of handshake you like, but that is the way it was. So
   we continued doing what we were doing, and they started putting a lot
   of emphasis on services. 

   I think that there is a very strong recognition, first on the part of
   Bob Palmer as CEO, that products are absolutely essential. We will
   fail as a company if we do not succeed in products. But even more
   than that, there is another individual in the company who thinks
   products are important. It is not me, though of course I think they
   are important as I hope you do. But we brought in a new head of
   Worldwide Sales and Marketing, Ed Lucente. And Ed Lucente is
   extremely product-focused. He goes out and he tells his sales force,
   "Services are nice, but if you do not sell products, the company
   fails." There has been nobody in the Sales organization at Digital
   that has given that kind of message to Sales for ten years. 

   So I believe that once again we are talking about something that will
   not change overnight, but I believe that the management of the
   corporation is recommitted to products in a way that it has not been
   done for at least five to seven years. I think we are going to go
   through a lot of pain this year. We are going to go through a lot of
   pain because, at the surface, the current structure of the Business
   Units is not in the business of pushing the products that we have. 

   We are also going through an extraordinary amount of pain in that
   the marketing area is being rationalized this year. And as far as
   the initial step at marketing, unfortunately, it is kind of just an 
   across-the-board cut, which means that the product marketing
   resources that we have are even smaller than they were before, and
   that is an unfortunate situation. But I believe that over the next
   three months to twelve months, we will have significantly changed
   things. We will significantly restructure product marketing to get
   it behind really marketing the products we are focusing on. 

   And I think you'll see an enormous resurgence of interest in products 
   from the Sales force. So you can't look at it right now and say it is
   really great, because it is not. It is really awful. But it is going
   to be a whole lot better. I think all the vectors, in terms of the
   attention of senior management, are in the right direction here. So
   hang on for a little bit longer. We will see a change here. 

Question: 

   I find that a lot of our customers, and myself included, get very
   confused about some of the products that we have got, in terms of
   describing what we are selling or what we are offering. An example of
   the internal confusion is, we tried to start with just a marketing
   name for a product, and I am in the regulatory area. We are trying to
   figure out what the regulatory approvals for that product were. And
   we discovered that we didn't know what the product was, even though
   we had the name and a model number, because we looked in Apex and
   found one description; we looked in the drawings and we found
   another; and we looked in the catalogue and we found a third; and so
   forth. And they didn't match, even to the point where we couldn't
   figure out whether it included expansion boxes and things like that.
   There seems to be a certain amount of clarity that could be added to
   what we are presenting to the world, in terms of our product
   offerings.

Answer: 

   I can't speak to this particular problem, but the fact that we have
   an enormous number of overlapping databases describing our products
   and our product plans, which are mostly filled with obsolete,
   inaccurate data, is something that is very clear to me. And we had
   started a major effort at the engineering staff level to restructure. 

   This is part of several of our activities, part of the Engineering
   Excellence process and part of other things, to restructure the 
   databases that we have to describe what we are doing and what we are
   delivering, and to get down to having "one database" with the right
   data in it, as opposed to multiple databases, all of which filled
   with inaccurate and wrong data. So as I said, I can't specifically
   address your question, but we are very well aware of the fact that we
   are in a pretty poor situation, in terms of the descriptive
   information about what we are doing. And we are trying to fix that.
   OK, well, thank you very much. My pleasure in talking with you.
   (applause)

                         DIGITAL Internal Use Only
    

    In Bill Strecker's q/a, count the number of times these words were
    used:
    
    VMS	= 12
    OSF =  0
    
    Is this important to anyone?
    
    	Regards,
    
    	Ron
    

    probably not.  You might have noticed that he did refer to UNIX in his
    talk and that name does show up in the Q/A.  I imagine that the people
    asking the questions are mostly responsible for the mix.
    
    Mark

    Being here in sales for 7+ years is getting noteworthy in itself.
    
    Bill's comments that the PBU's don't need a direct sales force 
    certainly makes my day.
    
    -Peddler in NY
    

    	On the one hand, we're saying we need to be #1 or #2 in everything
    we do. We'll focus on fewer things, presumably profitable things in
    high customer demand, and focus on products on those areas. On the
    other hand, we're saying we have a lot of work to do to be #1 or #2 in
    anything...if we don't start making money soon, we'll layoff more
    people (which, as we all know, also won't help us to make money
    sooner), and worst of all, I get the feeling from this speech that
    we've already accepted defeat in these very same areas. So I don't
    understand this contradiction. For instance, Novell is the network
    leader. We'll be picking up assorted pieces that they don't do and not
    be going head-to-head against Novell. Can we be #1 or #2 in networking
    this way? Is this a profitable business? If so, what makes us think
    other companies won't also focus on these areas and eat our lunch?
    The same is true in operating systems, databases, etc. I didn't hear
    anything which suggested our Unix offering would be #1 or #2. VMS is
    obviously being used as a cash cow and nothing more. And we're not
    going to make money from WNT (the OS), only from anything value-added
    that we do on NT. This value-added work that we do on NT will be more
    commodity based than any software we've ever done. So, the one software
    area where we seem to have a chance to be #1 or #2 requires
    high-volume, low-margin, retail-outlet sales. He said himself that this
    results in a much smaller company than we have now.
    	All these leaves me to think that if we don't make BILLIONS selling
    Alpha chips to someone, we're screwed. Even after all the speeches and
    Q&A sessions, I fail to see where we will be making our profits in the
    coming year or two. I'm glad they seem to recognize the problems.
    That's something, at least. They also recognize that they have some
    difficult decisions to make. I sure hope they make them soon and let us
    in on it.

    Re: .6-
    
    >	On the one hand, we're saying we need to be #1 or #2 in everything
    >we do. We'll focus on fewer things, presumably profitable things in
    >high customer demand, and focus on products on those areas. On the
    
    	I'm confused by this direction and how it relates to the labeling
    of the services portion of the company as "Digital Multivendor Customer
    Services", where we will cut a service agreement on most anything. Our
    lifeboat in this space was that we could often substitute a piece of
    our equipment if maintaining the original wasn't practical. Fewer
    products in our domain mean less options in this context. Integrating
    equipment to derive a solution is often easier than ongoing maintenance
    and bug-fix resolution involving multiple vendors. 
    
    Phil

    Well,
    
    It struck me most is that the many of the BU's will not report a profit.
    On the other hand, MVCS seems to be the real (only?) money maker.
    Someone said this to me before but I just could not believe it until
    Strecker said so also.
    
    What is the reason for this or is it just our way of bookkeeping again?
    
    Cheers,
    Henny

>    	I'm confused by this direction and how it relates to the labeling
>    of the services portion of the company as "Digital Multivendor Customer
>    Services", where we will cut a service agreement on most anything. Our
>    lifeboat in this space was that we could often substitute a piece of
>    our equipment if maintaining the original wasn't practical. Fewer
>    products in our domain mean less options in this context. Integrating
>    equipment to derive a solution is often easier than ongoing maintenance
>    and bug-fix resolution involving multiple vendors. 
    
Life boat eh?

Just heard about one fiasco... a DEC VAR (partner? Whatever they call
them now) sold an end customer a "brand X" 4mm tape drive, along with
a DEC service contract (seems this outfit can sell service as well).
The contraption breaks after just a month of use, the F/S guy finds that
he has to order parts... 2 WEEKS is quoted for ETA on the parts! Customer
goes balistic so we replace it with a DEC drive. Haven't yet heard if we
plan to put the "brand X" back in place after it is fixed.

I'd venture to say we've eaten up 1 or 2 years contract income thus far
on this deal.

Lee

I guess I should also mention the whole issue of spare parts.

I am told that in a recent MVCS meeting, a large square was drawn on the
board with a smaller square in the corner, maybe 10% the area of the
larger on. The 'small' square represented 'traditional' DEC business, and
large square new, MVCS 'other vendor' business.

The subject was raised about spare parts... that with cost cutbacks the
stocking level of spare parts for DEC gear was (in the opinion of this
F/S engineer) not all that good and how were they going to be able to fix
all of this other stuff with the same kind of level of parts? The
response was basically one of "wait and see".

It would seem in some cases that our support (in the form of parts and
maintenance information) of non-DEC products lags behind our zeal in
taking in the dollars for maintaining them.

Lee

    
    Sounds a lot like the Rusty Jones lifetime guarantee.
    
       They look great when your selling them.  Just remember to go
    out of business before the cars start to rust.
    
    Ken

    I work in sales and we in the field have traditionally had to live with
    what engineering has "thrown over the wall". For whatever reason, some
    of the stuff that's being thrown over is beginning to look rather
    attractive and it's refreshing.
    
    I read the entire script of the cafeteria prensentation. We have heard
    Lucente's homily but not Strecker's here in the field. I think is was
    encourageing.
    
    For the last 2 days, I participated in a relatively low key DoD event
    where we showcased Hot Pink/Kubota (OSF/1) and the AXP 150 running NT.
    At the end, I heard a very encouraging comment from two different
    geographically located customers from the same company: "We didn't
    realize that Digital had this class of products". This customer is a
    Prime Contractor to the US Government who was showing one of their own
    products running on a Sparc 10 (next to the AlphaPC, it looked like a
    '286 in perceived performance). For whatever reason, this customer only
    knew us based upon our DECstation (5000) product line.
    
    Maybe all of us are changing. We had a customer come into the demo area
    near the end of the conference just as we were about to pack things up
    and most had already done so.The "customer" commented that they did not
    realize the exhibits were schedule to close so quickly and we replied
    that we would be please to show/discuss whatever was desired...and we
    did. This customer was impressed and completly reversed in their
    opinion of Digital as a technology vendor.
    
    
     I don't believe it will happen overnight but the market is eventually
    realizing that Digital has some leadership products and they are
    beginning turn their heads (back) towards us. Maybe Engineering really
    is doing some of the right things these days.
    
    Please keep it up as the Sales Force's salary these days takes up to a
    20% hit if we are not able to sell the products that are
    produced/engineered from "Up North".

Strecker's talk sounds encouraging, but the day after I read this entry I
talked to someone in a marketing group who said words to the effect of "no
way will we push ObjectWorks, it's too closed, and besides Lucente just
said to sell TeamLinks".  It's enough to make your head spin!

re: .-1
It seems as if we're in a position now where the only really strong case
we can make for Alpha AXP is as a UNIX turbo-engine (OSF/1).  So the main
battle there is getting rid of all the Ken Olsen "UNIX is snake-oil" taint
that Digital probably has in the UNIX market.  Looks like you're starting
to turn that battle around!  Maybe in a few years NT will emerge as a
viable alternative, but for now it looks like "UNIX or die (with our VMS
installed base)".
							- paul

re: .-1:

=> way will we push ObjectWorks, it's too closed, and besides Lucente just
=> said to sell TeamLinks".  It's enough to make your head spin!

You mean Percussion, er, ummm, I mean, Linkworks?

I think Palmer et al are going to personally demonstrate Linkworks at the 
October 12th announcement. I don't see how Linkworks is perceived as closed
when it supports virtually any windows shrinkwrap software as an underlying
application. Teamlinks doesn't.

    	It sounds like one thing that might help morale in software
    engineering would be to have more direct customer contact/feedback. We
    see notes like the one a few replies back and receive Alpha "good news
    bulletins" which highlight positive customer feedback, but it just
    bounces off me. Maybe others find it uplifting, but I don't. It always
    just strikes me as propaganda, lame attempts to encourage us to work
    harder with fewer people. If things were as wonderful as these messages
    suggest, the money must be rolling in, right? Wrong. You see a
    potential bad quarter looming, and you begin to disbelieve all the good
    news.
    	Management is not exactly cheering us up either. We're constantly
    being told our stuff needs to be much better to be competitive. We need
    to be more productive and do more with fewer people...and that's with
    stagnant pay and decreasing benefits...and if the company doesn't do
    better (which seems out of our control), more of us will be laid off.
    	When I speak directly to a customer or even to the DEC account rep
    for a customer, I get a much brighter picture. They're happy with the
    products, want to know more, and really, really appreciate the direct
    contact, open communication and feedback loop. It actually makes me
    feel good about my job...until some other silly thing knocks me back
    down again. It would help to somehow bring more customers into contact
    with more of engineering, even briefly, to boost morale. Strange how a
    thank-you, pat on the back, and signs of appreciation can make you feel
    better about you job.