‘Naturally we know it’s (Linux) on the server.’–Joachim Kempin, Senior VP, OEM Division, Microsoft

As Senior VP, OEM Division, Joachim Kempin looks after Microsoft’s worldwide OEM sales, marketing and support activities. He is also responsible for developing the company’s relationships with its PC manufacturing partners and OEM distribution partners worldwide, as well as for licensing desktop operating systems and other Microsoft products for PC hardware vendors. 
A mathematics graduate from the University of Hannover, West Germany, Kempin is also a member of the Microsoft Business Leadership Team. In an interview to DATAQUEST, Kempin fielded questions on how the Redmond giant plans to deal with the threat from ASPs, Linux, the 64-bit operating system and other such issues which threaten to unseat the company from its dominant position. Excerpts:

does Microsoft view the evolving ASP business?

If you look at hosting in
general, the pre-requisite is that you should have adequate bandwidth. But that
is not the case. You have it in some countries but there are many others who don’t
have it. So whenever you come out with any type of application hosting, you must
have solutions which take you around to the countries where you can go and do
business. So you just can’t get up tomorrow morning and say that you can do
hosting and that’s the only way you can get my products. You will still have
to do the traditional distribution.

So for us it’s just another way
of licensing. In good old days, you bought packaged products and signed a paper
license for use. Tomorrow, you would use an electronic network and just choose
the product. And for Microsoft, this, except the fact that you have a network
challenge here, is truly not anything bold. The execution is a little difficult,
for you will need partners who will be able to run big data centers and this can
be generic for the public or specific for certain companies when they are
outsourcing things or they want to have their own security. But this is more an
implementation chance than an intellectual chance.

But, in a country
there may be one geographical region which is adapting more quickly to this
rental model than the others. Over there, how will you manage issues such as
planning the contribution and cash flow?

There too it is very similar to
what we do today– that is we would basically have resellers. These will be
companies who will sell services to the customers. Then it comes down to who
builds it and that too would probably continue to be the way we do it now. As
all the partners are same, it’s not really any different.

In shrink-wrap
the cash flow remains clear, that is you sell and then you have the credit, but
in this case you won’t be selling.

Due to the time element in
introducing this, I don’t think this is a big deal for us. And, may be, there’s
a quarter where we won’t grow that fast because of this. At the same time, you
probably have this system in place where you actually are able to forecast your
business with that.

Are you saying
that Microsoft has already got its pricing structure ready for ASPs?

I have no clue, I don’t think
we have done that yet. I think they are still working on that.

Many vendors,
especially in the OEM space, are looking towards the Linux model. How does
Microsoft plans to deal with this?

OEMs still continue to ship a lot
of PCs with Windows on it. And I think, here you need to differentiate a little
bit on what and where Linux may be attractive, and what and where it may not.
What we are seeing is that on traditional PC desktops, there is no Linux, there
is Windows and Windows applications. I have not seen a lot of people write a lot
about it. But we are definitely seeing Linux activity on the server. And this
has been there for last 3-4 years.

Then we are also seeing some
Linux activity in devices other than the PC–on small compactors and little
gadgets. But the environment there is so entrepreneurial and undefined, that you
don’t know what is going to succeed. Oracle, for the last 4-5 years, has been
talking about internet access devices taking over the computing world and PCs
becoming extinct. They have in fact also come up with Linux solutions for these
devices. But consumers are disappointing them, as they still want to buy PCs for
a long time, since it enables them to do many things that they can’t do
otherwise. And a lot of these devices depend heavily on high bandwidth network.
Here again, countries like India still have sometime to go before they can
really reach some where. I see some activity on the development side on small
devices, but not PC devices. And naturally, we know it’s there on the server.

Regarding WinCE,
what is the current competitive situation? How does Microsoft look at WinCE’s
usage in new information appliances where OEMs would not be PC OEMs, but where
one would see a lot of manufacturers coming in?

You are correct, traditional PC
OEMs are not customers for that. But I think we have done an OK job to see CE
into these companies. I have a special group that does nothing else but sell to
these types of companies. I also see a lot of potential there. The key for us
here is to be sure that they understand the value which CE can offer. And
hopefully, we should gain some good market share over the next 2-3 years.

You are saying
that the continuous uptake of PCs is not a surprising thing for Microsoft. So
what is the company’s perspective on these devices? Are you seeing them
picking up much later, even in the US and Europe?

A lot of these devices actually
work best with a PC. I mean, there are a lot of devices which you can call
companion devices. These are around PCs and some of them are standalone as well.
At the end of the day, one would actually sell more PCs because of all these

You would see people using little
more of what I call electronic gadgets. But this doesn’t mean that PC has to
slow down. The PC is adaptive and the PC technology can be used as a whole power
and so we are not only a part of the whole, we are part of the parts as well. So
we will evolve that.

Some years back
Sun talked about Network Computers taking over the computing world, apparently
strongly contended by Microsoft. But if you see today, thin clients are actually
a version of the same Network Computers.

Good for Scott McNealy. This is
his second or third round to sell his device, and what I am hearing is he is
facing the same disaster as he had in the first round. Think about this, the
device needs a Sun server, so how many servers does Sun sell, and how many
PC-based servers of, say, Novell or Win 2000 are sold. I don’t have the exact
data, but if there are half-a-million PCs sold around the world, Sun is lucky to
sell 150,000. So to force a customer to go from a better price-performance
server environment to a Sun server environment just because he can have cheaper
clients, I don’t think that is a viable proportion.

To put the
question differently, if it is thin clients versus PCs, then what is the

We have Windows terminal servers
built into NT as well. But let’s go back in history–we had this type of
computing environment when I started going to university. I grew up on
mainframes and later on Digital Equipment’s time sharing systems. Today, they
are roughly around 0.5 million of these dumb terminals being sold around the
world. And that is not going to go away, but compare this with 110 million PCs
sold. They don’t want to depend on a central administration to tell them what
to do and what not to do. And what if this network is unreliable. We think that
there should be computing on your desktop and there should be computing power on
your server and that it should be balanced. And in a way, people feel empowered
by PCs, think about what you can do with this thing today. I don’t think you
can take this away from anybody. This is just not going to work with end-users.
Sure not in homes or a company environment. I think at the end of the day,
customers will buy the balanced approach.

So to give it a
logical extension, the PC is getting fatter and fatter and so is the operating
system, as well as the applications. Where will all this lead?

I have a way to express this a
little differently. Instead of calling this a personal computer, I call it my
personal mainframe. But then, it is both good and bad. If I want that
multipurpose service then I will accept it. I think the real issue today is of
simplicity of use. We are working to make this better. It’s hard and might
eventually need a very different interface. You are used to a typewriter, to a
keyboard and to icons on a desktop with mouse clicks. So graphic viewers will
have to eventually develop more advanced interfaces with, say, voice type
technology, speech technology and even more in the future to come to a user
interface which gets you out of this nerdish type of environment.

Essentially, are
you saying that appliances are not too much of a challenge to Microsoft?

I am not saying that appliances
are not much of a challenge. I am saying we do have products in that space. We
don’t have a dominant market position in that space today and that’s for two
reasons. First, it is totally a virgin territory and you don’t know what’s
going to be successful. Second, there is lot of entrepreneurship and people have
not totally figured out how to do this. But yes, I am hopeful that we still will
have a good part of this market. And again, it might appear as big as the PC
market. It all just depends on how good we are in developing the operating
systems, and whole tools around it. I still believe that the environment which
we have built is the environment that enables you to get to the market fast. As
these services would probably have product cycles in the neighborhood of one
year compared to a PC life cycle which may be two-and-a-half years or three
years, meaning you need to have tools which are much faster than those you have
in the PC environment.

Among all these
areas, appliances, ASP and others, where does the technology challenge lie from
an OEM perspective?

In appliances we would like to
keep a wait and watch policy. We have better products on the NT side and Windows
2000 and they fit fine. We are also working on a lot of OEMs to get into this
area. Sometimes, it takes time before you see actual results, which is OK.

Most of the vendors today look at
our market very differently. We put software components and hard disk components
in them and then we sell these products. Now for the PC I get $500, for that
thing I might get only $80-90. Mind you it’s the same power. So that actually
brings down my gross margins, and I can’t keep my gross margins down. So I
will have to sell more devices if I have to make money. And the only way to
create money is to source some kind of servicing so you might find the whole
industry for these services to be very different in the future. For us, the
challenge is to work with the company that markets the products or that actually
manufactures the product. And to gain profit, you have to work with both, and we
do that.

What about the
64-bit OS? Will the OEM scene change as you will have all these guys looking at
Linux and SCO Unix?

There too, I will say the better
product will win in the end and that’s the reality. The 64-bit version is
basically written for the Merced processor and the next one, which will come and
we are making efforts to get the products out. Plus, we are also getting some
independent software vendors who have add-on products.

And yeah, we can compete with
Unix, I am not saying we can compete very nicely. Moreover, the PC
manufacturers, the way I know them, they can’t care less about that. There are
very few who really have investments there. And even they, over the last couple
of years, have come along and have been selling NT servers and Windows 2000
servers. It will depend heavily on us to create demand from large corporations
for this type of product and the OEMs will just gobble it up.

But are these
OEMs also not looking at Linux very seriously in the 64-bit?

I believe if 64-bit Linux is
actually done, it has to be a lot of Intel alliance not of OEM side, sure though
it will be there. And again, in the end it depends what’s the better product,
we are actually pretty confident that we can drive successfully there. Again,
you have to have a crystal ball, but what we have seen so far will be the
acceptance of Windows 2000 in the market particularly on the server side.

For example, SGI
is looking at Linux very seriously for its workstations?

I think that company is totally a
bad example, they are probably engaged in nothing but selling of parts for the
whole company. 
So I don’t know what their future might look like. I think probably it’s
fashionable to do that. Will SGI be successful in the market, we will ourselves
find out later.

Now tell us, how
much of Microsoft’s revenue comes from these OEMs today?

It’s around 30%, can be

With change in
technologies, will there also be any change in your general terms of agreements
with OEMs?

I don’t see that. The key for
us, when we work at OEM agreements, is to make sure that our intellectual
property gets protected and an OEM has the right to give licenses to end-users
on our behalf. I don’t think that’s going to change.

Finally, what
percent of its revenues does Microsoft lose to piracy?

Well, if you start with the OEM
business, here we are losing probably $2 billion to piracy. The total company
probably loses $7-8 billion. It can also be as much as $10 billion. It’s hard
to predict. But if you take some numbers out, get some surveys done, it’s in
the neighborhood.

Arun Shankar and
Manisha Singh
in New Delhi

Leave a Reply

Your email address will not be published. Required fields are marked *