their likely Vista adoption plans. Although the path of operating system (OS) upgrades is relatively
well understood, a bit of data never hurts to ratify the theory.
The theory, borne out by every OS introduction since Windows 95 (an exception; everybody
adopted it quite quickly), says that consumers get it when it's available because they mostly buy
what's on the shelf in stores. In recent times, retail has turned its product lines more rapidly to
compete with direct, notably Dell, which turned its entire line in about three months with Windows
XP. Enterprise (really, large and midsize businesses, government, educational institutions) tends
to be cautious and wait until the OS stabilizes before adopting. This broad segment could also be
characterized by the term "relationship customer," but the key characteristic of these firms is that
they have a knowledgeable IT department, which makes decisions for the company as a whole.
The others, consumers and small businesses, make their own IT decisions and could be called
"transactional"; that is, they buy one machine, or at most a few, at a time and are less constrained
by issues of compatibility and image management.
So, the short of it is, transactional customers get it fairly quickly and relationship customers adopt
over time.
Now, this basic model has to be superimposed over a seasonal pattern. Vista will be introduced
this week for relationship customers and in January 2007 for transactional customers. This
timing probably couldn't be any worse. The relationship customers, who have the opportunity to
buy first, won't be picking it up in volume for a number of quarters as they kick the tires, and the
transactional customers will get their opportunity to buy right after the high season, when their
interest is at ebb tide. Vista will be in the stores in January, but consumers won't show up until
back-to-school 2007, which is really August.
Straw Poll
The results of an informal survey of a few enterprise IT managers revealed a range of opinion, but
a key finding was that even the near-term adopters will begin rollouts only in 3Q07 or 4Q07, while
the bulk will wait for 2008 or Service Pack 2 (SP2). Answers to the question, "What are your plans
for Windows Vista?" are set out in the following paragraphs.
CIO
Major Hospital Complex
12,000 seats
We are focused on reliability and security. Although Microsoft notes Vista will offer enhanced
security, our experience is that installation of new, complex software leads to increased security
risk and less reliability in the short term.
As a matter of policy, we do not install Microsoft products until they have had a solid track record in
the marketplace. That usually means Service Pack 2 of any application or operating system.
We have installed Vista in our lab so that IT folks can learn about it, but we have no plans to install
it for at least a year.
CIO
Major Online Brokerage
10,000 Seats
No plans at this time. Presently we see no business value in deploying Vista given the cost and
impact to our operation. Of course, things change but our dependency on Windows in general is
minimal outside of basic office automation.
CIO
Upstream Oil Company
9,200 Seats
Refresh to begin end 2007 with all HW, OS, and apps migrated at once. At the same time, the
notebook ratio will rise to 60% from 45%. Last refresh was 2002 move to XP. Clients have lasted
2 years longer than anticipated. Most geological & & geophysics apps are in Windows. New life
cycle will be 4 years.
Network and Data Center Engineer
Large Basic Technology Supplier
9,000 seats
Our manager for Windows systems said he'll have his folks looking at it, but with no plans
whatsoever in the short term. We will not be on the ramp to an early adoption.
Desktop Technical Architect
Major Food Processor
28,000 seats
18 months out.
VP of IT
Large Real Estate Development and Gaming Company
5,300 seats
Vista in 2008.
Summary of Interviews
You get the idea. Six quarters out is when most of these companies are going to be adopting
Vista. The "no business value" guy can be explained by the fact that his firm has more servers
than clients, since most customers connect over the Web through a browser. His intention is to
shift the server plant to Linux over time. So, most of these firms will adopt Vista, but virtually all of
them groused about Microsoft as a partner, expressing mixed feelings about the software and
something akin to anger at the price. Several seemed to be afraid to criticize the Redmond
behemoth for fear of some sort of retribution, shades of the "old" Microsoft.
Forecast
So, where does that leave us? Endpoint recently did a rough forecast of Vista adoption for 2007
for Business Week. A back-of-the-envelope model gave about 76 million licenses for the year.
However, a number of requests have come in to refine this model, and so here is the more
considered version. First, the assumption for 2007 PC market growth over 2006 is 7%, below
some estimates but realistic and conservative. Mapping quarterly PC shipment estimates
against a proportional adoption picture by broad segment, the scenario goes essentially like this:
transactional customers will find Vista on their systems in retail and online pretty quickly after
launch. By assuming that, for these buyers, 25% will have Vista in 1Q07, 50% in 2Q07, 75% in
3Q07, and 100% 4Q07, the model accounts for old systems in the channel and any laggards
among the OEMs in the transactional space. Otherwise, by the end of 2007, it is assumed that all
systems sold to these segments will be based on Vista. Relationship customers will be much
slower to adopt (see the previous section). For this group, the quarterly adoption figures are more
like 5%, 10%, 15%, and 20%. Thus, the full year Vista license shipment figure comes in around
82 million, a bit higher than the Business Week estimate (Figure 1).
Figure 1
(in millions of units)
Further Out
Although 82 million units would make most companies happy, Microsoft has a big appetite and
might view these figures as less than stellar. The company can take comfort, then, in the fact
that by 2008, Vista adoption will be in full flood, defined as half of all shipments going out the
door loaded with the new OS. It is beyond the scope of this piece to argue all the reasons why or
why not to adopt the latest Microsoft product. Vista is a much better OS than XP, and it maintains
the ecosystem so important to the 60% of the market represented by relationship customers.
Vista is more secure and robust and better performing and looking than XP, and it has some
surprising features (e.g., integrated search, image management tools) that will please potential
customers. It is really not if, but when, and Microsoft's poor launch timing belies the fact that over
the next 2-3 years, hundreds of millions of copies will ship worldwide.
Roger L. Kay is the founder and president of Endpoint Technologies Associates
(www.ndpta.com).
© 2006 Endpoint Technologies Associates, Inc. All rights reserved.
