decoration decoration
Stories

GROKLAW
When you want to know more...
decoration
For layout only
Home
Archives
Site Map
Search
About Groklaw
Awards
Legal Research
Timelines
ApplevSamsung
ApplevSamsung p.2
ArchiveExplorer
Autozone
Bilski
Cases
Cast: Lawyers
Comes v. MS
Contracts/Documents
Courts
DRM
Gordon v MS
GPL
Grokdoc
HTML How To
IPI v RH
IV v. Google
Legal Docs
Lodsys
MS Litigations
MSvB&N
News Picks
Novell v. MS
Novell-MS Deal
ODF/OOXML
OOXML Appeals
OraclevGoogle
Patents
ProjectMonterey
Psystar
Quote Database
Red Hat v SCO
Salus Book
SCEA v Hotz
SCO Appeals
SCO Bankruptcy
SCO Financials
SCO Overview
SCO v IBM
SCO v Novell
SCO:Soup2Nuts
SCOsource
Sean Daly
Software Patents
Switch to Linux
Transcripts
Unix Books

Gear

Groklaw Gear

Click here to send an email to the editor of this weblog.


You won't find me on Facebook


Donate

Donate Paypal


No Legal Advice

The information on Groklaw is not intended to constitute legal advice. While Mark is a lawyer and he has asked other lawyers and law students to contribute articles, all of these articles are offered to help educate, not to provide specific legal advice. They are not your lawyers.

Here's Groklaw's comments policy.


What's New

STORIES
No new stories

COMMENTS last 48 hrs
No new comments


Sponsors

Hosting:
hosted by ibiblio

On servers donated to ibiblio by AMD.

Webmaster
Comes 2203 (Life in Wartime '95) | 131 comments | Create New Account
Comments belong to whoever posts them. Please notify us of inappropriate comments.
Comes 2203 (Life in Wartime '95)
Authored by: Anonymous on Sunday, May 05 2013 @ 06:33 PM EDT
http://groklawstatic.ibiblio.org/pdf/Comes-2203.pdf

<p>
PLAINTIFF'S EXHIBIT 2203<br />
Comes v. Microsoft
</p>

<p>
<b>To: Billg, Bradsi, Jimall, Paulma</b><br />
<b>From: Jonro, Richt</b><br />
<b>Re: Life in Wartime '95</b>
</p>

<p>
Over the last few months there has been a great deal of discussion between Bill,
Steve, Paul, Brad, Jim,
Bradc, etc on how to effectively respond to the competitive onslaught,
effectively leverage our strengths,
optimize Windows by creating a new revenue stream, and possibly change our
product and marketing
paridigms by taking advantage of our broad range of products and the emergence
of CD ROM as a
commercially popular technology. Outlined below is our iteration on this
thinking.
</p>

<h2>Life in trench warfare</h2>
<p>
The suite wars are close to making the applications business a commodity
business. Our API, UI, and
Object standards are being threatened. Competitors are actively trying to split
our standard. In addition,
operating system and application revenue growth will drop to approximately 10%.
Finally, we have a
large number of products who are individually not making money. This is the
equivalent of World War I
trench warfare. The result, if we are not careful can be stalemate (e.g.,
cost-plus pricing, commoditization
and stagnating markets).
</p>

<h2>Inventing the tank</h2>
<p>
But, there are some important technical, distribution changes coming that play
to Microsoft strengths.
Understanding the implications of these changes is really important to inventing
our equivalent of the
"land dreadnought," the tank.
</p>
<ul>
<li>CD ROM technology, which allows us to distribute our wide range of
solutions, making it easier for
customers to purchase products. Rather than having to activate customer to go to
a retail store,
purchase a product, find the time to "flip floppies for 30 minutes" to
install the product. Instead, with
a CD, our customers can just unlock software from a CD and lower the install
time dramatically.
Also, updating software is much easier with a CD since mailing a $1.50 CD can be
done very
frequently since it is so light. With 600MB, we can put 98% of our revenue
generating products on a
single CD.</li>
<li>OEM distribution. If the CD comes with their computer when they
purchase it, it is entirely possible,
they never have to go to the store. All their software can be fulfilled in the
comfort of their homes.
From our studies, it appears that 80% of users buy their software or change
operating systems at the
time of machine purchase. This means, we can make it much more convenient if we
put "locked"
software in a CD and place it in the CD drive itself <i>before</i>
the customer buys the computer.</li>
<li>Trial is expensive and getting more so. When we worked on the
applications business, the cost of an
additional trial was $20 and rising. As our base of customers expand, this will
get more expensive,
not less since we are moving beyond the easy to identify enthusiast
base.</li>
<li>Operating system modularity. It used to be that an operating system
was just a kernel, drivers and a
couple of utilities which usually fit on a single floppy and were monolithic.
With our DLL
architecture, multitasking, it is far easier for us to split up operating system
functions like RAS or
Mail, put them in separate modules and sell them separately. For customers, they
buy what they
need.</li>
<li>Full range and modulatiry of Microsoft products. We have the broadest
product line. We literally can
provide software from cradle (MS Kids) to grave (Server products). Moreover,
with OLE2 and other
technology, it is going to get increasingly hard to tell the difference between
the applications,
operating systems, etc. (It looks more and more like the somewhat arbitrary
divisions we have on
\products1release where it is hard to know if the MS Fonts are in sys or
apps).</li>
<li>Microsoft Select licensing. We have separated the licensing of the
product from the distribution of the
bits. This also potentially gives us a revenue stream (combined with CD insights
above) that is more
like an annuity. For customers, this makes it very convenient since they just
pay a small monthly fee,
rather than big lump sums every couple of years. Americans in particular are
trained to accept this
kind of purchase behaviour.</li>
</ul>

<h2>Life in Wartime '95</h2>
<p>
The challenge is to change our product and distribution model to leverage each
of these strengths and
maximize revenue. Let's see how this would work by mid calendar year 1995. We
pick this timeframe
because it is the first point where our operating systems are synchronized in
their user interface. Rather
than get confused about our current product quirks, let's see what our end point
should be. This will help
us understand where our product needs to evolve. Then, we will zoom back to mid
94 to see how we can
roll this out.
</p>

<h3>Distribution Channels '95</h3>
<p>
Let's see how what each one of our channel will want to sell in this timeframe.
This will tell us alot about
the products that we will need to build.
</p>

<h4>Software Retail</h4>
<p>
At retail, the primary product that resellers will want is something that
upgrades or adds new
software to existing machines (by definition). With most machines being floppy
only, this means,
they will primarily want an floppy products of applications or of operating
system upgrades
(again, by definition, the customer already owns a computer that already has a
operating system).
</p>
<p>
For a large percentage of these customers, they will want a Chicago-based
product since they
won't have the hardware to run Windows NT. Thus, the primary thing, we should
sell are floppy
upgrades.
</p>

<h4>OEM</h4>
<p>
The primary product that they will want will be CDs. By this timeframe, most
machines should
have CDs. This is much cheaper for them (since the backup floppies that
customers want will
cost $20 even for Chicago in that timeframe or about $20 vs. $1.50 for a single
CD). And, of
course, CD cost per megabyte is much lower than preinstalling on a hard disk.
</p>
<p>
Most of the machines in this timeframe will be NT-capable. That means, they
would probably
want a product that could install either Chicago or Windows NT.
</p>
<p>
There is still lots of space left on the CD. In the current Windows NT CD, only
60MB is actually
used. With all this extra space, we can put lots of additional Microsoft
software on it. This gives
them a revenue upside (either their resellers take margin, or the direct
sellers, take it for
themselves).
</p>
<p>
Ideally, an OEM would have two CDs. One would always have the Microsoft CD in
it, the other
would be for multimedia products. That would mean that a customer would always
have access to
the CD with Microsoft software<sup>1</sup>
</p>

<h4>Solution Provider/VARs</h4>
<p>
What will these people want? There are two cases. There are those who administer
and
implement prepackaged software. They will want the same CD as the OEMs for
machines that
don't come with the CD from the OEM.
</p>
<p>
The second case are people who do their own development. In this case, they will
want to blow a
custom CD that merges their software with ours. Custom blowing of CDs is
inexpensive. Setup
costs breakeven today is as 1,000 CDs which results in an average cost of goods
of $10 or so.
There are some licensing implications here, since customers have two ways to get
the same
Microsoft objects. In this timeframe, we will have the licensing APIs which will
let the reseller
track what is actually used. This continues the trend of focusing the channel on
value-added
services, not on software distribution.
</p>

<p>
<sup>1</sup> Technically, by this timeframe, we could implement a
hard disk cache that would let you keep 95% of the software on the CD. Then we
would keep the most recently used files cached on the hard disk. We haven't done
the technical analysis, but I bet that we would get access
times that would be 98% of that of the hard disk.
</p>

<h4>The Matrix</h4>
<p>
So, we actually have a pretty simple matrix for products sold in the 80% case.
Naturally,
Retailers will want to sell CDs, etc., but it's helpful to look at the mainline
SKUs.
</p>

<table border="1">
<tr><th></th><th>Retail</th><th>VAR</th&g
t;<th>OEM</th></tr>
<tr><td>Floppy</td><td>Floppy Chicage
upgrades</td><td>X</td><td>X</td></tr>
<tr><td>CD</td><td>X</td><td>CD with Chicago
and Windows NT on it.</td><td>CD with Chicago and Windows NT on
it.</td></tr>
</table>

<h4>What to do with empty space on CDs? The Silver and Gold
Editions</h4>
<p>
Even in this timeframe, it is hard to imagine both Chicago and Cairo taking more
than 20% of the CD
(e.g., 125MB). With licensing separate from distribution, we should just fill
the rest of the CD with our
locked products. This could include all accessories (Fonts, etc.), all MS Office
products (Mail, Word,
Excel, etc). In fact, when we did a simple disk usage on \products1re1ease, we
found we can get 98% of
our current revenue generators on a single CD.
</p>
<p>
In 1995, Cairo synchronizes the UI and PnP Support, so we can combine Chicago
and Cairo on one CD.
We could create one Desktop version of Windows. The enhanced security and
scalability of Windows NT
simply become modes in this new version. This is the orginal steveb thinking
about NT mode in
Windows. It is implementable now because we can put it on one CD rather than on
40 floppies.
Customers would simply have to consider the functional difference between the
products. This is
analogous to what we did with Windows 3.0 when we combined Windows 286 and
Windows 386 modes
into one product.
</p>
<p>
This would leave us with two major product editions:
</p>
<ul>
<li>Silver Edition. <sup>2</sup> This would be our base set of
bits that we provide as the standard OEM product on
CD. It would include the upgrade to the Gold Edition (see below) as locked bits.
A customer could
upgrade to any portion of the Gold Edition by paying some money and unlocking
bits. Or, we could
just have them buy the whole Gold Edition. This Silver Edition would be the
complete API set. If some
Office components become part of that base OLE API set, they would go on this
CD. (e.g., if we decide
the standard text processing object in Winword becomes a standard API for
developers, it would go
into this Silver Edition).</li>
<li>Gold Edition. This would include everything that we could stuff onto a
single CD. For instance,
putting the "premium Chicago/Cairo bits" (such as preemptive
multitasking, server components, etc.).
Also, we could then place the MS Office, the major MS Home components (MS Money
for instance).</li>
</ul>

<p>
This leaves us with the matrix below:
</p>

<table border="1">
<tr><th></th><th>Retail</th><th>VAR</th&g
t;<th>OEM</th></tr>
<tr><td>Floppy</td><td><ul><li>Floppy
upgrade to Chicago_Gold bits</li><li>Floppy versions of Office,
Home, etc.</li></ul></td>
<td>X</td><td>X</td></tr>
<tr><td>CD</td><td>X</td><td><ul><l
i>Gold CD for machines without the Silver CD.</li><li>Unlocks for
the locked bits on the Silver CD.</li></ul></td>
<td><ul><li>Silver CD as the base
product</li><li>Strong incentives to buy the Gold
CD.</li></ul></td></tr>
</table>

<p>
<sup>2</sup> We think that Silver and Gold are better, then our
current strategy of "No-name" Edition vs. "Pro Edition".
This solve four problems, the
generic name of the two products is simple (right now there is an ambitguity, if
I say the "Office", so I mean the standard Office or the
standard and Pro Office generically). Second advantage, you don't need a clumsy
identifier in your package. For instance, you can just
change the color of the product rather than having to always stick a long suffix
at the end. Third, localization&ndash;around the world, silver is
cheaper than gold and the big consumer marketing companies spend billions
driving home the difference. Finally, it is extensible across the
range of Microsoft offering. Home Pro is somewhat strange for instance.
</p>

<h4>What about if can't fit it all on the same CD?</h4>
<p>
It is pretty unlikely we can get all server products, all home products, all
office products into 600MB.
Moreover, the market for PCs is differentiating along market boundaries. We
already have a strong
division between server and personal machines. And, an emerging division between
office and home
personal machines. Compaq has introduced a special home PC for instance and they
market a high-end
desktop line (Deskpro) and low-end desktop line (Prolinea).
</p>
<p>
So, we should separate our CD products to match the market divisions that the
big players are spending
zillions of dollars to create. This would leave us with four major product lines
and a Silver Gold Edition of
each. Two would be for end-users. One would be for developers. Finally, we would
have a server product
line that would let you install Office Desktop and our line of server
applications. It also lets us decide
somewhat late in the game what we put on Silver goes on Gold.
</p>
<p>
This would be our implementation of "Master Branding" Rather than just
marketing the products as a
single set, we actually package them as a single set at the OEM and VAR level.
This gets addresses the
marketing challenges of distribution of the bits for our low-volume products. It
also makes trial of these
products much cheaper. Finally, it allows us to provide a consistent naming and
positioning model
accross each product grouping.
</p>

<table border="1">
<tr><th></th><th>Silver</th><th>Gold (all
Silver plus)</th></tr>
<tr><td>Home Desktop</td><td>Base Chicago/Cairo, Works,
Money, base Encarta</td><td>Premium Chicago/Cairo, Office, premium
Encarta</td></tr>
<tr><td>Office Desktop</td><td>Base Chicago/Cairo, Base
Office</td><td>Premium Chicago/Cairo, Office Shell, MS Mail, Ms
Project</td></tr>
<tr><td>Developer Desktop</td><td>Base Chicago/Cairo,
VB, Visual C</td><td>Premium Chicago/Cairo, VB Professional, VC++
Professional, MS Fortran</td></tr>
<tr><td>Server</td><td>Base Chicago/Cairo tuned for file
and print service. Would also include a over-the-net installer for the Office
and Developer desktops.</td>
<td>Premium Chicago/Cairo, SNA Server, SQL Server, EMS, Hermes, the
gold editions of Office and Developer with over-net
installer.</td></tr>
</table>

<h4>The subscription business '95, '96, '97,...</h4>
<p>
The next step is to add a six month or yearly subscription component to the Gold
edition of each of these
offerings. This moves us to a model that is very similar to the auto industry
where we would have a
1995, 1996, 1997, edition of each component. Users would simply sign a select
agreement and never have
to worry about software purchase again. The agreements could be administered by
Microsoft directly or
by the channel.
</p>
<p>
Basically, customers would get used to putting this special CD into the MS CD
bay.
</p>

<h2>Competitive implications</h2>
<p>
The biggest thing that we can do for customers is reduce the cost of trial and
simplify the long arduous
process for getting our software. This does have some competitive implications
that are quite nice for us.
</p>

<h3>Silver Server Edition includes full NetWare 3.11 server</h3>
<p>
Novell is currently cloning our base MS-DOS and Windows products. They also want
to prevent the trial
of our server operating systems. They want to do this by denying us client
software and also keeping their
server technology closed and available only through special resellers.
</p>

<p>
The above model changes the way we do business dramatically. In Silver CD, we
would include all of our
upcoming NCP server software and utilities. This is the basic cost of entry into
the networking business
currently dominated by the installed base of NetWare servers. For Netware 3.11
functionality customers
will be able to buy our Silver Server. For advance functionality we will upgrade
customers to the Gold
Server (NT AS).
</p>

<p>
This does have very dramatic implications for our product strategy and what we
need to do to increase the
ease-of-installation of our products assuming that servers will basically get
placed into a sea of NetWare
3.11 servers.
</p>

<h3>Silver Desktops include Cairo</h3>
<p>
Today, we have many companies in the business of cloning our base MS-DOS and
Win16 APIs. This
includes IBM, Sun and Novell. They want to split the operating standard before
we move onto our object-
oriented interfaces. Although most customers won't be able to use Cairo on
installed machines, this won't
be the case of new machines.
</p>

<p>
They should be able to get the full Cairo when they want to use it. We should
include enough of Cairo to
get the full API implementation on the base product. This will help ISVs since
they can be <i>guanranteed</i>
that these DLLs will be on the run rate of machines. We should also make it easy
for a Cairo appliaition
call to install flip from Chicago mode to Cairo mode and load the appropriate
APIs and go.
</p>

<h4>Rules of warfare for '95 and beyond</h4>
<p>
The nice thing about the model above, is that it should be generally extensible
through the next century.
For the next update of products, we would have the following rules:
</p>
<ol>
<li>Decide what new operating system features go into the new base
product. That is, all the APIs and UI
elements we want to make de facto standards. This should go into the Silver
Edition of all products.</li>
<li>As new market segments for computers systems emerge, add a specfic set
of editions for them. For
instance, if Winpad style machine stake off, we would offer a Silver Winpad and
a Gold Winpad. The
most likely case here is the widespread acceptance of the MiniCD (3.5"
format with 180MB data). Any
products that can include 100MB storage at a cost per &lt;$0.01/MB will
work.</li>
<li>As we continue to fight a brutal competitive key things that used to
be in this year's Gold will go into
next year's Silver. (Somewhat like the game that the car manufacturers have
played with Antilock brakes,
Airbags, Sunroofs, automatic transmissions).</li>
<li>Update the CDs on a regular basis. This is an easy way to get minor
upgrades out. And, to ensure an
annuity since customer really like the idea that it is easy to get the latest
bits.</li>
</ol>

<h2>What to Do in '94 Warfare...</h2>
<p>
Unfortunately, we won't be able to implement alot of the above for product and
organizational reasons
right away. The key is to start at the time of Chicago launch in mid 94 with our
operating system products
only. We can divide Chicago up into a couple of different product offerings. We
can essentially create a
base and premium Chicago which would result in the following offerings:
</p>

<p>
1) Silver Chicago CD. In the Silver Edition, we would put the following Chicago
features: New GUI,
Windows and MS-DOS app support, task-switching kernel, Plug and Play support,
small font set, OLE 2
support.<br />
1a) Silver Chicago Floppies. This will hopefully be a low volume OEM-only SKU,
but there will still be
some OEM machines without a CD-ROM driver. Sold primarily through MED and
SVED.<br />
2) Gold Chicago Floppy Upgrade for Windows 3.x users. This would be the product
that we would sell
through software resellers. This would upgrade the existing Windows 3.1/WFWG
3.11 machines to the
Gold Edition of Chicago. In addition, It would be all Silver Chicago bits plus
extended GUI, multitasking
kernel, MAPI/TAPI, mail client, fax transport, network client, peer
server&mdash;might want to put this in
separate package&mdash;additional fonts, accessories, Winpad PIM,
Doublespace, on-line service client.<br />
3) Upgrading Silver Chicago users to Gold Chicago users. For those OEM products
with the Silver CD,
this is just an unlock. For those with floppy Silver, we sell them #2.
</p>

<p>
The implication of this is that we do not try to create an Add-on businms.
Instead we include the
additional networking and mobile functionality into the premium editions of the
products. This keeps us
focused with simple offerings.
</p>

<h3>Introducing the Server Edition '94</h3>
<p>
Unfortunately, we can't yet introduce a merged desktop edition of Windows NT
1.0a and Chicago because
the UIs are so different. However, this should be much less of an issue on the
server side. So, we would
introduce the merged server edition in mid '94, We would have these products:
</p>

<ol>
<li>Silver Server CD. This would have a integrated setup that would
install Chicago or Windows NT 1.0a
depending on the size of the machine, etc. It would also install Chicago and
Windows NT 1.0a optimized
for file and application service (for instance, big net caches, etc.). It would
include locked addiitonal Gold
features.</li>
<li>Gold Server CD. This would have the Silver Edition above and include
our server applications like
SNA Server, Hermes, EMS, and SQL Server. We would also include the Chicago and
Windows NT
desktop installer could load Chicago and NT from the Server.</li>
</ol>

<h2>Product Implications for Chicago and NT '94</h2>
<p>
UI: The Chicago and NT UI need to be as similar as possible. Specifically, the
shell and common
controls (sliders, buttons, etc) DLLs need to be made available on NT. It also
means that the Chicago era
UI needs to be very similar to the Cairo era UI, and that the Cairo era UI be
implemented on both Chicago
and NT at the same time.
</p>

<p>
API: Elegant response when a functional call is made to a non supported API.
</p>

<p>
Drivers: principle issue here is how to resolve plug and play issues. The
registry between NT and
Chicago should be merged, otherwise the dual boot scenerio is very problematic.
</p>

<h2>Revenue Model for '94-96</h2>
<p>
This still needs to be done. However, a back of the envelope analysis is that we
should set the pricing for
our base Chicago/Cairo components between current NT OEM pricing ($200) and
MS-DOS+Window
spricing ($40). So as to achieve a small gain in average royalties on Silver,
Gold pricing should get tuned
to increase our revenue per desktop to meet our 20% per year gains.
</p>

[ Reply to This | Parent | # ]

Groklaw © Copyright 2003-2013 Pamela Jones.
All trademarks and copyrights on this page are owned by their respective owners.
Comments are owned by the individual posters.

PJ's articles are licensed under a Creative Commons License. ( Details )