Microsoft has a pending motion in Novell v. Microsoft, asking the judge to dismiss Novell's entire case as a matter of law without going to a second jury trial. Microsoft lawyers list many reasons why, in Microsoft's view, it did not violate antitrust law when Bill Gates
decided [PDF] not to publish certain APIs in 1994 even though it was "late in the day" to make such changes, because: "I have decided that we should not publish these extensions. We should wait until we have a way to do a high level of integration that will be harder for the likes of Notes, Wordperfect to achieve, and which will give Office a real advantage."
That's what the trial, which ended in a mistrial, was largely about, and we're waiting for Novell's response to Microsoft's motion.
Microsoft argues that being sharp in its practices, even being deceptive, isn't an antitrust issue. They had no obligation to do a thing for Novell, even have any dealings with them, because there was no contract and they were a competitor, Microsoft argues.
The first jury didn't accept that, but let's do so for a moment. Imagine that Microsoft is describing the state of the law in the US on antitrust. Let's assume that Microsoft can, indeed, escape responsibility on a technicality or two.
I found an exhibit from an earlier antitrust trial, Comes v. Microsoft, Exhibit #1495 [PDF], that is of interest in this context. It's an email thread involving undocumented Microsoft APIs in 1992, in which an employee named Collins Hemingway urges Microsoft executives to view matters from the standpoint of how things look, not just what the company can get away with. He wrote that sales would be lost if the company followed a "win without compromise" style and so was hated. Heminway wrote, "Don't take advantage just because you can, and act AS IF YOU CARE that there are
potential problems and conflicts. We're in danger of winning the battle and
losing the war."
There is a mention of Novell complaining, among others, about various Microsoft behaviors, and a Marianne Allison, in the same thread, wrote:
Finally I think it relates to the technical, rational minds who comprise Microsoft management. These people tend to relate in a binary way to facts. They have a concept of truth. And a belief in sort of an intellectual meritocracy, that if you get out the best facts, you will be OK, as if there are no filters through which your facts will be viewed. If it is wrong, you just provide so much information that you will convince people you are right. If someone disagrees, they simply need more information.
Microsoft unfortunately didn't alter as suggested, in that there was the next API incident involving Novell in 1993-4, the subject of the litigation. If your read the Findings of Fact (which the jury in Novell v. Microsoft had read to them), you'll understand what all the complaints were about regarding APIs, complaints that resulted in the US v. Microsoft antitrust litigation. And today, judging from Microsoft's super aggressive behavior in the patent wars involving smartphones, I don't think Microsoft ever got the message that while a company might "get away with" things, the public notices and so do partners and possible partners, and that is more true today, in a digital age, than it was in the '90s. Nothing stays hidden any more. And it certainly can, indeed, cost a company sales if people hate a company's behavior. I personally believe that is a factor in Microsoft's pathetic phone sales. To quote from the email thread:
This is best illustrated in a piece of mail sent near the end of the most recent undoc API episode. Lewisl asked a ton of people on an email train, “should we take the remaining undocumented calls out of Excel? We don't need them.” There was silence. My response was in effect, “why WOULDN'T you. To me your products will always be like the East German swimmers. People will believe they are on steroids and that's why they win so much. You should almost voluntarily submit them for “drug testing” to some kind of ombudsman to prove they ARE clean.” To which one of Microsoft's GMs answered, “no, we shouldn't because that would be like admitting we have done something wrong. We just have to convince the industry is's [sic] no big deal.”
I would submit that the industry has voted and MS's position in the industry makes the latter extremely difficult to achieve. At the least, it should not be the foundational objective of a crisis communications plan as we go forward.
Our failure to respond will cost us sales, if not on the issue then on the cumulative impact – IBM is out there saying, How can you trust MS, they cheated on OS/2? Novell is saying, they cheated on WFW. FTC is saying, they may have cheated on MS-DOS. ISVs are saying they cheat on sys/apps. Cumulatively, this
is killing us. Eventually this will sink in at the corporate level. No customer wants to do business with a cheat.
I do want to thank our volunteers for continuing to work on turning the
Comes v. Microsoft exhibits from PDF into text, so they can be keyword-searched. This is an example where a Groklaw volunteer found this PDF and made it visible, so to speak, by doing the work of creating the text and the HTML. I have more time now that I am at least semi-off duty on the articles, so I'm now going back to earlier article comments to look for your work, which is how I came across this exhibit.
[Update: I found another intriguing exhibit, #7374 [PDF], which is a
46-page collection of slides, titled "Windows API Strategy", dated just "September". It's particularly regarding Java and Netscape, but on page 10, it says, "Microsoft API inaccessible from Java -- MS has extended Java in IE3 with COM, SWT extensions, DirectX, but we need to do more." Note page 30: "Microsoft focused on Windows - Performance will be best on windows. New things appear on windows first". The next slide says, under the title Windows Innovation, that in the past, it came from "ongoing flood of new APIs". Does that not seem to match the Bill Gates memo, saying best to wait until it would be harder for Novell? By the way, it on page 3, it defines Java as middleware, which doesn't seem to match Microsoft's current definition of middleware.]
And here's exhibit 1495 as text:
From: Collins Hemingway
To: Brad Silverberg; Cameron Myhrvold; Jonathan Lazarus;
Cc: microsoft!anner; microsoft!garygi; microsoft!martyta; Waggener Group
Subject: RE: MS and crises in confidence
Date: Monday, December 14, 1992 1:41PM
This is good thoughtful mail, though I don't believe that the analogies hold.
The victims in all your examples are, in fact, innocent bystanders. The real
scenario is one in which the “victims” are fellow chemical companies and
airlines. It's very hard to imagine the public would respond positively to BP or
Shell attacking Exxon for spilling oil and saying, “Buy ours instead; we'd never
do that.” Or, if Borland were Exxon, saying they wouldn't have wrecked on the
rocks except that Shell was undercutting them and so they had new choice but to
steam full speed ahead in fog.
I'm being only somewhat facetious, but I do agree with Marianne that we have not
fundamentally stepped up to our industry role, that we think any concession to
outside perception is weak and downright unmanly rather than the best way to
optimize our long-term sales. And I would like us to look at the long-term sales
situation. In particular, I see a subconscious desire to “maneuver in the back
room” and “leave all options open” which are counter to the position of the
leader. (Consider the constraints on the U.S. today – we can attack anybody we
want, and win – yet the worldwide fallout would be worse than any short-term
gain, esp. if we picked on another Grenada.)
I think Jesse Berst made an interesting comment in one of the FTC articles last
week. He said, “It's become politically correct in the PC industry to hate
Microsoft.” That, I believe, is true – from our competitors to hundreds of small
ISVs who don't REALLY compete with MS but are afraid we'll mash them by accident
– the old mouse scurrying when the elephant dances kind of thing. I would like
us to try to address this issue on a global PR basis – this affects the whole
company. I would like to see high-level PR brainstorm – agency, MT, Anne et al –
on what is possible to be done about this: How do you, by actions and PR, change
such a perspective?
I DO believe the “hate MS” mindset this comes from some of the things Marianne
cites – a literal-mindedness about competition, a desire to win w/o compromise,
and a lack of sense of “what it's like out there” for other companies. It
doesn't mean we walk away from sales or shackle our products. It just means:
Don't take advantage just because you can, and act AS IF YOU CARE that there are
potential problems and conflicts. We're in danger of winning the battle and
losing the war.
In the undoc api situation, we so far as a company have not wanted to draw any
lines because we're afraid of the inconvenience or of just screwing up, or maybe
some people really AREN'T willing to give up any possible (however slight)
competitive advantage. Then the company is frustrated and angry that we lose
serious credibility points with the press and other companies for what is a
fairly trivial issue – running apps through some debugger at the end of
development and changing any undoc calls or doc'ing some we don't want to
change. We want the best of both worlds – being seen as fair and yet not being
allowed to be sloppy or doing what we want because we want – and aren't grown up
to recognize that sometimes you don't get what you want w/o making rational
tradeoffs. (Forget the public black eye: I'll bet the company has already spent
more $$ in time/energy in defending its position in the API war than it would
have taken to implement a solid system to prevent the use of undoc APIs. And
we'll get to keep on fighting this problem for months or years if we don't fix
We have to realize this is a business issue. We have to ask: How many sales will
undoc calls get us? My guess is, zero – and that's on the high side. Versus, how
many sales will “being dirty” (however slightly dirty) cost us in corporate
credibility and potential sales? In the short term, probably zero. In the long
term – lots. And how much more wasted time and energy will it take to cope with
public fallout? These are the issues to think about. And we should think through
similar issues the same way
Once you get to a certain size – and MS is clearly way past that size – the warm
and fuzzy attributes become important far beyond the literal facts of a case. MS
going out of its way for ISVs or to assure industry fairness is the perceptual
issue, not how trivial a certain api is.
Our failure to respond will cost us sales, if not on the issue then on the
cumulative impact – IBM is out there saying, How can you trust MS, they cheated
on OS/2? Novell is saying, they cheated on WFW. FTC is saying, they may have
cheated on MS-DOS. ISVs are saying they cheat on sys/apps. Cumulatively,
is killing us. Eventually this will sink in at the corporate level. No customer
wants to do business with a cheat. Regardless of the self-interest of the
accusers, we have not done anything to show that they might be wrong. We stand
accused and our response is “All our accusers are dirty too.” This has roughly
the weight of Nixon's “I am not a crook” speech.
To: microsoft!anner; microsoft!collinsh; microsoft!garygi;
Subject: FW: MS and crisis in confidence
Date: Monday, December 14, 1992 11:05AM
fyi, I had this stuff on my mind so I wrote it down this weekend
From: Marianne Allison
To: MSbradsi; Mscameronm; MSjoni
Cc: MS TEAM EXEC
Subject: MS and crisis in confidence
Date: Mon, Dec, 14, 1992 11:02AM
The recent undocumented API situation give us a chance to reflect back on
Microsoft's role of the industry and how the company is perceived. I am spending
this time because I think this relates to a fundamental Microsoft self-concept
that affects how the company communicates and is perceived. If we understand
this we have a greater choice on how much we wish to respond when something
similar happens again. This isn't a plan but does suggest a way of categorizing
First of all the assumption should be that this was a crisis. We usually think
of a crisis as a “disaster,” where lives are lost, for instance. But crisis is
essentially about loss in confidence precipitated by some event or disclosure.
The assumption is that some public trust and goodwill is essential to do
business; so the real issue for the company in crisis is not so much cleaning up
the actual event/problem as rebuilding confidence and trust.
There are three main things that companies have to communicate in a
1) That they are in control: the company recognizes the problem, is taking
responsibility for it and is taking steps to fix it.
Communicating this is step #1. It does not imply admission of guilt or fault. It
simply means saying, we recognize there is a problem that we are connected with
in some way. We may not even know what happened. But there are victims or
possible victims. We care about them. So we will do what it takes to participate
in fixing it.
This message has to be communicated quickly and broadly and by an empowered
person at the company who is the icon on behalf of the victims. This why the
president flies to Miami after Hurricane Andrew.
This action will never fix the problem. At best it will neutralize the situation
and take some of the power away from people who might benefit from filling the
vacuum created if the company doesn't take responsibility. And it can calm the
situation down. Fundamentally in a crisis, people are worried. They want the
company to say, “we have stepped up to this, we will take leadership.” The
premise is that the company is powerful and often the public feels powerless –
so if the company is not in control, there is chaos.
2) The second thing the company has to do is communicate that it acted
responsibly and in the best interests of public, i.e., that it didn't cheat, lie
or steal to precipitate the crisis.
Communication in this phase is usually oriented around getting lots of
information out about what the company did do or will do, bringing in outside
experts/ombudsmen to verify the company's points (because they are unlikely to
be trusted at this point in the process) and most importantly to communicate an
understanding of the problem SO THAT the company can generate confidence that it
won't happen again.
The company must move quickly during phase 2 to get the facts out but must be
VERY CAREFUL about moving into this phase until they can make assurances or
explanations that are reliable.
3) The third thing the company must do is indicate what will change as a result of
the crisis and prove that it is taking steps to compensate the victims or ensure
it won't happen again.
(There are also times where the crisis the result of a freak accident but the
company still ahs to clean up after the crisis even if can't guarantee it won't
happen again. In this case it still may need to show it will be better prepared
next time. Sometimes it is simple as saying, we apologize.)
During the undocumented API crisis, Microsoft focused on #2. We looked at “what
happened” and gathered lots of info/white papers and submitted them to the
press. The communication objectives driving these actions was proving that there
really wasn't a problem and that Microsoft didn't do anything wrong.
WHY? Because Microsoft did not/would not grasp that the information about the
undocumented APIs was a crisis in confidence. It did not recognize that there
are “victims” or potential victims. A victim is anyone who feels dependent on
the problem being fixed, yet who at the same time, feels powerless to fix it
himself. Apart from altruism, the reason we care about these “victims” is that
in the business context, they are our customers.
In the undocumented API crisis the victims were developers. Regardless of
whether they exploited the situation there was still a potentially injured
party. AND EARLY ON IN THE CRISIS, MICROSOFT DID NOT IN ANY WAY ACKNOWLEDGE THAT
THESE PEOPLE HAD A RIGHT TO FEEL INJURED, THAT IT WAS CONCERN OF THE COMPANY'S
AND THAT THE COMPANY WAS TAKING RESPONSIBILITY FOR ITS ROLE IN THE
Imagine if a train was derailed and a car with an impressive-sounding chemical
name emblazoned on it lay in its side. This substance might be completely inert
and of no danger. An environmental group that has been lobbying the chemical
company to move materials safely actively starts complaining. Neighbors are
worried. There is really nothing wrong, no real damage done. But there is a
In effect, the company says, “there is nothing toxic in that railroad car and if
you think there is you don't know anything about chemistry.” When people say,
“we are worried,” the company says, “don't get hysterical.” When the people say,
“But the environment group says this is a problem.” The company says, “They
have been after us for a long time. They are simply exploiting the situation to
meet their agenda.” When the media say, “why aren't you taking this more
seriously?” the company says, “we would if there was a real problem but there
isn't. Here's a report on the chemical substance. Read it. We will remove the
car, it's no big deal, chill out.”
This is an overstatement and oversimplification but it contains the kernel of
our response. The focus was on facts. This was important – it was part of Step
2. But we essentially had the wrong victim. Our premise was that MICROSOFT was
the victim and being unjustly accused. This shows a fundamental lack of
acceptance of the company's stature and responsibility in the
What if the first thing an airline spokesperson said after a crash was, “well it
wasn't our fault because the weather was bad and we'll prove it to you”? The
absolute number #1 priority is to define the problem from the outside point of
view and to express sympathy and commitment to the affected parties. The fact
that an airline crash will cost an airline riders, is incredibly inconvenient,
etc is the LAST thing it can communicate in a crisis.
We also have not done #3 – still not taken the definitive steps that have teeth
in them to show that there is a change taking place, nor has it in a super
public way said, “We apologize.” I know Claire is working on this with
Why did this happen? I believe it has something to do with attitude and
arrogance. But it also has to do with a naivete about how dependent the
“victims” in this case are on Microsoft and how vulnerable they feel. How
powerful Microsoft is. What it feels to be NOT Microsoft but a company affected
by Microsoft. It is hard for Microsoft to internalize this because the
businesses it is in are so competitive. Saying we are big and powerful seems
like being complacent and no one wants that.
It also has to do with the fact that the “victims” in this case are in a
position to exploit Microsoft in this crisis and in fact do. But how is this
different from many other instances of crisis? Environmentalists can exploit the
Exxon Valdez situation. Does this make it any less horrible that the environment
was completely fouled?
Finally I think it relates to the technical, rational minds who comprise
Microsoft management. These people tend to relate in a binary way to facts. They
have a concept of truth. And a belief in sort of an intellectual meritocracy,
that if you get out the best facts, you will be OK, as if there are no filters
through which your facts will be viewed. If it is wrong, you just provide so
much information that you will convince people you are right. If someone
disagrees, they simply need more information.
This is best illustrated in a piece of mail sent near the end of the most recent
undoc API episode. Lewisl asked a ton of people on an email train, “should we
take the remaining undocumented calls out of Excel? We don't need them.” There
was silence. My response was in effect, “why WOULDN'T you. To me your products
will always be like the East German swimmers. People will believe they are on
steroids and that's why they win so much. You should almost voluntarily submit
them for “drug testing” to some kind of ombudsman to prove they ARE clean.” To
which one of Microsoft's GMs answered, “no, we shouldn't because that would be
like admitting we have done something wrong. We just have to convince the
industry is's no big deal.”
I would submit that the industry has voted and MS's position in the industry
makes the latter extremely difficult to achieve. At the least, it should not be
the foundational objective of a crisis communications plan as we go