How Microsoft kills cool projects
There’s a good article detailing the death of Microsoft Courier, a tablet device project from 2009/10 led by J. Allard, of XBOX fame. The core story rests on this observation:
Within a few weeks, Courier was cancelled because the product didn’t clearly align with the company’s Windows and Office franchises, according to sources. A few months after that, both Allard and Bach announced plans to leave Microsoft, though both executives have said their decisions to move on were unrelated to the Courier cancellation.
Most interesting products for today’s world can not easily align with business models created in 1995. I know many smart people who had great prototypes for new products while at Microsoft, who were saddened to learn the escape velocity of a project is, at minimum, greater than the gravity of its two largest businesses (Office & Windows). They’d watch with sad eyes as their well conceived plans were smashed to pieces against the massively successful, but ultimately boring, twin leviathans of Microsoft. The details of the Courier story, a well designed product, fully staffed with 100+ creative employees, is sad indeed. A very different future for Microsoft was ready to born, but never saw the light of day (photos and demos).
During the browser wars, a similar, but rarely told, story explains why IE4 was the pinnacle of browser innovation in 1997, and then took a right turn into stagnation.
Brad Silverberg, VP of internet things circa 1997, intended for the web to replace Windows. He wanted Microsoft to make the web a platform, and launched versions of IE on Mac and IE on Unix (to the dismay of the industry. It’s the only UNIX application Microsoft has released). The idea was to leave OS’es behind, and focus on the web as the core way people will interact with computers. A prophetically Googlean strategy.
But when it came time for Gates to make the call, Jim Alchin, the VP of Windows, won. Windows was more important. As a result, after IE4 (and the implosion of Netscape), plans for making the web the future platform for the company were shut down, in favor of protecting the Windows franchise.
Many lament these choices. It seems boring to continually protect the status quo. But when your status quo generates $60 billion annually, a rate of income only a handful of companies in history have achieved, few complainers would have the courage to act differently if they were in charge.
Surprising, but it really isn’t? People, hence organizations act according to a script usually:
” I don’t have much, so I can’t act differently. (Some time later) I have so much, why should I act differently?”
Microsoft or anyone with 60B$ can bet on something new, as long as they can ensure in the process they do not stop milking the current cow. If the new demands the current cow to be killed, then its correct that one has something to ponder upon?
The cliche business advice is to kill your own business before someone else does. Its rare to see anyone practice it, but as I pointed out, Silverberg was on to what Google is doing in 1997. He wanted to move on to a new landscape ahead of the industry, rather than trailing it.
Historically first movers don’t do all that well, but most people find this more heroic and inspiring despite the typical outcomes.
Good read, Scott.
I’m with Deepak. Not surprising. It’s The Innovator’s Dilemma in action. The key is to keep cannibalizing your business model; to morph your platform. Easier said than done, especially for a large public company.
Just a quick note to say that IE5 for Solaris was, at the time, by far my favorite browser on any platform. I loved having full event handlers and MSXML in a browser that would use my $EDITOR when I selected View Source. For ages, I ran an X server on my Mac just so I could export the IE5 display from a Sun machine for web browsing.
Sounds an awful lot like how Oracle works, too. Glad I left that place.
The article and your analysis are great examples of confirmation bias in action.
Ok Bruce – happy to hear how *you* think Microsoft kills cool projects :)
Jokes aside, with some effort I can name at least 10 projects that fit this profile. I think that’s sufficient to be more than cherry picking.
With XBOX, Windows 7, and Mango, there is a story for how Microsoft ships *cool* projects, but that thread doesn’t negate the reality of this one.
My point is that you are predisposed to believe the Courier was a “cool project” killed solely because of corporate infighting, executive power plays, lack of alignment with dinosaur-era businesses from 1995 – “invalid” reasons perhaps. The story on CNET has you nodding along in agreement, remembering all the times past when you felt the same thing had happened to other “cool projects”.
I very doubt if you put much thought into considering if Courier was killed for what you would consider valid reasons. Such as, say, having a fatally flawed, myopic vision.
Actually, after reading the article I agree with Bruce a bit… the scope that Allard had in mind for the Courier was very narrow.
For example, it should have been designed with e-mail in mind (Not unlike the iPad… :-) )
David Carrington noted that:
“MSFT released other apps for Unix: LAN Manager for AT&T Unix (yeah, that one was weird), Services For Unix (thank you, Interix) and MS Mail on Xenix. And, of course Xenix itself.”
there was also Word on Xenix, which was supported by SCO – so not sure quite how you count it, but it was a Microsoft product on UNIX ;-)
Beyond old vs. new, Allard vs. Sinofsky, the trouble with Courier is that the market for content creators is much smaller than the market for content consumers. Some empirical estimates say 1:10 ratio. I am convinced Courier was a great “magical” product for creators and that J is a visionary, but would it have turned into a billion $ business?
Silverberg’s intuition in 1997 was correct re: Internet platform, but no one has figured out how to make money from cloud apps at the scale of Office. Google’s main revenue is from advertising and the main reason Google Apps exist is to weaken Microsoft’s revenue not to be the core business through licensing. How much money would Microsoft’s cloud apps be making 15 years later if Silverberg in 1997 built IE into a ubiquitous browser like Chrome is today.
Intuitively we all know the Windows and Office product lines as we know them today have an end of life. But it’s so hard to predict when. 15 years have gone by since the moment you mention and they’re still strong. Sinofsky’s strategy is to transform them very very slowly.
I’m all about change and disruption, but sadly you can’t argue with $60 billion a year.