MarkL and Selection Bias

People normally think their car is a lot cooler after they buy it than just before they buy it. Cognitive psychologists call this ?Selection Bias?. Mark Lucovsky, Google employee who was the brains behind Hailstorm at Microsoft, demonstrates selection bias in actionin his blog entry (link via Dare). He says, ?I deeply believed that Microsoft knows how to ship software. We know how to build it, test it, localize it, manufacture it, charge lots of $$$ for it, etc. Mark and I talked about this briefly at lunch that day, and I have been thinking about it from time to time since… I am not sure I believe anymore, that Microsoft “knows how to ship software”.

His logic seems to be centered around the fact that API platforms (like Windows or .NET) get revved infrequently, and take a long time to ship at Microsoft — and people like Amazon or Google can ship updates directly to a web site. His selection bias is leading him to compare apples and oranges, and ignore some important points:

  • MSN ships a ton of code directly to websites, messenger clients, and directly into the hands of consumers. SQL Server may take 5 years to ship, but MSN ships every 6 months, at least as efficiently as Google, Yahoo, or Amazon IMO.
  • It’s not about APIs and platforms. Did Hailstorm fail because MSN didn’t know how to ship, or did Hailstorm fail because it was an ill-advised attempt to jam Windows-stye API play on top of the web? With all due respect, I desperately hope that Google wastes a few billion trying to ship Hailstorm. Actually, I hope they don’t, because it would retard the momentum they’ve built in shipping smart tags.

My guess is that the focus on ?platform APIs?shows that Google istrying to diversify her revenues, since it’s clear that some big gorrillas are gunning for a piece of the ad revenues pie. Perhaps they will take on a platform play only as a way to grow the ecosystem of content ultimately paid with ad revenues, but I don’t think that web service APIs are useful for this. Only open standards like RSS accomplish thison the web. ?Platform? style APIs might be useful for defensive lock-in and licensing revenue, but I also think this is a non-starter on the web. I can’t envision a scenario where Google gets big enough to shut out all of the del.icio.us, flickr, and other similar innovations that keep popping up based on wide-open standards. They seem to be testing this theory a lot lately, but it’s not all that successful.

On the other hand, I wouldn’t blame Google for trying. Being utterly dependant on ads revenue is dangerous. Lucovsky is right when he points to the different ship profile for ?platform? pieces like SQL and .NET. But would you rather have a company that has ad revenues *and* platform revenues, or one that just has ad revenues? He presents it like an either/or, but in his mind I bet he is thinking ?crap, we need some of that .NET and SQL Revenue too?. In addition to ad revenues, a company could get revenue from several streams:

  • Service Subscription Revenue: Xbox Live, MSN Premium, etc.
  • Digital Content Transaction Revenue: Music Download, etc.
  • Platform License Revenue: SQL Server, Windows Server, etc.
  • Consumer Products: Office, Windows, etc. (updated through Windows Update, no less), XBox

Google is kind of weak in all of these categories at the same time as they are trying to protect their slice of the ads pie from Yahoo and others. I’m not a strategy guy, so maybe that’s why I don’t see a really obvious strategy for them. It’s either:

  1. A devious and strategic master plan that nobody can figure out yet
  2. Slow but steady growth and moderate diversification, with little room for error
  3. A lot of brick walls

The general consensus on the web is that it’s #1; a brilliant master-plan. ?Google is shipping an OS!?, ?Google will make a web browser!?, ?Google will take over the platform and programming language!?. But I just don’t see it. They’re funded enough, and have enough Microsoft employees now, to be a formidable competitor. So I wouldn’t downplay their importance. But I think it’s a bit silly to be projecting world domination at this point; #2 is perfectly reasonable IMO. And Microsoft will keep shipping software.

Leave a Reply

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