Why Geoffrey Moore doesn't get it. Over the summer I read Crossing the Chasm by Geoffrey Moore and Information Rules by Varian and Shapiro back to back. The former is a business book about high-tech marketing, the latter a business-y book about information economics. Reading them together illustrates how poorly Moore understands his industry.

That's not to say Crossing the Chasm has no value. The basic strategy he lays out--first target a niche market until everyone suddenly starts to buy your product--is sound, but Moore doesn't understand why it works. He illustrates this limitation through his remarks on Microsoft: "to the best of my knowledge, Microsoft has never followed the niche strategy that I have been so strongly advocating...Microsoft's history is so unique it makes it virtually unusable as a precedent for strategy decisions in other companies." (p71 1999 paperback).

What Moore never twigged, but Varian and Shapiro understood right away, is the strength of demand side externalities in these sorts of information goods. The value of Word to me increases as more people use Word--it becomes easier to share documents. Therefore if everyone uses Word, it's hard to use anything else. Note that this is not true for a web browser, if everyone else uses IE I can still use Netscape and be OK (a fact Microsoft is furiously trying to change). Moore's niche strategy is nothing more than focusing on a fragmented market and gaining the 30% share needed to kick-start these demand side externalities that will drive market share to 100%. And an unassailable 100% at that.

Also note that other examples Moore wheels out do not work like this. For example, the value of my Palm Pilot increases only marginally if other people also have a Palm Pilot. Likewise, the value of Amazon only increases marginally if others use it (although their book reviews do create some demand side externalities, and undoubtedly they have great economies of scale on the supply side.) Hotmail has weak demand side externalities, eBay's are strong. And so it goes on.

The point is that the technological game is played by exploiting strong demand side externalities to lock up a fragmented market, and then start forcing that market through upgrade cycles by controlling the interface (this is what Microsoft has to do each time they release a new OS or Office application). So long as the market dominance is maintained, this can happen endlessly.

Obviously, this model sucks for consumers. It does not take long until upgrades are driven by the supplier's need for more money, not the customer's needs for a better product. This explains Microsoft's legendary "we've added features that just one person requested" attitude--they're desperate for features they can roll into upgrades. If they cared about actual customer needs they might decide that the increased complexity of single user-requested features actually harms the experience of the program overall for all the users who want something simpler. This model also drives complexity in technology, which is why development and productivity gains have been so stagnant. Eventually, firms are going to figure out that sometimes simpler, cheaper technology suits their needs better than complicated bloatware designed to justify yet another upgrade. But consumers share in the culpability. By being so passive in their technology choices, they've positioned themselves as the spoils of war, so that's how they're treated.

Open source software and open standards have a particular advantage in the standards control war--they're the one thing all players in a fragmented market can agree on. To prevent a dominant player from locking-in, all other market participants could support an open standard (and so forfeit market lock-in themselves). The Apache server and the FreeBSD software at its core pretty much cemented the TCP/IP stack's role as network standard of choice--it's a platform without a platform vendor.

Microsoft has two strategies to continue to capitalize on its desktop lock-in. First of all, Microsoft's biggest competition on the desktop is older versions of Microsoft. Leasing policy, backward (but not forward) compatibility, retiring support for old OSes etc. all eliminate the market for old Microsoft products and make room for new ones while protecting the lock-in. Secondly, Microsoft targets new markets by essentially merging them with its existing monopoly position, i.e. it ties the software into its desktop OS. They've done this with the word processor, spreadsheet, database, browser, media player, photo editor, text editor, server (in XP) etc. etc. It does not matter if they instantly commoditize the product (which for information goods means its price falls to zero) so long as it supports the central platform which the company can still charge for. Given that this is how information goods work, the court's hesitation in the Microsoft case because the company had not grown through acquisition makes no sense.

Moore would have written a better book, and no doubt delivered better service to some of his clients, if the understood the economics behind his advice.
Link to this article