Things that interest me.

Hal Varian, who is generally someone worth listening to, has a post up on the Google Blog about why he thinks his company (yes, he works for Google) continues to dominate the market. After dismissing economies of scale, lock-in, and network effects, he goes on to say:

The answer, at least in my opinion, is a much older economic concept called “learning by doing” that was first formalized by Nobel Laureate Kenneth Arrow back in 1962. It refers to the widely-observed phenomenon that the longer a company has been doing something, the better it gets at doing it.

This may, indeed, be a part of Google’s success. I think a lot of technologists think that they can come up with a golden bullet that will lead to Google’s demise as a market leader. No doubt they think this because Google had its own golden bullet, the use of linkage analysis to filter results. But it would be a mistake to think that this particular approach to search filtering is what defines Google as a search engine or as a company.

First, while Varian offhandedly dismisses the “usual culprits,” clearly Google benefits from all of them. He suggests as much in his dismissals. Yes, there are large datacenters available, but not without a fairly significant amount of capital layout. I can’t, personally, go and buy a Google-sized datacenter tuned for search, even if I had the capital available to me. In fact, I suspect the only way to win in the search business at this point is distributing as much of the work (crawling, indexing, etc.) as possible. But as the Wikia rollout has shown, it’s not the scale of your datacenter that matters, it’s the scale of your index. Google can rely not only on the systems to mine that index, but on an existing “balance” of mined sites to draw on.

He suggests that there is no lock-in, since a user can easily choose to visit another search engine, but this ignores the significant–if largely hidden–cost of learning to navigate a new site and the results pages it throws up at you. Indeed, the only way to overcome those switching costs would be to first clone Google’s user interface and then slowly ween users to your alternative offering. (Think Word’s ability to clone WordPerfect’s commands.) And then, how do you convince people to switch for what appears, at least on the surface, to be a clone?

He dismisses the network effects model, and certainly the classical view (that the value of the product is affected by the number of users) doesn’t seem to apply to Google, but the common sense view that the name is familiar, and becomes more familiar as people use the internet and find it on pages that they have found through Google, means that it is hard to break into that cycle.

The idea that people go to Google simply because it has become, through experience, the best search engine, has merit. They have managed to “tune” their engine to best match the average user’s needs. But this isn’t the only reason–and may not even be the most important reason–it is the most popular engine.

Or, to use his metaphor: yes, Google’s got a great recipe. But if you really think your recipe is all there is to the business, you shouldn’t mind too much sharing your raw ingredients. No? I didn’t think so.

Update: Not surprisingly, Eszter says it better.