Until Google came along, internet search suffered from two big problems as a business. First, it was hard to make money off the end users (as a result, search engines had become commodity services sold to and rebranded by portals) and, second, switching costs were low (there was little to stop users from hopping from engine to engine). Google solved the first problem, cracking the nut on search-based advertising, but it has never really solved the second problem. Users flocked to Google because its PageRank algorithm provided clearly superior results to those of other engines, but most Google users remain ripe for the picking - there's little to stop them from switching to another engine if they're so inclined. (Habit can be strong, but it's not that hard to break, at least on the internet.)

The low switching costs could turn into a big problem for Google for a simple reason: basic internet search is once again a commodity. Do a search on Google or MSN or Yahoo, and you'll find little differentiation in the relevance of the results. Yes, if you're a super-sophisticated searcher, you may be able to point to variations that you think are important, but casual searchers won't notice any difference - and the vast, vast majority of searchers are casual searchers. There may be another great, proprietary breakthrough in internet search in the future, but for the moment Google has lost its lead. As for expanding search to more specialized areas, like 18th century manuscripts or academic working papers on quantum physics, that's not going to make much of a difference to Joe and Jane Searcher, neither of whom gives a toss about musty books or egghead treatises.

Of course, Google knows this, as do its competitors. They're all looking for ways to increase switching costs, or, as we used to say, make search sticky. One way is to make it easier for a user to default to your engine - by embedding a toolbar on his desktop, say, or putting a search box into his browser window. That can be pretty powerful. I've continued to use Google for most of my searches simply because Apple stuck a search box in the corner of my Safari browser window. But it's also a tenuous advantage. If in the next Safari update that box gets switched to Yahoo search, I doubt I would go to the trouble of hacking it back to Google. I'd start using Yahoo. Unless you control the desktop or the browser, in other words, you're stickiness is in somebody else's hands. Somebody like Microsoft, who may just happen to be your biggest competitor. Or somebody like Apple, who sooner or later is going to sell its real estate to the highest bidder, squeezing your profit margin, or incorporate internet search into its own Spotlight engine. Or somebody like Dell, or HP, or even IBM.

A better approach is to do what Yahoo's doing with My Search 2.0 - using personalization to embed proprietary data into search services. With My Search, you can tag pages that interest you and then restrict future searches to the tagged set. (You can also join up with friends to tag interesting pages, and then restrict future searches to the "community pages.") Because you can't take your tagging data with you when you go to Google or MSN, you suddenly face a real switching cost. Of course, it remains to be seen how attractive personalized search services will be to Joe and Jane Searcher (who may not give a toss about tagging or communal browsing, either). And you can bet that other search providers will quickly mimic any such service - so while it will increase switching costs, it won't necessarily enhance competitive differentiation.

There's also the old-fashioned portal strategy: provide an array of useful services to get users to spend a lot of time at your site, and they'll tend to default to your search service. That's still a good strategy - even if in the long run portals become relatively less important in people's everyday use of the net - which is why Google, despite its promises to the contrary, now offers a portal. But in this model, search inevitably becomes relatively unimportant again - differentiation and switching costs lie elsewhere in the business.

Maybe, then, what we've seen in the last few years is an aberration. Maybe the basic internet search engine is fated to be a cheap commodity running behind the scenes. And maybe those who control the search function - and most of the related ad revenues - won't be the guys running the engine but those who own the desktop or the portal (or whatever replaces the desktop or the portal). Maybe search doesn't really matter.

6:36:05 AM    comment