It’s not often that I find myself disagreeing with both Tim O'Reilly and Fred and instead agreeing with Mike Arrington. 

First a quick recap of the discussion (in case you have not followed it).  Fred supported the idea of Yahoo outsourcing search to Google as far back as February.  Over the weekend, Tim reiterated that view and extended it to Microsoft.  Mike responded with a piece titled “The Importance of A Competitive Search Market."  Mike points to competition in search being the driver for innovation and for a more reasonable distribution of the profits from search advertising. Tim then followed up with a rebuttal on why search competition isn’t the point (which in turn was picked up by Fred).

I had written on the same topic of competition in search a few days earlier, arguing that the economics of search were so compelling that people would always try to take parts (e.g., indeed with the search for jobs) and do a better job at those.  This would continue to exert pressure on Google which would help drive innovation and potentially force Google to share more of the profits with sites and searchers.  But these competitive forces would be significantly weakened if Yahoo and Microsoft were to outsource search to Google.  Google is already using their huge cash flow from search advertising to make forays into many other fields with free offerings that are free only because they are subsidized.  While that may look like a boon to consumers (or in some instances developers), it’s a long term loss in diversity and innovation for the Internet ecosystem.

Now Tim argues in his rebuttal that focusing on competition in search misses the point because search is just one of many services in this ecosystem.  As an analogy, Tim points to the many different small programs available in the Linux operating system which can be put together in many different constellations.  I think that’s a misleading analogy.  Search is more like the local loop is in the network.  Yes, it’s just one small part, but it’s the part everyone has to pass through to get to the rest.  Just check for yourself how much of your online activity starts with search.  For me it’s well north of 50% and I don’t see that shrinking in the future.

A lack of competition on such a crucial "choke point” is bad for the Internet.  Exhibit A here is what happened after Microsoft had defeated Netscape. We went for several years with a fairly horrendous version of Internet Explorer that hardly improved until Firefox started to get some momentum. Ceding core search to Google would likely leave us in a similar situation for search and for other areas that Google choses to cross-subsidize.

P.S. All of this is from a macro view.  Tomorrow, I will post on the same question from the perspective of shareholders.