TechCrunch interviewed Powerset founder Barney Pell and Microsoft Live Search GM Ramez Naam yesterday, following the formal announcement of the acquisition. After a half-dozen or so questions about whether and when the deal would be formalized, Arrington dug in to the specifics of what the acquisition might do for Microsoft (NASD:MSFT) in their efforts to become competitive in the search market.
The key takeaway - according to Ramaz Naam, is that roughly 5% of searches now done on the Internet could be aided by the application of natural language processing. Considering that Microsoft has roughly 9% market share of the Internet search market, that means that roughly 0.45% of searches will get better results with Powerset.
Now, one could argue (and they do) that by returning superior results, users will move to Powerset/Live Search and their market share will grow. But, to make that happen, they need to show that they deliver radically better search results most of the time. You can't displace what's become a verb (Google) by offering a little better results once in a while.
I think that semantic search will, one day, play a role in improving search results. But my experience with semantic processing at ClearForest and in watching technologies such as Powerset and Hakia suggests that we are years away from seeing it deployed on a mainstream basis. So the fact that Microsoft has added a team of 50+ developers, focused on semantic search, to their Live Search team is a good thing. I just don't see it having any significant impact on Microsoft's ability to become a player in web search in the next three-to-five years.
The full TechCrunch interview is available as a blog post and also as a podcast.
