Is a Google Monopoly a Negative Externality?

Contemporary cricitisms of Google revoles around the fact that they’re pursuing a monopoly. Google, of course, denies this—if they didn’t have to, PR departments wouldn’t be a cost center. But it’s clear that Google’s business will function better if they own as much data as possible about their users and users’ social connections. So, rather than ask if Google really intends to be a monopolist (they do; their constraints are PR and anti-trust violations), a better question is: would this be good for users and for other sites?

The Atlantic channels Paul Graham to claim that this would be generally bad for the web: even though user-level data supports the idea of customizing searches based on the searcher’s behavior, that ends up being a negative externality since we don’t have a common set of data to refer to.

But that begs an important question: should Google be the standard of truth? That sounds like a noble mission, but it also sounds like a special case of what Google is: a way to get answers that you’re looking for. To get answers that you’re looking for, you need two ingredients: an idea of what answers are, and an idea of who you are. Right now, Google knows what the general answers are for any given category of queries; they need to know who users are to put together the rest of the puzzle. That works at a large scale, but even at a smaller scale it’s powerful: Google Docs now uses the web for spell checking.

For anyone who’s used Google long enough, it’s unfortunate that Google is no longer a universal arbiter of truth. But that’s best understood as their wedge product—it’s annoying the way Facebook got annoying to Harvard students once it started letting Yalies in. As it turns out, universally agreed-upon truths are less popular and less commercially viable than individualized answers. (And of course, Google offers un-customized search results for anyone who’s really intested: just log out and add “&pws=0″ to the end of the URL for your search, and you’ll get something close to universal results.)

If Facebook can link up their data across properties—Google+ data for social connections, Youtube and Google Play data for leisure, Doubelclick and AdWords for commerce—they’ll have an unparalleled ability to show people the ad messages they’ll respond to, in a medium they like. As always, the question is whether Google can figure out social faster than Facebook can figure out search. Google’s big bet was that if they figured out the web, they’d have a superior product for users; Facebook has made them worry that they should have figured out users, instead.

Ultimately, a Google that dominated social would be about as useful as a Facbeook that co-opted search.