Rust Valley

Google is at it again.  They are taking the bread and butter of someone else’s business and giving it away for free in order to further their massively profitable ad-selling business.  The Wall Street Journal is reporting that Google will offer a free tool to measure web traffic targeted at helping advertisers select the right venues for their products.

The new offering is likely to ruffle feathers on at least three fronts and will shake up the analytics industry over the next several months.  It should be interesting to see it play out.  In the process of trying to shore up a key part of their own business, Google is likely to make a few new enemies and may even endanger a current stream of data.

The offering is supposed to detail which populations of users are visiting what sites.  If it is integrated into the adwords system, then ad buyers would be able to see more about the audience of the sites in the content network in order to determine which of them are worth advertising on.

Adwords users already know that the content network is the weakest, most underdeveloped part of Google’s ad-selling empire.  What we’ve seen, in placing small ads for Babbledog over at Renesys, is that the click through rate (CTR) of the same ad on various parts of Google’s content network gets 1% or less the CTR of the exact same ad placed on search.  The low CTR on the content network is a known problem whose official answer is “those content pages are just so darned interesting that people don’t always see the ads!  honest!” The more likely explanation is that there is a behavioral difference involved.  People who are searching are in the act of reaching out and trying to answer a question or acquire information.  In that mindset they are much more receptive to well-targeted paid answers to their questions than they are when reading icanhascheezburger.com.

This is trouble for Google.  Search is already monetized and its growth will be relatively small.  On the other hand, the content network could, conceivably, be as large as the Internet as a whole.  That’s probably bigger.    But in order to make more money off of the Internet as a whole, Google has to do two things: 1) get more ads on all those pages; 2) get more viewers of those pages to click on those ads so that Google gets paid and advertisers are happier.  This new website analytics product is designed to do just that.

The WSJ says that advertisers might be worried about giving too much of their money to Google.  Since Google already has a dominant market share for online advertising, they might worry that they should not use Google for website metrics, too.  It’s a matter of trust.  Still, free is hard to argue with.  Comscore and Nielson Netratings should be worried.

There’s an interesting backstory here that is barely covered in the WSJ article.  Where is Google getting the data for this product?  ” Google’s new offering will be based mostly on data from Web servers, allowing for a deeper and broader view of Internet use.”  Hrm.  “Data from Web servers”.  What webservers?  Yours.

It’s obvious that Google is finally cashing in on their free Analytics product by aggregating all of your web hits and putting them to work for advertisers.  This should be of no surprise to anyone, since Google’s entire enterprise is about aggregating all of the data out there and putting them to work for advertisers.  However, the nakedness, the blatantness of this particular scheme may surprise some people and may make it more difficult to justify that “free” Google Analytics account.

All of this should have people thinking about the subject of website analytics and why this is big business in the first place.  Jay Adelson, of Digg and Revision3 fame, wrote an excellent piece about the sorry state of the web analytics industry last year.  He wondered why large sites couldn’t just publish audited web stats from their own logs rather than relying on silly panel-based metrics such as Comscore and Nielson offer or extrapolated data such as Google is likely to offer.

I have to say, he has a point.