Google pocketed $3.8 billion in advertising revenue during Q2 2007, an increase of 58% over the same period in 2006.
If Google is to keep growing at this rate in the future two things need to happen. The first is a question of reach – at present over 300,000 advertisers use the Google pay-per-click advertising model so if Google can reach out to the millions of other websites in existence they can grow the market and increase revenues.
However increasing the size of an already multi billion dollar market is much harder than making the existing market more efficient. If Google was able to get even a small percentage increase in the click through rate of its Adsense adverts it would translate to millions of dollars in additional revenue.
At present Google uses a complex algorithm to determine not only which adverts to show but how much the advertiser needs to pay when you click on it. Factors such as the advert text and website content in relation to the search query are the main variables but there can be no doubt Google incorporates your previous search history into the equation.
Imagine a user who uses Google to perform navigational searches, they visit Google.com and type in Amazon a few times per week and have a tendency to click on links to Amazon.com in the Google search results more than any other site. Google tracks this behaviour and knows that if it shows a sponsored listing from Amazon at the top of the results the user is extremely likely to click on it. Hide the Amazon sponsored listing away at the bottom and the user might decide to visit an organic listing instead.
This kind of behavioural targeting is what could make a few percentage points difference and millions in extra profits for Google
Whats this got to do with Digg and Facebook?
Digg, Reddit, StumbleUpon, Facebook and pretty much every other social network have a huge amount of data about their users. For years I’ve been telling Digg exactly what I like, what I hate and what I read.
StumbleUpon knows the websites I visit every day thanks to their toolbar and seems to be better than anything else at finding cool stuff for me to look at.
Imagine if Google had access to this data. For every search they would know that I don’t use eBay but that its a good idea to show me some sponsored listings for Amazon. They could tell that since you always Digg Linux articles you might not be interested in buying the latest copy of Microsoft Office when you search for “software”.
With data from sites such as Facebook, Digg, StumbleUpon and even Myspace Google would have a huge arsenal of information to apply behavioural targeting to its users.
Even a 1 percent increase in revenue would be more than eBay paid for StumbleUpon.