March 17th, 2007
As a Javascript expert (ha!) I was recently given the fascinating task of deconstructing Adsense - reverse-engineering it in a sense, to pick apart all the scripts to see what makes it tick. Of course I have no way to hack into Google itself and see the machinery that parses the page and delivers relevant advertisers. Anything happening on their server stays on their server (sorry if you came to this blog expecting more).
What I did get was a thorough look at how they deliver the ads, and what information they are gleaning from the client as they do. It took about 8 hours in total to pore over every line and de-obfuscate all their mangled code (after all they have no reason to make it easy), and in the end I figured it all out.
My findings were at the bequest of my employer, and the 8 hours spent on my journey were paid for. So I'm not sure if I'm allowed to blog out the details I found. For now I'll just report that I did it, it was interesting, and if I'm allowed to I'll post everything here for you all to read... it'll be a big one
I'll give you one juicy tidbit
did you know Adsense reads the length of your history collection and sends it back to Google?
When you visit a page with Adsense on it, Google knows the number of items in your browser's history.
They don't know
which pages you visited before the current one. They only know
how many.
Why would they ever need to know that?
What could they possibly do with that information?
Why do they care enough to have written it into the code?
Filed under Uncategorized