Does the Secret Adsense API Exist?


And if the secret Adsense API does exist, is it funding spammers and allowing them to break the TOS' that apply to the rest of us plebs?

There exists a private API in Adsense for retrieving ads via XML. This would allow a customer to completely customize the look of the ads and avoid the Ads by Goooogle appendage entirely. The API is RESTful, extremely lightweight and includes some very interesting options like returning only adult ads.

Google have been attributed with funding web spam on one hand, with Adsense, whilst publically making a show of trying to stop it - so if this post linked above is for real, how does THAT fit into the equation?

via ig


Google Ad Services API

I seem to recall during my travels that I have stumbled across a couple of sites using a non-publicised Google feed. Further investigation proved, mainly by a process of illimination, that the ad-feed was infact a mirror of Google adsense.

I would estimate that this 'API' or XML feed has been around for at least a year. And seems to be available only to a few select affiliates. I myself do a fair bit of cash with the Google Adsense network, (£75,000+) and my own inquiries have been met with blank faces and clandestine answers by Google - so Lord only knows what the current criteria for having access to the API are.

Of course this assumes a few things also, firstly that the ad-feed was not infact Googles R&D dept. playing about with ways of monetising the content they have in the adsense feed, and secondly that the API isn't just a Third Party hack that sits ontop of an existing Adsense account. (I know the latter is possible because I have wrote one myself just for Proof-of-Concept purposes a little while ago, although I don't use it as it's outside the TOS of Adsense.)

What does this mean? Well it may open the door once again for SEO/PPC marketers to use ad-feeds as content once again (at least for a short while), until of course the feed is blocked by the other engines. But this poses another question too? Would Google block the indexing of their own ad-feed appearing in Googles SERPs, and would doing this financially benefit them?

I suppose it's inevitable

I suppose it's inevitable that these things will run on an API. Banner Blockers (and the scum who use them) will quickly be able to ditch ads served from, or with a specific size. Running the thing from an API allows publishers to customise their ads, even on a per-page-view basis, preventing scum from turning off the ads that usually make the provision of their content possible!

I wouldn't exactly call it a

I wouldn't exactly call it a "secret" API, but the ability to retrieve AdSense ads via XML (and in a JavaScript array) has always existed.

I created this tool almost 2 years ago, and it's always gotten ads via XML:

Of course doing it with a production AdSense account without Google's permission would be a quick way to get yourself banned. It's not like Google couldn't see how the ad request was made (XML/JS, etc.) and probably an even bigger flag would be that every impression was requested by a single IP address (your server) since it's done on the backend, rather than the end user making the request.

I kind of just found it too

This weekend, inspired by Shawn's AdSense sandbox tool, I set out to create my own version. From my findings, there are at least 3 output options for the AdSense ad feed: xml, js, and html. There seems to be a limit of 20 ads returned for every request. I'm pretty sure all regular publisher AdSense code outputs using the html option. I also know gizmodo uses this "secret api" to display their ads using custom js code.

BTW: Here are the results of my experiments I also use the info on my Google search counts page at:

AOL is your friend

probably an even bigger flag would be that every impression was requested by a single IP address (your server) since it's done on the backend, rather than the end user making the request.

Need an ip that *G* won't block?

proxy your server request to a machine you have running on AOL broadband, then let that machine be proxied by an AOL proxy. blocking the AOL proxies would be financial suicide. AOL is the scraper's friend.

of course, to deal with the uplink speed you *may* need to cache your requests. on the other hand, the returning data is fairly light. an enhancement would be to have the AOL box build up an inventory of results to return by having it run full tilt all the time.

the final enhancement: if the AOL box runs independently, it can then do port 80 GET or POST transactions to the server for the purpose of supplying inventory and totally avoid the necessity of trying to get around AOL port restrictions. POST is better because proxies are not supposed to cache POST requests. Of course, it is also possible to have the AOL box query the server for anticipated inventory requirements.

in case you wanna try...

some months ago had some interesting comments in their html that were easily reverse engineered.

guess you figure the parameters out yourself. I just snipped the original useragent and IP out of the url - just enter your own ;)

think i'd get hung for

think i'd get hung for alternately integrating that in with the RSS? heh...

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.