Feeds

Google behavioral ad targeter is a Smart Ass

Distributed ad crunching

Secure remote control for conventional and virtual desktops

Google is handling its interest-based advertising behavioral ad targeting from a custom-built ad server known internally as Smart Ass, or smart ad serving system.

According to a former Google employee, the system was under development as far back as 2006, and a second person with knowledge of the server says it was deployed sometime before the end of 2008.

The system is designed to simultaneously crunch various sets of stored data related to each individual user before deciding which online ads that user should see. Like the rest of Google's vast back-end infrastructure, it relies heavily on distributed-computing techniques. The task of serving up an ad is broken up into myriad pieces, and each piece is handled by a separate computing resource

One thread, for instance, might examine a user past search history. Another might look at past ad clicks. And so on. It operates in much the same way as, say, applications built atop Google App Engine, the Mountain View web service that lets you tap into the company's distributed infrastructure. With App Engine, a system request is terminated if it takes more than 30 seconds or returns more than 10MB of data. Everything must be broken into relatively small chucks that can then be distributed across the Google back-end.

In March 2009, Google announced that on YouTube and across its AdSense network of third-party sites, it had begun showing ads to websurfers based on the pages they've visited in the past. The company preferred to call this "interest-based advertising." But it's typically known as behavioral ad targeting.

"We think we can make online advertising even more relevant and useful by using additional information about the websites people visit," Google vice president Susan Wojcicki wrote at that time in a blog post entitled "Making ads more interesting."

"Today, we are launching 'interest-based' advertising as a beta test on our partner sites and on YouTube," she continued. "These ads will associate categories of interest - say sports, gardening, cars, pets - with your browser, based on the types of sites you visit and the pages you view. We may then use those interest categories to show you more relevant text and display ads."

Google called the new scheme a beta test and indicated it had not been deployed on its search engine. It did hint, however, that it had done at least a modicum of behavioral ad targeting before this. "To date, we have shown ads based mainly on what your interests are at a specific moment," Wolcicki said. The key word is "mainly."

In an effort to fend off privacy complaints, Google launched the scheme in tandem with something called an Ad Preferences Manager, which lets you view and edit the ad categories into which Google has placed you based on your past behavior. It also offered a cookie-based opt-out and a browser plug-in that kept your opt-out even when cookies were cleared.

Though Google did not discuss this during its official launch, it later told The Reg that the new scheme meant that Google was now using the same user cookie across both AdSense and its DoubleClick display ad platform. But it would not be drawn on how much user data was pooled across the two platforms.

We asked Google to discuss its Smart Ass server and its behavioral ad targeting schemes. But it has yet to respond. It's unclear whether behaviorial ad targeting occurs on services beyond YouTube and AdSense. And it's unclear what data - from what Google services - is fed into the Smart Ass system.

Clearly, Google's long-term goal is to serve ads that match the preferences and habits of individual users. This is why Google toys with the definition of "anonymization" when anonymizing your data history. It wants to know what you've done in the past so it can serve you ads accordingly.

This is the goal of Smart Ass - and it's significant step beyond the company's original ad serving system, which served ads only in response to search keywords or website text. The original system, says a former Google employee, is now known as Dumb Ass. ®

The essential guide to IT transformation

More from The Register

next story
Assange™: Hey world, I'M STILL HERE, ignore that Snowden guy
Press conference: ME ME ME ME ME ME ME (cont'd pg 94)
Premier League wants to PURGE ALL FOOTIE GIFs from social media
Not paying Murdoch? You're gonna get a right LEGALLING - thanks to automated software
Caught red-handed: UK cops, PCSOs, specials behaving badly… on social media
No Mr Fuzz, don't ask a crime victim to be your pal on Facebook
Ballmer quits Microsoft board to spend more time with his b-balls
From Clippy to Clippers: Hi, I see you're running an NBA team now ...
Online tat bazaar eBay coughs to YET ANOTHER outage
Web-based flea market struck dumb by size and scale of fail
Amazon takes swipe at PayPal, Square with card reader for mobes
Etailer plans to undercut rivals with low transaction fee offer
Call of Duty daddy considers launching own movie studio
Activision Blizzard might like quality control of a CoD film
US regulators OK sale of IBM's x86 server biz to Lenovo
Now all that remains is for gov't offices to ban the boxes
prev story

Whitepapers

5 things you didn’t know about cloud backup
IT departments are embracing cloud backup, but there’s a lot you need to know before choosing a service provider. Learn all the critical things you need to know.
Implementing global e-invoicing with guaranteed legal certainty
Explaining the role local tax compliance plays in successful supply chain management and e-business and how leading global brands are addressing this.
Build a business case: developing custom apps
Learn how to maximize the value of custom applications by accelerating and simplifying their development.
Rethinking backup and recovery in the modern data center
Combining intelligence, operational analytics, and automation to enable efficient, data-driven IT organizations using the HP ABR approach.
Next gen security for virtualised datacentres
Legacy security solutions are inefficient due to the architectural differences between physical and virtual environments.