Akamai has just announced the launch of Advertising Decisions Solutions, which they say will work with any of their existing clients to "apply behavioral-targeting layers to ad campaigns," reports Advertising Age.
Acerno, a predictive modeling business, which Akamai bought recently for $95 million, tracks user buying habits across a number of ecommerce sites, then apparently applies a new algorithm to predict what they might buy next — and serve ads accordingly.
Repurchase and loyalty programs are already very sophisticated. One that I worked on in the past integrated purchasing information from WebSphere, statistical information processed in Omniture and dynamically modified and delivered messages through a Responsys email system. This definitely resulted in purchase and repurchase lift, but it was very hard to attribute that lift to purely the technologies in play. Sometimes customers just come back and buy other stuff.
This solution from Akamai seems to go further in that it promises that this information can somehow be shared across multiple sites and advertisers, and I trust that Akamai is no "dummy" when it comes to all things privacy. What I found most exciting was the apparent promise that this information would be shared on a pure performance basis, in exchange for
The system that I worked on across multiple vendors was definitely not paid for with performance dollars. It was expensive to setup and maintain, and while it definitely had ROI, it was not risk-free as this promises to be.
This begs a few questions.
How does this differ from Deep Packet Inspection technologies that make this type of information available to site owners and advertisers, but has met with resistance from privacy advocates?
Doesn't this sound like the e-commerce equivalent of Facebook's beacon?
How do you feel about being tagged in your purchase flow and re-marketed to based on what you do?
Akamai shall offer the service to publishers, ad networks and advertisers, taking a cut of revenue gleaned from purchases resulting from the ads.
Tuesday, October 21, 2008
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment