[tweetmeme source=bradmehl only_single=false]
Google’s acquisition of Teracent last Fall may be bearing more fruit soon. It’s a platform that enables marketers to optimize the performance of online display ads using data from both online and offline sources. Teracent dynamically generates ads by combining creative assets and putting them into a multivariate testing blender. It then optimizes the ad units shown based on marketing objectives and data such as on-site conversion rate. Think of it as Optimost for online display ads, in some respects.
I suspect there’s a strategic connection to this functionality and the recent launch of Google’s new user interface which enables users to refine their search results based on Shopping, Images, News and other buckets. Here’s why: search refinements are – to a large extent – a proxy for user intent. A refinement is in many respects a declaration of intent. When someone searches for “blackberry’s new model” and refines by Shopping, Google’s algorithm knows they want to buy one as opposed to reading news about it. When that sort of user intent is married with on-site conversion data (from Google Analytics), powerful things can happen. Google can go beyond associating keywords with conversion rates, they will know which creative units should go with which keywords to maximize sales for specific use cases like shopping intent. The ads might be different if a user is in information gathering mode, like when they refine searches by “News.” And because the number of creative units that can be put into a test pool is essentially unlimited based on dynamic ad building technology, that creates a lightning-fast speed to market for the optimal messaging. In fact, the optimization happens in real time. In hyper competitive areas like e-commerce and CPG this could give astute marketers who experiment with and master this approach a leg up.