Saturday, February 21, 2015

Contextual Advertising

I was reading this article about how Oyster, a "netflix for eBooks" app, recently updated their recommendation engine to include contextual recommendations:

The gist of the article is that depending on the time of the day (for example) users will be recommended different titles. It made me think about how contextual recommendations could make their way to advertising. Imagine if display advertising could change depending on the time of day, what the weather was that moment, and perhaps even be location-aware. If display ads could be smart enough to start advertising umbrellas when they know its about to rain, then advertising starts performing a utility function for the user. The lines between advertising and immediately-useful-recommendations begin to blur. Certainly search-based advertising already has the potential to be very utilitarian for the user (because it is so intent-based), but their utility is limited by the instance of the user query. Contextual advertising has the advantage of personalizing the ad to potentially going the next step and actually being very useful for the user. Will advertisers go down that route and invest in the necessary algorithms and computing power to serve personalized ads to the hundred of millions of potential ad viewers out there? If initial testing show increase in conversion, then this could be a race for startups and big firms alike.

No comments: