• by Steve Smith , Staff Writer @popeyesm, October 12, 2012

“Recency” has always been a watchword in data and targeting. For the last decade behavioral intent data has always prized finding people who are chronologically at the bottom of the purchase decision funnel. But what if the most accurate predictor of your next consumer behavior actually occurred a year or more ago? When it comes to goods and services that become relevant on a seasonal basis, it would seem that knowing what a DIYer was doing online last spring would be more relevant to a hardware vendor in March than data around that person’s activity from February.

That is the simple premise behind a new targeting model from the social media targeting company LocalResponse. This is the company that in the past year or so has been scouring the public social postings terrain for check-ins, mentions and other relevant user broadcasts to target Twitter offers and now banner ads to them in real time. In a new effort it calls “Historical Intent Targeting,” the company is deliberately reaching back beyond the traditional “intent” zone of weeks to see what people were broadcasting about themselves during the last buying cycle that would have been important to a particular marketer.

For instance, Sony Pictures used the technology this past summer to test campaigns that targeted off earlier declarations of moviegoing or movie taste. LocalResponse sifts through billions of Tweets, public Foursquare check-ins  and many other public explicit and implicit indicators of intent. Rather than target people with Twitter offers only, LocalResponse is now finding these people on the PC and on mobile devices in order to target them with relevant banner advertising. According to co-founder Nihal Mehta, LocalResponse is working with a network of data companies that are able to drop cookies  on users so that their social ID can be correlated to their desktop and evem mobile IDs.

Mehta says that the possibilities for moving outside the usual short time-frame box when it comes to identifying intent is boundless. Consider how the cycle of entertainment media now has this long anticipation cycle. Video games used to the only consumer product I recall that had built-in six- and nine-month preview cycles where anticipation seemed to be part of the consumer experience. But that same pre-release hype technique has been extended to blockbuster films (superhero and franchise releases), gadgets (iPhone 5) and even books (J.K Rowling).

The method can be extended to a host of scenarios. One way to look at social media is that people express their affinities for celebrities, events, causes, consumer goods in unpredictable ways. I am not typically a social poster. But the right confluence of inspiration and opportunity will lead me to post a mention of a new book or movie, a gadget launch or a response to a news piece. There is a tendency for social media cheerleaders to believe that the rest of the world embraces rampant and regular self-expression on the social nets as fully as they do. In fact, I suspect most of us are popping up pretty randomly on social media radar, so any targeting off our postings must cast a wide net.

It's hard to say whether on the consumer side historical intent targeting is any more or less “creepy” than the usual techniques of retargeting that tend to focus on last week or month’s online behaviors. If anything, recency in targeting only enhances the ad stalking effect, since the more recent visits to a specific vendor are fresh in one’s mind. Who the heck recalls what stray mention or check-in I posted last summer, anyway? The Internet does.