Can Interest Targeting Improve Facebook's Ad Model?

Most people don't hear the name "Leighton Meester" and think "jewelry," but there's a strong correlation between the Gossip Girl star and consumers who are in the market for some bling.

The relationship had been unexploited and probably unknown until SocialFlow, the New York-based social media marketing firm, crunched the numbers. Working on behalf of a large jewelry client, SocialFlow was able to buy ads against Facebook conversations about Meester.

The result, says Frank Speiser, co-founder of SocialFlow, was a 2x to 3x jump in click-through rates for ads.

Since CTRs for Facebook ads are notoriously low, the actual figures are unlikely to impress outsiders, but getting a 0.03% rate instead 0.01% is a big jump. Speiser also says that there's a big viral lift -- a.k.a. new conversations -- for brands that run such so-called "interest targeting" campaigns. (Though his term for it is the far geekier "linguistic prototyping.")

Speiser was in full pitch mode on Thursday for linguistic prototyping as SocialFlow brought its Crescendo product out of beta. While SocialFlow's eponymous product lets marketers help target and measure their activity on Facebook and Twitter, Crescendo lets them buy ads on Facebook as well.

The value-add is that the ads are based on interest targeting so unlikely associations like Meester and jewelry can be employed instead of the comparatively inexact demographic targeting.

In the marketing world, there's nothing new about psychographic targeting, a concept coined in 1974. The idea is that the more you know about a consumer's mindset, the better you can market to them. For example, fans of Lifetime's America's Supernanny are more likely to be fans of Walmart while Target customers are apt to be devoted to CMT's Top Secret Recipe, according to Bluefin Labs.

To use a classic, if likely apocryphal example, data mining firms discovered in the early '90s that there was a high correlation between sales of beer and diapers between 5 p.m. and 7 p.m., so savvy retailers put the two close to each other.

Although Amazon has excelled at introducing customers to new products they might like based on such logic, it's a relatively new idea in social media. One issue is that the SEO template set by Google encourages a more literal approach. As Heather Pidgeon, VP of services at SEO firm iProspect notes, if you were to run a jewelry ad for the search term "Leighton Meester," Google would show preference for other ads that have a clearer connection.

The other issue is automation. Facebook rolled out Interests Targeting in August 2011, but the drop-down menu is likely to be useless unless you have access to the kind of fancy algorithm that SocialFlow has. "You could guess the targeting, but humans don't have the ability to listen to that number of signals," Speiser says.

Those barriers help explain why demographic targeting has so far been the norm for Facebook. However, with so much scrutiny of late on Facebook's ad model, the company would be wise to do a better job marketing interest targeting as third parties like SocialFlow work out the kinks.

"The whole targeting of interests and groups has to change," says Pidgeon. "You have to not only be able to buy the hot terms, but the right hot terms."

Image courtesy of Flickr, david_shankbone

This story originally published on Mashable here.

Source: http://news.yahoo.com/interest-targeting-improve-facebooks-ad-model-193235398.html

Fatboy Slim Rio de Janeiro Shark Week London 2012 closing ceremony Shark Week 2012 evelyn lozada UFC 150