Facebook isn’t crazy about clicks as digital advertising’s be-all, end-all performance metric. Rather, the company is more concerned with conversions.
In the fall, Facebook partnered with data provider Datalogix to launch a tool that helps advertisers figure out how many ads it should show a given user. The tool compiles this Goldilocks recipe based on Facebook’s ad impression data and Datalogix’s offline purchase data. Now, Facebook is opening up a bit about results.
Across 29 campaigns, Facebook found that splitting households into four subgroups based on how much they’ve purchased a brand’s products or a product in a given category has produced an average 22 percent lift in profits, said Brad Smallwood, the company’s vp of measurement and insights, in a preview of a presentation he'll make today at the Advertising Research Foundation’s Re:think 2013 conference.
In an interview with Adweek before the presentation, Smallwood acknowledged that not all advertisers will achieve such results given the number of variables involved in constructing the Goldilocks recipe. Instead, the idea is to change how advertisers view digital advertising, which obviously includes ads on Facebook. That is, to get away from a narrow focus on an ad’s immediate return on investment (did it get enough clicks to justify its cost?) and refocus on whether it fulfilled a brand’s ultimate goals (read: sales).
As a hypothetical example of what the Facebook-Datalogix tool can do, Charmin could determine the number of ads it should show someone who regularly buys their toilet paper compared with someone who doesn’t buy their toilet paper, someone who buys a lot of toilet paper but isn’t loyal to any brand and someone who doesn’t really buy toilet paper (i.e., a gross person). Charmin could then retool their ad buy to make sure they’re not wasting impressions or dollars on any of the subgroups or leaving potential sales on the table by not investing enough.
Smallwood referred to this mix-modeling as finding the “effective frequency” for a campaign or advertiser. Again, he stressed that the mix changes from brand to brand, even product to product, and especially user to user. That’s where Facebook comes in.
Not only does Facebook have almost the widest reach of any online property, but the wealth of demographic and interest information that users supply gives the company a unique position from which to evaluate and tweak their campaigns. It’s easy to see how this tool could be Facebook’s biggest bait in luring TV budgets to online.
As with any conversation around getting advertisers to release their reliance on clickthrough rates, Smallwood, who was joined in the interview by Facebook’s head of measurement platforms and standards Sean Bruich, invoked TV advertising, particularly the broad reach of TV and the evolution of TV advertisers’ ability to measure reach and fine-tune frequency. Bruich cited research from in the early 1980s that helped advertisers determine the right reach for their TV campaigns and balance the frequency with which those ads were shown to ensure a profitable campaign.
“What we’ve found is those truths researchers back in the ’80s found about TV seem to be true about online,” Bruich said. That said, online, and more specifically Facebook, takes things a step further in being able to dial that reach and frequency mix for more specific audiences than available through TV.
“The ROI of a campaign is made up of a lot of pieces of a campaign. We want to figure out what’s good and fix what’s not as good,” Bruich explained.
When Adweek asked Smallwood what the ROI of his talk on Wednesday would be, he replied simply, “Success is getting [industry executives] to think beyond just ROI.”