Facebook to Partner With Acxiom, Epsilon to Match Store Purchases With User Profiles
Use a loyalty card for discounts at the drug store? The ads you see on Facebook could start looking familiar.
Facebook is testing out a new kind of ad targeting that will let brands market to users based on what they've bought in stores, according to execs briefed on their plans.
Facebook is partnering with data giants including Epsilon, Acxiom and Datalogix to allow brands to match data gathered through shopper loyalty program to individual Facebook profiles, much like it's done previously with marketers' customer data from their CRM databases. Datalogix, a company with a rich trove of loyalty-program data, gained notice last fall after Facebook partnered with the firm to decipher whether users exposed to ads on the social network ended up buying any of those products in stores.
The targeting would hypothetically enable Coca-Cola to target to teenagers who've bought soda in the last month, or Pampers to show ads to North Carolina residents who've recently bought baby products, since Facebook's own array of demographic and interest-based targeting options can be added to further refine audience segments. But adoption will be contingent on acceptance by corporate legal departments wary of becoming embroiled in a consumer privacy scare.
The targeting will function through anonymized matching of loyalty-program members and Facebook users through email addresses and phone numbers, according to sources with knowledge of the product. (Holders of loyalty cards from retailers are asked for their email or phone number when they register, and Facebook users sign into the site using one or the other, and a match between two corresponding data points needs to be detected to enable delivery of an ad.)
The product seems aimed at CPG marketers, whom Facebook has been assiduously courting of late, hosting its first CPG Summit in New York last month. That event put a spotlight on "custom audiences," which lets brands upload emails, phone numbers and addresses from their CRM databases to show ads to their existing customers on Facebook. The ad targeting powered by shopper data uses the same technology, called "hashing," to find a match without allowing Facebook data to be intelligible to data vendors, or vice versa.
OMD is part of the ongoing beta test, and its social media director Colin Sutton said he's enthusiastic about the measurement potential implicit in being able to connect the dots between Facebook ads and offline sales, as well as the specificity of the targeting.
"Our CPG clients can begin to micro-target specific sets of consumers based on their in-store activity and buying behaviors and customize the messaging," he said.
The notion of targeting people online based on their offline purchase history isn't new, and Datalogix partners with any number of ad-tech companies, including AppNexus, Invite Media and MediaMath. A key distinction is that until now, it's just been possible to buy an aggregate population of cookies, and it can be difficult to discern how many unique users out of the pool have seen the ad or whether 20% of them have been targeted several times, according to Tousanna Durgan, MEC's senior director for audience buying.
"Because of the recency and accuracy of Facebook, the likelihood of seeing the whole population throughout your campaign is greater," she said.
Though purchase-based targeting will be enticing to brands, adoption will hinge on Facebook demonstrating that it's crossed every "t" and dotted every "i" with respect to protecting consumer privacy. MEC's social lead Kristine Segrist noted that adoption of "custom audiences" -- or targeting using data from CRM databases -- has been slow, since approval often needs to be obtained from beyond the marketing function at brands.
"Facebook's challenge is going to be breaking down the process in ways that are simple to understand and fostering confidence that this powerful data can be handled in a responsible way," she said.
Epsilon referred a call to Facebook, which declined comment on the new tool. Acxiom and Datalogix didn't immediately respond to requests for comment.