As AdWeek’s Tim Peterson explained, to accomplish this, Facebook turned to techniques that are familiar to direct marketers:
A brand uploads a file including the email address or phone number a consumer provided when he or she made a purchase in a brick-and-mortar store or online; an app developer can upload the user ID it received when a Facebook user connected his or her account with the app. That file is hashed…Facebook then looks at its own hashed data to see where it matches with the advertiser’s information. Facebook informs the advertiser of the number of matches it found—actually, the number of an advertiser’s current customers it can target with a Facebook ad—and then the advertiser can run the campaign against those users.
According to Facebook, some of the advertisers it has been testing with have seen impressive results. One, a financial services firm, reportedly “ended up doubling its fan base.” Another, an app developer, used Facebook’s new targeting functionality to let existing users know about an update and saw an ROI that was at least double that of previous campaigns.
That companies would be able to use this type of targeting to good effect is not surprising: if you have an existing relationship with someone, you stand a better chance of getting them to pay attention to your message, particularly if that message is relevant to a product or service that they expressed interest in or purchased from you in the past.
Will the creep factor kill the goose that laid the golden egg?
Facebook generates well over a billion dollars a year selling advertising, but thanks to its exorbitant public valuation, it is under a lot of pressure to prove to investors that it can grow its ad revenue in a meaningful way. As a result, Facebook is increasingly looking to prove to advertisers that its ad offerings can generate an acceptable return.
But at what cost?
Inside Facebook’s screenshot of the Custom Audiences setup tab indicates that Facebook hashes the email addresses, phone numbers and UIDs advertisers provide before they are uploaded to Facebook. Although one could argue that the hashing should be done by the advertiser (and not a Facebook process), assuming that Facebook really doesn’t have an opportunity to look at personally identifiable information being submitted, the question still remains: how comfortable will Facebook users be once they realize that companies they’ve interacted with — including through other online and offline channels — are effectively stalking them on the social network?
At the end of the day, Custom Audiences may be the creepiest ad feature Facebook has ever experimented with, and it creates a substantial amount of risk for the world’s largest social network. Will the company’s willingness to allow advertisers to stalk users be the straw that breaks the camel’s back, encouraging users to, at a minimum, share a lot less about themselves? Time will tell.
But for advertisers enticed by Facebook’s increasingly aggressive ad capabilities, it’s worth remembering: when it comes to how you employ the personal information your customers entrust you with, your customer relationships could be in just as much peril.