Facebook has expanded its custom audience targeting feature to now include what it calls “lookalike audiences.” By porting their CRM databases to Facebook, advertisers can reach their existing customers and users who share similarities with their customer base.
Facebook’s custom audience algorithm will now pull additional clusters of users for advertisers to target based on similar interests, demographic data or location. If the new beta feature is effective, brands will be gaining a unique insight and direct connection with Facebook users who are more likely to become customers.
“I think this gives an opportunity for advertisers to get a better understanding, and another way of looking at, who their audience is,” notes James Malins, VP of cross-channel solutions at Adconion Direct. “By creating a lookalike profile against an advertiser’s existing CRM base, it means advertisers are going to be targeting the audience most like their existing one.”
“It’s based on all the Facebook data that’s targetable right now,” says Elisabeth Diana, communications manager at Facebook. “What we’re seeing is that advertisers are gaining new customers through it, but they’re also seeing lower cost per acquisition (CPA).”
The level of probability for conversions, purchases, and other marketing objectives tends to increase when advertisers reach users who share similarities with their customer profile and target them with different ads or unique offers, she says. Facebook declined to name specific brands in its early results for lookalike audiences, but Diana tells ClickZ that a travel site is reporting a 70 percent reduction in CPA while a shopping site is reporting a 56 percent reduction in CPA and a 94 percent reduction in cost-per-checkout.
“The lookalike audiences could have huge potential compared to traditional targeting as it removes significant guesswork out of the process. Facebook is really taking the approach to improve the reach and engagement of its advertising,” notes Dave Williams, CEO and president at Blinq Media, a longtime Facebook advertiser partner that was acquired by Gannett last August.
“Traditionally, advertisers have focused on targeting precise interests, topics, and demographics to reach new audiences within Facebook. Using this targeting is very subjective to the perception of the brand’s customer and often that perception is not necessarily correct,” Williams adds.
“The benefit is getting a message in front of qualified leads, in this case a segment identified by the advertiser that would otherwise be difficult to impossible to match,” notes Rebecca Lieb, analyst at The Altimeter Group. “Because Facebook has an almost unprecedented amount of data around its users, the platform can slice and dice audience segments in a highly customizable way. I therefore do believe Facebook can match advertisers’ criteria for targeting.”
Custom audiences, and the new lookalike audiences beta feature, can be applied to any ad unit on Facebook. Once advertisers extrapolate new leads they can further refine the clusters of Facebook users who match a desired profile by either optimizing for highly accurate similarities or broadening the scope to a larger population of users with fewer similarities.
“It is a step in the right direction in that advertisers will start to realize their audience might not be the cheapest one they want. It is time to start paying and targeting users that actually fit their profile and aren’t just the cheapest,” notes Malins. “It doesn’t change the dynamic for advertisers that were previously focused on getting the right audience, but it might surprise and shock other advertisers that won’t believe the differences in the audience they have on Facebook, and the true audience that likes their brand.”