The call for an audit of Facebook’s metrics comes a week after the social network acknowledged inflating its video metrics.
The social network says clicks aren’t the right way to measure all ad campaigns.
Facebook Inc. reached 1 billion users on Sept. 14, Mark Zuckerberg, the social network’s founder and CEO, announced today in blog post. “I am committed to working every day to make Facebook better for you, and hopefully together one day we will be able to connect the rest of the world too,” he wrote.
The social network also announced it has more than 600 million mobile users. Finding ways to present those mobile users with ads has been an increasing focus for the social network as it seeks ways to monetize its massive user base.
Facebook this year has rolled out multiple mobile ad units that appear in consumers’ news feeds, which display updates from friends and from companies they’re fans of. Reaching consumers in their news feeds is essential, says Facebook, because that is where they spend 40% of their time on the social network.
While most marketers have used consumers’ clicks to gauge the success of ad campaigns on the social network, that isn’t always the right approach, said Brad Smallwood, Facebook’s head of measurement and insight, this week in a presentation at a conference sponsored by the Interactive Advertising Bureau.
Ad impressions, not clicks, are how merchants can evaluate campaigns that aren’t about a direct response. That could be “the outcomes that happen in the grocery store, in the car dealership or in the local coffee shop,” he said. Or, it could be multichannel retailers’ efforts to use ads to boost awareness of their brand or drive in-store sales.
Facebook worked with data analytics vendor Datalogix to conduct a study based on 50 ad campaigns to reach those conclusions, he said. The findings are based on a new tool Facebook and Datalogix developed that connects Facebook ad impressions, that is a consumer seeing a brand’s ad on the social network, with in-store purchases.
Datalogix says it has purchasing information about more than 100 million U.S. households that it gathers via loyalty cards and programs at more than 15,000 retail stores. It then matches consumers’ e-mail addresses or other relevant information with the e-mail addresses or information associated with consumers’ Facebook accounts. While all the information is anonymous, according to Smallwood, Datalogix can detect differences in sales among consumers who viewed a Facebook ad compared to those who didn’t.
Facebook says Datalogix uses hashing technology, which takes an e-mail address or other information and converts it into a string of numbers and letters that cannot be reversed to create the original information. The hashing technology enables Datalogix to share the hashed version of its customer data and information, like their e-mail addresses, with Facebook, which hashes its own users’ information.
The report found that 99% of sales generated from online branding ad campaigns stem from consumers who were presented with, but did not interact with, the ads. Because the delivery of the marketing message to particular consumer segments creates value for advertisers, Facebook advertising is, in some ways, like TV advertising, Smallwood said.
Brands that used targeting to present ads to particular consumer segments saw, on average, a 70% jump in their return on investment, he said. However, there is a “sweet spot” that marketers have to hit. Consumers who see too few ads are unlikely to remember the message and those who see an ad too many times are likely to tune it out, he said. That sweet spot differs from brand to brand and campaign to campaign, and finding the right formula requires experimentation, Smallwood said.
Facebook has long given consumers advertisers tools designed to maximize their investments. When a marketer creates a Facebook ad targeting specific customer groups, he sees how many consumers on the social network will likely see the ad.