Nielsen: Facebook's Ads Work Pretty Well
BATAVIA, Ohio (AdAge.com) -- It pays to have fans on Facebook if you want your ads to work there too, according to the first public study to come out of the collaboration of Nielsen Co. and Facebook.
The study of more than 800,000 Facebook users and ads from 14 brands in a variety of categories shows a marked increase in ad recall, awareness and purchase intent when home-page ads on the social network mention friends of users who've become fans of the brand in the ad.
The impact on awareness and recall is even more pronounced when a home-page ad coincides with what Facebook and Nielsen term "organic" social advocacy, i.e. an item in a user's news feed indicating a friend has become a fan of a brand.
In short, so-called earned media generated when people mention or advocate brands makes the paid media considerably more effective, according to the study. Nielsen and Facebook plan to discuss results of the study in a session at Ad:Tech in San Francisco on Tuesday.
Facebook-home-page ads on average generated a 10% increase in ad recall, a 4% increase in brand awareness and a 2% increase in purchase intent among users who saw them compared with a control group with similar demographics or characteristics who didn't.
But the increase in recall jumped to 16% when ads included mentions of friends who were brand fans, and 30% when the ads coincided with a similar mention in users' news feeds. Brand awareness saw similar bumps: up 2% from just a home-page ad, 8% with a "social ad" bearing mentions of friends who were brand fans and up 13% when a home-page ad appeared along with a mention of friends who were brand fans in the users' news feeds.
Purchase intent was 2% higher among viewers of home-page ads vs. nonviewers, but got a four-times-bigger bump, up 8% either from social ads or when ads appeared alongside organic mentions of the brand in the news feed.
Earned and paid media
One major takeaway from the research is that paid and earned media work together in ways that could have implications well beyond Facebook, said Jon Gibbs, VP-media analytics at Nielsen. "The market has been talking very much about how to buy paid media and earn earned media, but there's been very little attention to the types of hybrid impressions and hybrid experience that blends these two," Mr. Gibbs said.
While Facebook's social ads present a fairly unique way of blending the paid and earned impressions, Mr. Gibbs noted that it's not a totally isolated example. He cited rich-media vendors that allow for Twitter feeds, social commentary or other kinds of consumer input within their ads. But he said having specific friends linked to a brand, as Facebook does, appears to have more impact than just incorporating social commentary broadly.
The recall levels for home-page ads on Facebook appear "slightly higher than standard norms we've done on other projects," Mr. Gibbs said. "What we've seen in both social ads and organic [mentions] are much higher than we've seen in other campaigns along these lines."
Rex Briggs, CEO of the analytics firm Marketing Evolution, which has conducted numerous online advertising effectiveness studies, called results for Facebook's regular home-page ads "unremarkable and in line with banner ads [generally]," but he added that the results for social ads and the impact of organic mentions make for "a really interesting story."
Nielsen appeared to employ a good methodology used since the first online ad effectiveness studies in the mid 1990s, Mr. Briggs said.
"It does what Facebook wanted to do, which is legitimize the advertising and business model of Facebook," he said. "What it doesn't do is give the cross-media understanding of how does this piece fit into overall marketing plans."
What Facebook also hasn't done, he said, is open its doors and data to a variety of research companies as others, such as Microsoft, Yahoo or AOL have done. That its internal data remain largely under wraps, and its template for creating fan pages remains relatively limited compared to what marketers can do with their own sites or other networks may also be limiting revenue for Facebook, he said.
Paid media cheaper
In all, Nielsen projects around 18 million Facebook users saw ads measured as part of the study, of which around a million also saw organic mentions of their friends in social ads. Roughly another million saw organic mentions of the brands featured in the study without seeing the ads.
Based on those numbers, it's still a lot easier -- if not necessarily cheaper -- to buy scale on Facebook than earn it by winning fans. It's also an indication to Mr. Gibbs that marketers need to focus on winning Facebook fans over the long haul if they want to improve their odds of success when advertising there.
Of the 18 million users exposed to the ads, only around 130,000, or less than 1%, "engaged" with them by clicking on them. But around 40,000, or around 4%, of users who saw organic mentions of their friends become brand fans clicked on those news items. The higher click-through on organic impressions is another indication of the power of earned media on Facebook, Mr. Gibbs said.
"I do think it requires a level of ongoing investment in social media," Mr. Gibbs said, as opposed to a series of short-term projects. He also said marketers who have large e-mail databases should probably be encouraging consumers in e-mail programs to join their Facebook pages.
Mr. Gibbs said he doesn't believe Facebook's plans to move from "become a fan" to the more click-prone "like" as a means of joining brand pages would have much impact on the numbers in the study. And he believes, though it wasn't part of the survey, that users by now have been exposed to enough of Facebook's social ads to realize that when they become fans of a brand, they may also become endorsers in that brand's Facebook ads.
The Nielsen BrandLift polls used to survey Facebook users was a "lightweight" poll, generally with only two questions, aimed at maximizing response rates.
Nielsen didn't incorporate actual purchases, as opposed to purchase intent, "because this is the first generation of this research," Mr. Gibbs said. "We wanted to stick to branding because it's language the market is very comfortable with. In next generations, I would assume we will start incorporating offline purchase and other transactional data as part of the analysis."