Study: ROI May Be Measurable in Facebook, MySpace After All

Ed: Ironically, how much did it cost to figure out the ROI on this campaign? Sounds expensive and time consuming.

Package-Goods Brand Earns $1.28 Million in Sales From $1 Million Social-Media Campaign

by Jack Neff 

Published: April 13, 2009

BATAVIA, Ohio (AdAge.com) -- Package-goods brands are still cautious about social media, figuring that the return on investment can't be accurately measured. After all, marketing on Facebook or MySpace might generate a conversation but not necessarily a sale. Now, however, a method is emerging to relate one to the other, potentially eliminating a major impediment.

What it reaped

MySpace marketing ROI for unnamed personal-care brand: Total consumers exposed: 76.9 MILLION Percentage of internet population: 40% Total impressions: 1.1 BILLION Media outlay: $1 MILLION Offline sales generated from campaign: $1.28 MILLION Sources: ComScore, MySpace, Dunnhumby

Recent research from ComScore, MySpace and Dunnhumby presented at the Advertising Research Foundation's Re:Think 2009 conference in late March suggests that even relatively small outlays on social networks by package-goods brands can result in offline sales impact and deliver positive return on investment.

Generally, the ROI tool of choice for consumer package goods -- marketing-mix models that rely on econometric analysis of changes in retail scanner data -- can't pick up the impact of the relatively small five- and six-figure outlays package-goods brands make on digital media.

To overcome that, MySpace teamed with ComScore, which uses a panel of more than 1 million people in the U.S. to track internet usage, and Dunnhumby, which runs loyalty programs for supermarket retailers and has access to loyalty-card purchase data from 59 million people in the U.S. The two panels include 60,000 people who are part of both databases, creating a single-source database that allows a definitive look at how internet ads affect offline purchases.

ComScore Chairman Gian Fulgoni last year approached fellow Information Resources Inc. veteran John LaRocca, VP-U.S. insights at Dunnhumby -- which runs the loyalty program for supermarket heavyweight Kroger Co., among others -- about combining efforts to measure digital campaigns.

Measuring sales
One of the first studies was for an unnamed personal-care brand that ran a $1 million campaign on MySpace last year, including a contest in which members submitted videos of themselves and friends for others in the network to vote on, said Heidi Browning, VP-client solutions at MySpace. The program also included online couponing.

By the standards marketers sometimes use to measure digital-ad effectiveness, the MySpace effort wasn't overwhelming. Of 76.9 million people exposed to the campaign in four months, as estimated by ComScore, only 765,000, or fewer than 1%, visited an advertiser page on MySpace, though roughly half who did (358,000) visited the advertiser's website.

But by the measure that matters most, sales, the campaign appeared to pay off nicely. It produced $1.28 million in offline sales, as measured by Dunnhumby, which compared purchases among shoppers not exposed to the campaign with purchases among those who were. That amounted to a 28% return on investment, not counting returns from repeat sales among consumers the brand won via the campaign. Only about 17% of the sales were of products advertised in the campaign; the rest of the sales lift went to the parent brand, in what's frequently called the "halo effect."

Particularly by package-goods standards, that $1 million digital outlay with one site was large. But Dunnhumby has run similar studies with smaller digital campaigns.

"Just about every one has seen some positive results," Mr. LaRocca said. "And the ones that have had negative results ... just had a poor message or poor targeting."

Promising method
While a campaign that reaches nearly 77 million people is certainly large enough to generate a read in marketing-mix models, the combination of the ComScore and Dunnhumby panels into a single-source database -- assuming the numbers are correct -- holds promise for more-accurately measuring many smaller efforts, said Gregg Ambach, managing director of the Cincinnati office of research firm Analytic Partners.

The bigger question is whether the ROI will hold up for bigger efforts, he said, justifying budgets similar to what consumer-package-goods brands spend on TV and magazines. Digital is "incredibly efficient, because the cost per thousand is low," he said. "But it's just not moving a lot of volume yet. And, of course, what you always grapple with is if they suddenly went [from $1 million] to $10 million in digital, would the return stay where it is? In my heart of hearts, I think the answer is no."

But it's also a question he said no CPG brand appears to have tried to answer yet.