Reach and Frequency vs. Buying Segments: Media Planners Walk Through the Process
Through 2003 until about six months ago, there were enthusiastic collaborations underway to create an acceptable system for forecasting and reporting reach and frequency of online advertising campaigns across multiple sites.
One such effort, the ARF’s Online Media Council, chaired by Dave Smith of Mediasmith, brought together ComScore and Nielsen, Atlas and Doubleclick, among others, to work side by side with agencies and publishers to make this happen. But activity seems to have stalled. Moderator John Durham asked the panelists if we really need R/F. And, if so, why?
Moderator: John Durham, Managing General Partner, Pericles Communications
Panelists:
Sean Finnegan, Midwest Director, OMD Digital
Carrie Soriano, Group Director, The Digital Edge
Mike Zeman, Associate Director, Insights & Analytics, Starcom IP
The panelists, all from large integrated shops, agreed on both the need and the reasons. Carrie Soriano’s clients are comfortable with general media. Having comparable terms such as R/F and TRPs for interactive means The Digital Edge can ease more clients into putting more of their budgets online.
Sean Finnegan called both R/F and GRPs “largely directional” but necessary to provide a “channel-neutral, one plan, one language” for OMD clients.
Starcom IP’s Mike Zeman pleased some publishers in the room with the opinion that “media planners will be more willing to pay higher CPMs” for delivering 100 percent audience composition against the target. Zeman’s agency has used tools from Atlas, Nielsen and ComScore and has learned plenty about their strengths and weaknesses. But neither Starcom nor the other agencies on the panel have found the single right solution yet.
“Atlas gives us historical views based on ads served,” Zeman said, but does not address cross-site duplication or allow drilling down into the sites. “Nielsen has an optimization feature we like, but we can’t measure R/F beyond one month.” And while comScore is more flexible – allowing them to measure between one week and six weeks – “it lacks an optimization function.”
Finnegan gave thumbs-up to one suggested solution: combine ad server data for site-side demos and depth, coupled with panel-based data for network reach that can be normalized in the agency’s own proprietary system.
So buyers want a solution and will pay for it. Why aren’t the vendors delivering? And why aren’t publishers insisting on selling audience and providing frequency distributions that make sense?
“Publishers are not presenting R/F data,” said Finnegan. “We give them our own equations and ask them to use our formulas to show us what we need to show our clients.”
The Digital Edge has one client for which Soriano’s RFPs demand that the sites account for frequency distribution. “If our goal is to reach a large audience over a specific time frame and it turns out that we’re reaching 10 percent of your audience 211 times, nobody wins.” So far the agency has taken this tack with one client, but others may do the same.
Zeman noted the medium’s ability to target beyond demographics. “I am pleased that relevance targeting and behavioral targeting are gaining momentum, but they don’t necessarily fit easily into a R/F planning model. I don’t think planning against R/F will ever be as core to online campaigns as it is to offline. So, while tools can help support online media plans - and provide solid complementary info - I don’t think that they should drive them, nor that we should necessarily want them to,” he said.
Is R/F the next step towards media equality for the Internet? As Soriano said, “GRPs don’t make sense.” Reach and frequency goals change with the objective, and GRPs, like any gross number, don’t reflect these inverse relationships. “So,” she concluded, “we don’t need GRPs, but reach and frequency? Yes. That makes a lot of sense for our medium.”
