CIMM, the Coalition for Innovative Media Measurement, was formed in 2009 to advocate for improvements in media measurement and advertising analytics. Its members include TV content producers like The Walt Disney Company, CBS and NBC, media buyers like GroupM and Starcom MediaVest, and advertisers like Procter & Gamble and Unilever. Jane Clarke, the coalition's managing director, talks to Adweek about TouchPoints, its latest attempt to achieve this.
Adweek: Most people in advertising would agree that existing measurement tools and systems are broken. And one area where CIMM is really advocating for reform is cross-platform forms of measurement. What progress is being made there?
Jane Clarke: We’re thinking of cross-platform measurement like an advertiser. In a cross-platform world, you need planning tools that help you to understand how people use media. Its gotten complicated over the years.
We’ve found that you need a “day-in-the-life” context planning tool that includes media in all of life’s activities. That’s a project we’re working on called TouchPoints. It was gratifying last week to see Nielsen and GfK MRI each bought 25 percent of Media Behavior Institute to help officially launch USA TouchPoints next year.
TouchPoints is a research analytics product?
It’s an innovative use of an iPhone app where you can keep track of media usage throughout the day, along with what you’re doing, who you’re with, and when you’re shopping. It’s a life context planning tool for advertisers.
It's being done as a “recontact” of a big survey that [audience measurement firm] GfK MRI does. They’re going back and recontacting people who have used it and having them fill out a diary, which adds a time element into all of these media currencies. They can then combine it with Nielsen Television and Nielsen Online Data. It’s creating a synthetic, big single source panel. It’s as if the same person who is a Nielsen household is then in every other media currency survey.
So it’s like a workaround because Nielsen doesn’t collect that other data?
Yes. Because you can't. The premise is that the kind of person who would be willing to let you monitor every single aspect of their life is not the kind of person you would want in your panel anyway.
How is it useful then, if it’s not real?
It really is “as if.” You statistically create the data as if they answered all the surveys. Most people wouldn’t agree to hand over every single detail about all of their behavior. But once they’ve answered this huge MRI survey that already asks a gazillion things, then do an e-diary for a week, (the data collector) can link it to all the other surveys the user didn’t answer. It’s called “fusion,” and you do it by finding like people through these data sets and then applying the data to more people who are similar.
So then you can predict behaviors that you didn’t survey on?
You find a person who looks like another person, and the more ways they can look alike, the better. Then you apply the remaining data you don’t have to each of them.
And the benefit of that is?
It becomes a cross-media planning tool so that when smaller agencies and media companies go to sell these big cross-platform packages, they can go in, pick a target audience and go into this data set and find out all kinds of information about them and put together scenario planning. They can ask, “What would be our estimated unduplicated reach and frequency across all these people, on various channels within the context of their daily life activities?” That’s just the first bucket of what CIMM is doing.