By combining art, science, math and humanity, UM achieves analytic breakthrough
We were so impressed with Universal McCann’s new media analytics technology that we created a new category and declared the Interpublic agency this year’s winner. UM has found an innovative way to refine the media plan as it’s playing out. Therefore, it can make adjustments accordingly and, it is hoped, reach the targeted audience more effectively.
How does it work? Think about Wall Street, where stockbrokers constantly evaluate which stocks are working harder than others and moving money around to their advantage. Well, over on Madison Avenue, this media agency has produced a continuous cycle of “measure, analyze, optimize.” It’s called UM 3.0.
Huw Griffiths, executive vice president and head of global performance, describes the methodology as a “one-two punch, an accelerated market-mix approach with macro feedback [that] tells you broadly to move money to national TV, to print, say, complemented by [audience measurement] research.
“These are high-level, strategic learnings of what’s working and how to execute the change at the channel level. We hand that to the planners and buyers, and it’s not conceptual any more; it’s very practical.”
Clearly, the agency has made a significant investment in 3.0. One reason is that it is moving toward a pay-for-performance compensation model. “We have to have a high degree of confidence that we can impact and optimize this,” Griffiths says.
“The key part is that we’ve created a next-generation measurement first, about measuring performance, the business impact our plans are having for our clients,” he says. “Not just reach of frequency but how many units are sold, media performance. We try and take that and blow it out. These techniques allow you to link media with sales.”
Whereas the methodologies of the last 10 to 20 years are to do this annually or every six months, Griffiths says that is too late to change anything that isn’t working. 3.0, he says, does the analytics every month or weekly, even, depending on how fast data is available, making mid-campaign changes possible.
“It’s a combination of having the technology to handle data sources and deploying the right people — having enough people to build and execute models faster,” he says. UM has invested in modeling and analytics talent and has the bandwidth to build the technology out.
Hari Abhyankar, executive vice president of global performance enablement at UM, says the level at which you make the buying decision is more granular than mere analysis. He notes the yin and yang of 3.0.
“The first part is a combination of doing analytics more frequently and the planner, buyers and analytics working together. No amount of math will ever solve this problem. Math is not be-all and end-all. It’s all top down and stops where human judgment takes over. It’s art and science, math and humans contributing, closing the last mile, to a degree.”
Abhyankar’s been on a five-year journey to where he is now and excitedly speaks about having arrived. “It’s happening! Data’s coming in, clients want it; it’s all coming to life in meetings. We are at the proving stage.” (Among the clients already using 3.0 are Charles Schwab, Chrysler, Johnson & Johnson and Sony.)
Keith Camoosa, UM’s senior vice president and head of North American research, is the guy behind UM’s Audience Measurement Platform or AMP, the engine that drives 3.0.
It is “the umbrella term for data platforms, involving lots of sources of data across the media industry. We customize and create data with suppliers in the industry,” he says, noting that the types of data NPD Group and Nielsen offer “are not things that can be purchased off the shelf.
“Agencies have invested in a lot of data intelligence to plan and buy, but not in the data that tell us how well those are working or not working. It’s a new way of thinking about research within the agency.”