Moneyball Meets Marketing as Ad Research Game Changes
Data has played a role in advertising ever since the first retailer started wondering about the effectiveness of his promotions. But in this age of big data, the capabilities that retailers and marketers have to track, analyze, and model shopping patterns and consumer sentiment is driving a fundamental change to how the industry approaches research.
It wasn’t that long ago that companies lacked the computational and storage resources to track all customer interactions. “When we started in the mid 90s, the technology didn’t exist to capture ever single shopping transaction, so we sampled everything,” said Nishat Mehta, executive vice president of global partnerships at the marketing consultancy dunnhumby during a recent panel discussion hosted by the Advertising Research Foundation.
“Because of that, we didn’t necessarily feel comfortable sending one to one communications,” Mehta continued. “Today big data gives us the capability to look at individuals individually. It’s a direct democracy, not a representative democracy, and that enables us to do things on the activation side that are much more direct.”
Advertisers’ databases have gotten bigger and offer much more detail on consumers’ buying patterns. But that hasn’t totally eliminated the need to conduct primary research, said Bill Pink, senior partner of client solutions at Millward Brown, which is billed as the world’s second largest market research organization after Nielsen Company.
“The irony of the moment is that the more big data insights we get, the more [we have a] need for small data as well. We’re doing more primary research, traditional surveying, than ever before,” Pink said during the panel discussion.
“It’s liberated research, because we’re no longer using primary research for things it no longer needs to be used for,” Pink continued. “We have endless supplies of behavioral data sets. But what keeps coming out of it is, the models and the analytics are so fancy, yet they continually generate new questions that require more primary research.”
With so much data pouring through the marketers hands, it’s more important than ever to find a productive way to sort the wheat from the chaff. One company that has been at the forefront of identifying positive signals amid the noise is the cereal manufacturer Kellogg Company.
“We’ve chosen to be very systematic in what we measure so we can identify the stats that are most indicative,” said Aaron Fetters, director of the insights and analytics solutions center at Kellogg, during the panel discussion. “Just like Moneyball. They said, home runs and RBIs are not the most indicative of us building a winning team. We’ve kind of said the same thing in digital.”
This process may suggest that the click-through of a given ad campaign is not indicative of its potential success. But then the question becomes, what is indicative of success? “And that’s why being consistent over time, we’ve been able to weed out the things that are not indicative and focus on the things that are,” Fetters said.
Even with the best data, the best models, and the best analysts, taking a data-driven approach to marketing can fall flat without somebody who can do the hard work of showing executives how insights from big data will translate into tangible benefits for the company.
“We need to be turning the conversation into, ‘That’s great I collect location data and there’s all sorts of wonderful things I can do, insights about where my customer are, but it will allow me to better measure out of home advertising, which will result of an ROI, etc.'” he said. “All those things we were used to doing when it didn’t have to do with data, all of a sudden we’ve moved away from because we have all this data and we believe insights are the only thing we can derive from.”