Robots, 'Facial Coding' and Instant Focus Groups: Unilever Wants Better Metrics Faster and Cheaper
Unilever wants to move from such standard ad-industry metrics as recall and persuasion toward measuring how well people engage with marketing. And it's willing to look outside the conventional research industry to get what it wants, according to Keith Weed, chief marketing and communications officer of the world's second-biggest ad spender, who spoke at the Advertising Research Foundation Re:Think 2015 conference in New York on Monday.
"If we are going to get to a more structured approach to this incredibly fragmented and challenging world, I do think some of these things need to be coming through to the top of your agenda," Mr. Weed said to researchers at the U.S. industry's largest annual conference. "We'd prefer to find the answers with you guys. But we shall find the answers. And if not, we'll have to look elsewhere."
Indeed, Unilever is already looking elsewhere. Stan Sthanunathan, senior VP-consumer and market insights, said during the same presentation that the company already has around 15 pilots for innovative research methods, looking largely at non-traditional providers. Some of that work is coming through the Unilever Foundry, the company's London-based incubator for ad-tech startups, Mr. Weed said.
Unilever joins rival Procter & Gamble in laying out challenges for research and measurement firms, albeit with a slightly different focus. In a talk earlier this month to the Association of National Advertisers Media Leadership Conference, P&G Global Brand Officer Marc Pritchard called on the industry to "break the cycle of addiction to outdated metrics" and move toward measuring sales-based outcomes of marketing.
In a presentation that might be seen as inspiring, or chilling, for research firms, Mr. Sthanunathan also laid out some stark realities both for his company and the industry.
"Everyone today talks about margins getting squeezed," he said. "My response to that is: Welcome to our world. Data is increasingly becoming a commodity."
That also cuts at Unilever's own advantages as a company that Mr. Sthanunathan said did its first marketing-mix modeling in 1939 and started an in-house research agency in 1962 (ultimately spun off and now part of WPP).
Today Unilever's old advantage in spending millions on global research can be replicated by "a pizza-shop owner in a strip mall" who can get a survey done overnight for $500, he said.
While data collection revenue grows 2% to 3% annually, analytics grows 8% to 15% and drives the industry's real growth, Mr. Sthanunathan said.
Robots, "facial coding" and instant focus groups
The pair, apparently only half-jokingly, brought out a programmable Nao robot from Aldebaran that can do facial coding to measure people's responses, which helps it carry on a conversation. Mr. Sthanunathan mentioned such machines might do "facial coding" research on a large scale. That's something Unilever has been working on with WPP's Affectiva unit, which uses a global panel of consumers using webcams to track their facial expressions in response to ads.
As Unilever looks to expand research capabilities, it looks for opportunities that can double the impact at half the cost in half the time, he said -- and without adding internal head count. Among the things the research department can do now is assemble a global virtual video focus group of five or six consumers for a senior executive within 10 to 13 minutes -- provided the recruitment parameters aren't too exacting.
But it's not just about robots and margin compression. Unilever is also looking to make researchers at vendor companies more like parts of its internal team, Mr. Sthanunathan said, with a "partner reward and recognition" program that gives individuals bonuses for innovation and good work directly -- rather than funneling those through their companies. Some of the thank-you e-mails he's received from that, he said, "brought tears to my eyes."