Nielsen has partnered with IBM to put its Watson technology to work for advertisers. IBM will make the data-parsing system which famously outwitted "Jeopardy!" champions in 2011 available to companies for CRM, customer call centers and other purposes through its new Watson Engagement Advisor offering. Over the next several months Nielsen aims to apply the Watson technology to improve measurement of ad effectiveness and media planning.
"There's lots of information and data in the public domain that the average practitioner … typically isn't taking advantage of and Watson would use all of this information to come up with the best possible answer," said Randall Beard, global head, Advertiser Solutions at Nielsen.
A 26-year-old media planner....asks questions like, 'Who are my best prospects in the telecom category?' and 'How much should I budget for next year,'" said Mr. Beard. Using Watson, that media planner could tap into data she might normally ignore.
Watson evaluates social-media data and publicly available data in conjunction with proprietary data from clients and their partners such as response data from previous marketing efforts and continuously learns from that information using its natural-language processing and machine learning capabilities.
If a banking marketer is preparing a credit card offering for instance, the system could help determine who best to target. "What if I don't know what to ask? What if I just want to start to have a conversation about past activity?" said Stephen Gold, VP- marketing and sales operations for IBM's Watson Solutions
Conversation, not a query
"It's not going to bring back an answer," said Mr. Gold. Rather, the system will deliver a set of weighted responses. "It's not an IT query…it's a conversation," he said.
Nielsen is just embarking on its Watson application research, which will be spearheaded by Nielsen's Innovation Lab, a project launched about a year ago in conjunction with the Stanford Graduate School of Business. The effort will be tackled over the next few quarters, according to Mr. Beard.
"What we're hoping to do with Watson is use its capabilities to get answers to questions about ad effectiveness easier and faster," said Mr. Beard, adding, "We've already had some initial discussions with the big media holding companies and there's certainly a lot of interest there."
Watson is 75% smaller and around 2.5 times faster than it was when it burst on the scene for its trivia game show debut two years ago, said Mr. Gold. Back then, it took the computer system around three seconds to analyze 200 million pages of information. Its footprint has also decreased in size.
"We can actually put Watson down on a single server today…so that opens up a whole new world of opportunity on how Watson can be put to work," said Mr. Gold.