Revenue Science Pumps Up Behavior Targeting

A notable player in the hot behavioral marketing space has launched a new tool to help publishers find audiences more quickly and easily.

Revenue Science’s Audience Search aims to make it easier for publishers to put together their own audience segments and calibrate the tradeoff between targeting and reach. The product launched Monday with the Wall Street Journal Online, one of the company’s eight online publishers, two of which just signed up in mid-March.

The tool incorporates another fast-growing sector of online advertising: search. Publishers can type in a phrase from an agency or advertiser’s request for proposal (RFP) letter, and get audience segments based on those words. To generate these results, the search engine goes through the words in every article and page in the publisher’s site.

When a given phrase is plugged into the search engine, Audience Search returns a group of people. Every person in the anonymous group has a proven metric: they have read some content around the specified words.

“If I search for ‘sports car,’ the engine finds the people who read about sports cars,” said Omar Tawakol, senior VP of marketing for Revenue Science.

Audience Search, an upgrade of the company’s Audience Select product, also supplies a slider bar balancing reach against relevancy.

“The fundamental tradeoff of targeting is reach versus relevance,” asserted Tawakol. On the extreme left side of the slider is reach — everyone who has done anything to show an interest — and on the right is people who are extremely interested.

“You can say, ‘I only want people who are extremely interested,’ and it comes back with 10,000 people, so then you might want to adjust the bar a bit and get 50,000 people, sacrificing some interest,” Tawakol said.

The idea behind behavioral targeting is that one can infer, by tracking where visitors go and what they do on a site, what type of advertising a person might be receptive to. In the case of Revenue Science’s new product, publishers infer that because people have read content featuring certain words, they are interested in related products and services. Because such targeting can reach people across the site — and not just on the page or in the section where such words appear — it can be especially useful for publishers that sell-out inventory in popular sections, a growing phenomenon.

But competitor Tacoda takes issue with Revenue Science’s assertion that because someone views pages upon which certain words appear, one can infer that person is interested in that subject.

“Tacoda customers have used search in their behavioral targeting since the earliest version of our product three years ago,” said George Simpson of Tacoda. “But our solution is purely behavioral, not contextual.”

For example, Simpson said in an email message, when visitors to Tacoda customers’ sites ask for the weather in Detroit or about a Toyota in Los Angeles, “the sites can immediately target those visitors with appropriate ads. This is a far cry from inferring that someone who opens a page about last night’s game between Detroit and Los Angeles is interested in traveling or buying a car, for example.”

Still, media buyers are showing interest in what both companies can do. “Revenue Science and Tacoda are providing tools that allow publishers to better optimize their inventory. The more, the better,” said James Hering of TM Interactive. “If they can help the publisher get smarter, faster, better cheaper, and improve the clients’ ROI, that’s great.”