Revenue Science today rolls out a beefed-up behavioral targeting tool that it says will help online publishers create their own audience segments, and adjust reach and frequency in real-time for
greater effectiveness in ad targeting.
Audience Search, an upgrade to Revenue Science's Audience Select product, launches today with Dow Jones & Co.'s Wall Street Journal Online, which is the
first of the company's eight online publisher partners to use the tool.
Audience Search uses a natural language search engine that enables publishers to plug in the precise words from an agency
or advertiser's request for a proposal letter. The search process delivers specific audience segments based on those words. The tool searches for the words in each article and page throughout the
publisher's site. Publishers can perform multiple iterations of a search in order to configure and build customized audiences. The tool also allows them to optimize relevance and create optimal
reach for the best ad buy via a manual reach/relevance slider bar.
The WSJ.com regularly sells out its ad inventory in the Technology section. "With targeting, publishers can create more
valuable inventory than they previously had," says Omar Tawakol, senior vice president of marketing, Revenue Science. Tawakol says the tool addresses the problem of audience quality: "How do you
know when you've created a good audience segment for your advertiser?"
In a demo of Audience Search for the MediaDailyNews , Revenue Science showed how entering the keywords "wireless,"
"cell phone," and "Nokia" yields all the audience segments associated with relevant sections of a publisher's site. Each of the sections is paired with the following metrics: a behavioral relevance
score, content score, reach score, total number of page views, and the average page view per visitor. The tool can examine every word on every page of a publisher's site.
"Traditional sales
teams come up with category level sales approaches, but this allows us to offer publishers a customized audience for a media plan," says Nick Johnson, vice president of sales. Tawakol puts it
differently: "Audiences aren't always prepackaged at a vertical level-- they're created at a campaign level. We believe we'll be closer to the needs of the advertiser."
The tool's reach/relevance
slider bar helps address the inherent problem with targeting, which is "do I reach everybody, or only the right people?" Tawakol adds. "It's like going from a grocery store to a restaurant ... no
one has ever addressed the notion of what is relevant to a person."
The online Journal was active in product development with Revenue Science on Audience Search. "For us, this is an evolution--an
upgrade to what we've been doing," says Randy Kilgore, vice president of advertising, WSJ.com. "The technology allows the tool to read virtually every story on every page," he says, adding that
travel, consumer electronics, personal finance, and automotive category advertisers of the Journal are the most interested in using the new tool. Audience Search can look at the words at the page
level, story level, and across a publisher's entire site.
Singapore Airlines is already using Audience Search to target frequent business travelers as part of its campaign with the Online
Journal. "We'll be able to identify a list of key words or phrases that are an indication of the relevance for our target audience," says Todd Fraipont, vice president, online media director, WPP
Group's Mediaedge:cia, which handles media planning and buying for Singapore Airlines. For example, using the words "Southeast Asia," "travel," and "Asia business," Singapore Airlines will be able
to target any visitor who has read any article on WSJ.com that includes those words.
"By adding more words or less words, or by reducing the time window, you're able to dynamically control your
reach and relevance depending on how broad or specific you made those parameters," Fraipont explains. For example, a media planner can configure the tool to search for readers who read an article
with the desired words in it last week or during the last 30 days. "This just takes behavioral targeting to a whole new level," Fraipont adds.
"Behavioral targeting [really] allows us to talk
about audience composition [with advertisers]," Kilgore says.
Other publishers are more skeptical about behavioral targeting. "You certainly want message relevance, but when you break it down to
such granularity, there is an issue with inventory management," says Scot McLernon, executive vice president of sales and marketing, CBSMarketwatch.com. McLernon points to the difficulty of
projecting how active particular category sectors are going to be at any given time. He says that it's difficult to project out how much inventory is available in a particular area of a site, and
worries about setting himself up for under-delivery. "I'm a big fan of behavioral targeting, but I still have an inventory issue in order to be able to fluidly project out [availability]."