Does The Web Know Which TV Shows Will Be Hits?

Say you buy TV ads for some of the world’s biggest advertisers, and you’re about to enter negotiations with the networks to purchase time in upcoming midseason shows. How valuable would it be to know in advance that Fox’s American Dad or NBC’s The Contender will be among the next big breakout hits?

Of course, it would be hugely valuable, especially as hit shows are fewer and farther between each year, as the TV viewing universe becomes ever more fragmented, and as clients continue to demand more of a return on every ad dollar they spend. The ability to heavy up on hot shows before everybody else knows about them would be a great advantage.

Interpublic Group media agency Initiative believes that both Dad and Contender, debuting in the next two months, will be hits. What’s the secret? With partners Trendum and TV Tracker.com, the agency has developed a new Internet research tool that it says can pick the winners in each season’s crop of new network programs, about 95 percent of which fail each year, according to network estimates.

The tool, dubbed PropheSEE, combines three elements: technology from Trendum that monitors conversations in chat rooms, news groups and blogs; TV Tracker’s data about program development, scheduling and ratings; and Initiative’s analysis. The partners think it is so hot that they plan to sell it to broadcasters. But they will have competition. At least one other company is selling a similar tool—Toronto-based Brandimensions. Company COO Bradley Silver said most of his clients are Fortune 500 companies and include some broadcasters and ad agencies.

With PropheSEE, after the crunching the data, Initiative gives new shows a “buzz” rating, which quantifies how much each one is talked about each month on the Web, and a “sentiment” rating, which breaks out the buzz on each show by positive, negative and neutral comments.

Back in July, few foresaw that ABC’s Desperate Housewives would be the most talked-about show of the new season. But with the aid of PropheSEE, Initiative claims it predicted Housewives would generate hit ratings as well as buzz—two months before it premiered.

Stacey Lynn Koerner, director of global research integration at Initiative, said PropheSEE is much more than a hit predictor. In addition to helping clients know where to shift weight around for TV buys, it can spot branded-entertainment opportunities, she said. “We can really drill down and look at the depth of conversation, what people are actually saying about shows and characters,” Koerner explained. “What are the most engaging factors of one show compared to another, and what’s compelling about different characters.”

Added Initiative CEO Alec Gerster: “PropheSEE will give us a better contextual understanding of consumer engagement with the programs, in addition to giving us the potential to spot breakout shows that are below everyone else’s radar.”

Initiative has just begun sharing data with its clients, and at this early stage, many of them think the company is on to something. “I’m excited about it,” said Richard Taylor, svp of brand marketing at AOL. “We’re just beginning to scratch the surface on what we can leverage the Internet to do for us in the research area, but this taps into what people are really saying about current programming and shows and branded property and things that are going on.” Taylor expects to use the data when placing scatter dollars this midseason.

David Poltrack, evp of research and planning at CBS, said he is considering several Web-monitoring tools, including the one offered by Brandimensions, which culls feedback on TV shows and also about any brands a marketer wants to monitor. “These tools are a means of creating contact with the consumer that we are very interested in and that we are exploring,” he said.

Such tools would compliment the 20,000-member online research panel that CBS regularly surveys via e-mail. In fact, Poltrack said CBS knew back in June that Housewives would be a hit, based on results from its own panel.

Some outside the Initiative client base are more skeptical. Ad research consultant Erwin Effron dismisses such prediction techniques as ineffectual. “Change ‘buzz’ to ‘chatter’ and you have the CIA tracking al Qaeda, and you know how successful that’s been,” he quipped.

Competing agencies give PropheSEE and tools like it mixed grades. Said the head broadcast buyer at one top-10 media agency: “You can’t predict hits and misses with panels. It has to do with the quality of the show, promotional activity, time period and pixie dust.”

Brad Adgate, svp and director of corporate research at independent Horizon Media in New York, agreed. “I think one of the reasons Desperate Housewives did so well was that HBO wasn’t airing Six Feet Under, Sex and the City or The Sopranos,” he said. “Factors such as time period, competition and lead-in are more important indicators” than Web buzz.

Having said that, Adgate believes there is value in tools like PropheSEE—not so much as hit predictors but in terms of gaining insight into consumer engagement in programs. That, he says, ought to lead to more effective branded-entertainment approaches. “That’s a lot more interesting and probably where the future of TV is headed,” he said.

Horizon itself is now developing its own Internet tools for TV research, although Adgate says it is too early to release details.

While you can argue the finer points of predicting the popularity of TV shows, there’s no question that marketing research will increasingly rely on Internet-based tools. “Seventy-five percent of our research is probably related in some way to the Internet,” says Kate Sirkin, evp and global research director at Publicis Groupe’s Starcom Mediavest Group. “It’s cheaper and quicker.”

As to PropheSEE, Sirkin believes the jury is still out on techniques that use chatter to predict hit shows. “I think it’s interesting but unproven,” she says.

Still, SMG is experimenting with new techniques to get an early read on the types of shows consumers develop appetites for. “We’re developing an influencer panel where we’re trying determine who talks about the types of shows that they want to see,” says Sirkin. “The idea is to predict which genres are going to be hot in the future rather than individual programs.”