Cable's Big Bet On Hyper-Targeting
Imagine the scene. You're relaxing at home, engrossed in the flickering images on your big-screen TV. Ahhh, nothing like really great programming. Except in this case, it's not a show that's got your attention; it's a 30-second commercial.
Preposterous? Cable operators hope not. With their small, hard-to-measure audiences, cable companies and channels haven't commanded anywhere near the ad revenue of the networks. Now they're starting to fight back. Time Warner Cable Inc. (TWX ) will soon begin testing a system designed by software firm Invidi Technologies Inc. that uses the digital set-top cable box to track what each TV in any house has on. It's expected to go into a household test by yearend, with other cable companies following in 2006. By analyzing where a viewer channel surfs, Invidi says the system can figure out the age, gender, and probable interests of the viewer. Most important, the system then matches TV ads to the consumer based on what might actually interest them.
That's a long way from merely trying to match ads to TV programs that seem likely to attract consumers an advertiser desires. If a 50-year-old father and his 14-year-old daughter each tuned into a Seinfeld rerun on different TV sets, for example, Invidi's system would send each a different ad during the same commercial break. The father might see a pitch for after-shave lotion while his daughter would see one for acne medicine. If this experiment in hyper-targeting works, it could spread fast. As the share of cable TV households with digital set-top boxes rather than analog boxes rises from 38% at the end of 2004 to nearly 100% by 2007 as part of federal law to make all TV digital, operators such as Time Warner and Comcast Corp. (CMCSA ) hopes to make consumer sharp-shooting commonplace.
Advertisers have long been looking for a better way to measure and track TV viewing than the ratings from Nielsen Media Research that are used to set ad rates. The payoff could be big for cable operators. They get only about 30% of TV ad revenues despite having 48% of prime-time viewers because so many cable channels have minuscule viewerships, making it hard to set ad rates. If Invidi works, viewers might actually watch more commercials rather than zipping past them. That's something advertisers would likely pay a premium for. "Not wasting consumers' time is the new currency," says Invidi CEO David Downey (for a Q&A with Downey see, "This Ad's for You -- Just You").
While some are nervous about the Big Brotherish nature of having a TV that watches back, Invidi says the process is anonymous -- viewing patterns are not matched to addresses. Still, privacy watchdogs are wary. "It does tip my creepy meter," says Marc Rotenberg, executive director of Electronic Privacy Information Center, a nonprofit research center. But Rotenberg concedes that technology to make consumer profiling anonymous on the Net and through digital cable has grown fast.
This is how the system works: Invidi's software allows the cable company to constantly and anonymously monitor viewers and their channel surfing to build a database of digital profiles. It takes about 30 hours to detect the makeup of a household -- say, frat house, vs. nuclear family -- and store it, and about 25 clicks of the remote to tell who's clicking. Based on that information, the cable box determines if the viewer is a man, woman, teen -- or if they're, say, a likely pet owner. Invidi then adds in income assumptions gleaned from each viewer's Zip Code. The cable boxes digitally "vote" for which ad gets downloaded and triggers the same ads to all the matching profiles in the cable system. Invidi says its tests show the system predicts gender correctly about 95% of the time. Age plus gender accuracy is about 75% accurate. Not perfect, but that's about six times more accurate than current targeting, notes Tim Hanlon, vice-president/director for emerging contacts at Starcom MediaVest Group who has seen the system demonstrated.
Cable companies are out to boost that accuracy even further. Right now, they earn only about $20 per thousand viewers on average for a 30-second commercial, compared with closer to $25 for broadcasters. But "our advertisers have told us they'll pay as much as 10 times more for this kind of hyper-targeting," says Comcast Spotlight Inc. President Charles Thurston, whose company will test Invidi's software next year. "If the targeting is 10 times better than today, that price is not out of the realm of possibility," says Betsy Lazar, General Motors Corp.'s (GM ) director for advertising and media operations. And instead of selling one commercial per ad break, cable companies could sell six, though each would be reaching a much smaller group.
Advertisers such as GM are obsessed with better targeting, partly to reach TiVo-wielding viewers who skip ads. GM's latest tactic: telescoping, in which it creates spots specially for TiVo-equipped homes that flash an icon over a Cadillac ad, for example, indicating that a longer Caddy "film" is available in the video-on-demand menu. GM says 10% to 25% of viewers have been clicking on the longer ads. GM execs think coupling that strategy to Invidi's system could boost results.
Other marketers are eyeing the new technology as well. David Verklin, CEO of Carat North America and an adviser to Invidi, says clients, including Pfizer (PFE ), Hyundai, and Orbitz, (CD ) are interested. Orbitz Chief Marketing Officer Randy Wagner says reaching consumers with ads that they won't skip is a priority. She likens the goal to Amazon.com's strategy of recommending products based on past purchases. "It's so relevant, it feels like a service instead of selling," says Wagner.
The real test for Invidi will be if it can sell the system to the broadcast networks. Hyper-targeting consumers who are watching a relatively small cable show is fine, but what advertisers really want is to use this technology on big prime-time programs. For ABC (DIS ) or CBS (VIA ) to offer advertisers the option, however, they'll have to go through cable operators, who own the set-top boxes. That may be unsettling for network execs, but demand from advertisers is likely to drive cooperation. "It would be significant if we could just slice off consumers over 50 in any program for pharmaceutical companies," says David Poltrack, CBS' executive vice-president for research and planning. Note to couch potatoes: The boob tube is about to get a whole lot smarter.
By David Kiley in New York