Vast amount of feedback hinders more than helps
For television research, think of this as the best and worst of times. Never has more information been available at such a granular level, allowing networks and marketers to slice and dice demographics, tastes and preferences.
Yet given the rapidly shifting nature of media consumption and habits, this flood of data threatens to be more confounding than clarifying — especially with the audience being divided and subdivided as never before.
Marketing and ratings gurus have new methods of contacting subjects and amassing data. Such breakthroughs, however, create their own new pressures and headaches, and the vast amounts of feedback raise questions as to how representative much of it is.
Certainly, little that can be culled from the Web and chat- rooms — witness the instant polls many programs and sites conduct — would pass the sort of academic rigor applied to any legitimate opinion survey.
Nielsen Media, meanwhile, is frantically working to keep pace with what’s known as the three- or four-screen environment, including, along with traditional TV, the Internet, mobile and (most recently) out-of-home viewing in venues like bars and hotels.
To see the new veins of info that can be mined, look no further than TiVo, the pioneering digital video recorder, which has become more aggressive about probing the company’s subscribers and sharing that info with paying clients. These go well beyond mere demographics to Orwellian-sounding treasure troves, such as identifying the percentage of people who saw a certain on-air promo and then later watched the actual program.
While Nielsen’s sample is in the tens of thousands, TiVo boasts 350,000 subscribers for its Stop/Watch service, which anonymously aggregates data from random boxes. That’s augmented by Power/Watch, an opt-in pool of 40,000, which can cross-reference program selections with political affiliations, income, even what kind of car somebody drives.
Obviously, such details — including “second-by-second measurement capabilities,” identifying when people fast-forward or change channels — are useful not just to advertisers but networks, which in theory could tweak shows based on such profiles. Tracking which commercials are zapped most often, meanwhile, illustrates “how the ads are resonating with viewers,” says Tara Maitra, TiVo’s VP of advertising and media sales.
In addition, the ability to funnel material into those boxes — already wired into millions of homes, as DVR penetration nears 40% in the U.S. — and cull data back out of them has significant implications for everything from ad sales to focus-group testing. People can be enlisted to sample shows without needing to round them up at shopping malls in Los Angeles or the Las Vegas strip.
Will such research tools be as valuable as traditional methods? It’s too early to say, and so far most innovations are supplementing rather than replacing old models, such as the biometric testing employed by Innerscope Research, which analyzes subconscious responses — namely, heart rate, breathing, moisture (or sweat) levels and movement — while test subjects watch a movie trailer or pilot.
For all that, the failure rate of new programs remains dishearteningly high. And the proliferation of media options makes securing accurate measurements increasingly difficult.
To underscore the research challenges fragmentation can pose, consider the following: Not long ago, NBC officials were left scratching their heads when CNBC’s ratings suddenly plummeted in the key news demographic, viewers age 25-54. Had people suddenly stopped obsessing about their 401Ks and day trades?
The network nosed around in conjunction with Nielsen and discovered the problem’s source: Several regular CNBC viewers within the Nielsen sample happened to turn 55 — enough, given the relatively small number of Nielsen-monitored homes that watch the business-news channel, to cause a precipitous decline.
Another, more nebulous issue is the tendency to view with suspicion — or filter through a self-serving prism — the information new technology can provide.
Take Bill O’Reilly — who has worked in TV most of his career — launching an assault against Nielsen in October 2008, seeing potential shenanigans behind “wild swings in the ratings that have benefited MSNBC.”
What sort of conspiracy? O’Reilly suggested the “unprecedented” fluctuations might have something to do with Nielsen executives being “overwhelmingly liberal,” which might be responsible for “big-time cheating going on in the ratings system.”
Not long after that, of course, Barack Obama was elected president, ratings for Fox News (including “The O’Reilly Factor”) soared, and the host has stayed relatively silent on the subject of tainted Nielsen data.
Then again, as research folk would tell you, the numbers don’t lie. But thanks to the innumerable ways findings can be spun and interpreted, people do.