Nielsen's Brave New World

Paul Donato has spent much of the past year passionately defending Nielsen's TV ratings against charges that they under-represent minority viewers in big cities.

But it's what the chief research officer of Nielsen Media Research has not been talking about that really excites him. Donato, a 52-year-old lifelong number-cruncher, has been concentrating on the future of television itself and a myriad of science-fiction–worthy ways he plans to measure it.

The electronic gadgets and systems envisioned by Donato—“talking meters” and artificial intelligence—represent entirely new ways of thinking about how people view television.

At press time, Nielsen was expected to announce it will spend $2.5 million to fund independent, third-party research as part of a consortium to develop better TV-rating methods. Nielsen researchers are expected to be involved, but the budget would not be controlled by Donato, who would continue to pursue Nielsen's own R&D efforts.

Nielsen has to do something—fast. The 82-year-old ratings service, say a legion of critics, is a monopoly past its prime, a creaking giant about to be overwhelmed by new technologies. Viewers with digital video recorders, who skip ads as a rule, pose a threat, which Nielsen plans to counter with a new service counting DVR homes next year. Moreover, cable operators are moving increasingly to digital, on-demand programming.

Because of these fundamental changes in the way programming is distributed—and consumed—older, static forms of audience measurement are already obsolete in some ways. Nielsen's local people meters have only begun to replace what is acknowledged to be an antiquated form of TV-audience measurement: paper diaries.

“Media technology is moving at lightning speed, and you have to ask yourself if Nielsen is up to measuring it. I mean, it took them 10 years to develop the latest version of their meter,” says Brad Adgate, SVP/director of corporate research at Horizon Media, New York. “The problem is the lack of competition. The only time Nielsen moves fast to develop new technology is when there is competition in the marketplace. And right now, there doesn't seem to be any.”

Cable operators have already discussed banding together to create a TV-ratings service based on a census of digital–set-top subscriber data, but they've yet to go public with it. Comcast led the charge and held discussions with Nielsen and Madison Avenue, but to date, there doesn't seem to be a consensus within the cable industry. Comcast is expected to unveil soon a new type of ratings service for VOD viewing, which could give advertisers a new way to reach audiences.

To stave off competition—and obsolescence—Nielsen has turned to its in-house Buck Rogers, Donato. Next month, he will unveil his vision for the future during a series of national and local client meetings across the country.

One of the hottest new products is a talking meter, which Nielsen has begun installing in the homes of employees to study how the devices work, how people interact with them and whether they lead to better audience measurement.

Unlike Nielsen's current generation of people meters, on which flashing lights prompt viewers to register what they are viewing, the talking meters can literally hold a conversation with viewers, asking who they are, what they are watching and a few other things Nielsen might want to know about them.

“Why not?” says Donato. “Your phone can talk to you. Your car can talk to you. Why not a TV-ratings meter?”

As Orwellian as that concept may sound, the talking meters are just one in a series of methods Nielsen is developing to measure how people watch TV. The research firm's conventional diary, meter or even its state-of-the-art people meter cannot keep pace with the evolution of the TV.

Donato began formulating Nielsen's plan for the future about three years ago, in a manifesto he dubbed the “Six Children of Nielsen,” which outlined a half dozen futuristic methods for measuring TV viewing. Those concepts became part of a working R&D program two years ago when Nielsen unveiled its “MeterWorks” project, which Donato plans to tell clients about during next month's meetings.

Artificial intelligence

As he prepared for the upcoming presentations, Donato previewed an early draft of the project's latest iteration, MeterWorks II, for B&C. “This is where we're at and what we're about to install in the field,” he says. “It's our roadmap to 2010.”

He adds that Nielsen is in the process of developing or testing an array of new metering devices. Among them:

  • Electronic “tags” capable of telling Nielsen who's in a room when a TV set is on.
  • A super-secret “imaging” technology that would tell Nielsen if people are sitting in front of a TV set while a program is on.
  • Inexpensive “mailable” meters that can be sent via the postal service, returned and reused.
  • Arbitron's portable people meters, which can measure radio and, potentially, other media, in addition to TV.

Beyond these new products, Donato also is developing “systems” that would augment or even replace components of Nielsen's current research methods, including an approach that would “harvest” data from computer servers compiling TV-usage information from electronic programming guides, and another that would incorporate “heuristics,” or artificial intelligence that would make Nielsen's data-processing systems capable of learning whether people were cooperating with Nielsen's ratings research or not.

For example, talking meters could deliver immensely useful market research, says Donato. “It's 3:00 in the afternoon, and there's a 13-year-old boy turning on his TV. Should we ask who that is? Or should we assume it is him, especially if he is tuned to a channel that is consistent with what he normally watches? Or should we ask who he is if he tunes to something that is an unusual channel for him, like a soap opera?” Donato asks. Nielsen would incorporate the answers into its own judgment calls in finalizing ratings estimates.

Under Nielsen's current edit rules, if the 13-year-old boy didn't press his people-meter button, Nielsen would not credit any viewing on his TV set. If heuristics—which involve pattern recognition—were used, Donato says, a judgment call might be made to credit the 13-year-old with viewing based on his historical behavior. A talking meter could simply ask him, “Are you the 13-year-old boy who lives in this room, or are you a friend visiting after school?”

The talking meter needs to be tested further, Donato says. “Is it going to be perceived as an invasion of privacy? Or is it going to be perceived as greatly simplifying things? We might find that young people love it but it spooks people 55-plus.”

That in a microcosm, says Donato, is the future of audience measurement: different methods required for different audiences, even different markets.

Harvest data from program guides

No single silver bullet exists; Donato says Nielsen's plan is organic and evolving. “What we have is a line of sight on how to deal with the most significant potential changes that could occur in the viewing environment,” he says, wielding a chart he sketched on a legal pad. Called the “matrix,” the chart is a grid he uses to check off possible solutions for TV measurement in the array of markets Nielsen serves.

Next to “small markets,” Donato has checked off diaries and mailable meters. Next to midsize markets, he has mailable meters, heuristics, portable people meters, voice meters. Next to large markets, demographic markets and the national TV-ratings sample, he adds people meters, tags and imaging.

One solution for all markets will be harvesting data from electronic program guides. “There are almost 40 markets with 100,000 households. Even by 2010, it's questionable whether a market with 100,000 households can be metered,” he says, “but even in a small market, if you can get server data from an interactive program guide, then you're in a position to significantly improve the quality of the data.”

Donato says Nielsen is also still trying to crack some potential delivery systems, such as futuristic “in-home wireless” television, as well as TV distributed via Internet Protocol to PCs, cellphones and wireless devices.

Altering nielsen's plans

The changes in the way TV is distributed, how it is consumed and how advertising fits into the equation will fundamentally alter Nielsen's plans for measurement. For example, says Donato, if television evolves into a free or premium video-on-demand marketplace in which consumers essentially craft their own programming schedules, “the actual concept of an individual program rating becomes less important.” Instead, Nielsen might offer a “a rating for the distribution mechanism that is serving up a large portion of the ads.”

The biggest threat to Nielsen's future, say critics, isn't any new technology or the entry of a rival research company. It's Nielsen's clients. If the advertisers become irritated enough with Nielsen or lose confidence that the data is truly credible, they may simply stop using it to make business decisions, says Kate Sirkin, EVP/global research director, Starcom MediaVest Group. In the late '90s, some networks discussed backing a competitor, but the effort flopped.

Sirkin says advertisers and agencies already are developing far more sophisticated ways of measuring how consumers use media, including television, and those methods ultimately will lead them to plan their media choices better than they can with Nielsen ratings. And while TV advertising buys have historically been based on Nielsen's ratings results, she says, if agencies develop alternative methods that they are more confident in, they may simply stop using Nielsen's ratings altogether.

Nielsen has been criticized for introducing its innovation on its own timetable and occasionally without any input from clients. Nielsen recently anounced that it will begin reporting ratings for homes with digital video recorders, something it has been bypassing for years. While most clients would like to see Nielsen accurately measure DVR usage, its abrupt action without warning to clients left many miffed.

How Nielsen's R&D might ultimately affect TV-audience measurement remains to be seen, but media buyers praise Nielsen for innovating. “Nielsen always gets beat up because people feel they are not looking into the future,” says Rob Frydlewicz, VP, research, at Carat Insight, New York. “They should be credited with looking ahead.”