Who’s Really Watching?How Cable’s Digital Box Will Rock the Ratings World 5/15/2005 08:00:00 PM Eastern
As advertisers and networks hunt for a better way to measure who’s watching which shows, the answer may be in front of them: little black boxes on TV sets around the country.
|AND GROWING DIGITAL CABLE...|
|Projected number of TV homes with digital cable|
|SOURCE: PricewaterhouseCoopers Global Entertainment and Media Outlook 2004-2008|
|PUT PRESSURE ON NIELSEN|
|To update its ratings service to adjust to changes in viewing habits caused by DVRs, VOD and high-definition viewing, Nielsen will:|
|Increase its sample size from 8,000 homes/18,000 viewers to 10,000 homes/26,000 viewers by May 2006.|
|Include time-shifted viewing in national and local people-meter samples by January 2006.|
|Experiment with different types of meters, including those with voice and motion triggering.|
|Test a mailable people meter, an electronic device that can be sent to homes to capture demographic viewing.|
Some of the largest U.S. cable operators are quietly testing a service in nearly 2 million homes that may offer the first real competitor—or complement—to Nielsen ratings. The technology can record every click of the remote control by every digital subscriber. And it offers an instant census of millions of homes that dwarfs Nielsen’s current universe of thousands.
Cable operators have already begun tracking digital-cable viewership on a massive scale—without Nielsen. Comcast is collecting viewer data from 1.2 million homes in Philadelphia, for example, while Time Warner’s Oceanic Cable is crunching viewing patterns in 200,000 households in Hawaii.
At the moment, Nielsen Media Research has teamed up with selected cable partners to experiment with new technology. The ratings giant is negotiating with cable giants Comcast and Time Warner Cable on ways to access and use the data to monitor viewing in non-Nielsen households. And it has gone further with Time Warner, developing software to track channel changes in Nielsen homes within Time Warner’s subscriber universe.
Nielsen concedes that data from digital-cable set-top boxes could greatly enhance its offerings and expand the accuracy of current ratings, particularly for small networks. “Set-top data can have value if it’s linked to existing Nielsen metered points,” says Scott Brown, Nielsen Media Research senior VP, strategic relationships, marketing and technology.
The need for such a system, say critics, is obvious. In a media-obsessed nation of more than 109 million TV homes, Nielsen household ratings track the viewing habits of a sliver of that: a mere 8,000 demographically correct TV homes. Yet this archaic system is the creaky underpinning of an annual bazaar in which networks and ad buyers exchange more than 60 billion advertising dollars.
“As audiences get smaller and smaller across literally thousands of viewing options, it stands to reason that Nielsen’s current system of measurement won’t come close to cutting it,” says Tim Hanlon, chairman of the advanced television committee of the American Association of Advertising Agencies and senior VP/director of emerging contacts for Starcom MediaVest Group. “The crude process we use now has gone as far as it can go.”
Cable operators are already using remote-control technology to understand the likes and dislikes of their subscribers, fine-tuning their digital channel lineup and VOD offerings. They say the data may also boost local cable advertising as the sales teams have more- reliable data on viewership. “Local cable salespeople and even national TV salespeople will begin thinking more like direct marketers,” says Hanlon.
But turning the raw data from local markets into a tool for tabulating national viewing patterns faces many obstacles. Any real competitor to Nielsen would need the technical ability to process hundreds of millions of channel-changes, an industry-wide agreement on how the data is audited, and an interface to Nielsen’s current ratings, now widely accepted in the TV universe. Says Tony Palermo, VP of business development and emerging markets at Scientific-Atlanta (S-A), “There are a lot of strategic discussions that will have to take place.”
The technology itself is surprisingly simple: The software tracks every click of the remote control, keeping tabs on every channel watched by a household. Then, at preset times, the data is pushed out of the set-top–box and back to the cable headend, where it is gathered with other viewer data. To ensure privacy, the set-top box address is separated from the data. At the headend, software sorts the data by channel and time, giving the operator an inside look at the viewing habits of potentially every digital-cable subscriber.
The capability was discovered and perfected as an outgrowth of an obscure maintenance program by cable operators. The Scientific-Atlanta Retriever software, now in three trials, was developed to monitor the health of the set-tops and cable infrastructure. “This is sort of like trying to cure smallpox and accidentally curing cancer,” says Paul Kosak, Scientific-Atlanta VP of market development, Media Networks/Headend Systems.
Surprisingly, what makes census measurement so appealing is also its main weakness: its ability to pull in vast amounts of viewing data without viewer participation. Scaling out the census is as simple as adding software to the boxes (in the case of Retriever, that cost is about $1.50 per set-top box).
TNS Media Research, which is currently involved in the Oceanic Cable service, sorts the data recorded by Hypergate software from Navic Networks. As early as this year, predicts TNS COO George Shababb, “we’ll see some examples of how this data can be commercialized.”
Nielsen Still Rules
While the census model holds promise, Nielsen, the de facto standard for TV ratings, still holds the advantage. Its detailed demographic information and samples are representative of the entire population, not just digital-cable set-top–box owners. Nielsen, which staunchly defends the veracity of its ratings, shows advertisers and sellers who is watching the show, not just what channel the box is tuned to. In the census model, for example, if a viewer leaves the cable set-top box on while he or she is on vacation for a week, the box will report 24/7 viewership from an empty home.
To solve such problems, TNS has designed a product to monitor whether the TV set is actually on. It has also developed “UltraTag,” a small luggage-tag–size device that is worn by members of a household and receives ultrasonic signals to time-stamp who is in the room at a given time.
To address privacy concerns, no names are used. Data can be sorted on a ZIP-code-plus-four basis, which networks believe will be useful without being intrusive. “You can still know the ethnic makeup or if a household is upscale or downscale, enabling localized advertising opportunities,” says Jack Wakshlag, chief research officer for Turner Broadcasting System, which is looking into the system.
Next-generation Retriever software made by S-A can monitor such DVR habits as what programs are recorded and when they are watched, opening new advertising opportunities. Car companies, for example, could send out specific commercials to specific neighborhoods. Cable operators could also offer “telescoped” commercials, in which viewers choose to watch a longform commercial stored locally on the DVR before returning to the same spot in the original program recording.
Says Starcom MediaVest’s Hanlon, “I fully expect the industry and the media buyers to start using linear channel-tuning data, VOD-usage data, DVR-usage data and interactive-TV data. I don’t think the Nielsen sample methodology will go away, but it will be only one leg of the table.”
Reams of Data
Ironically, some advocates believe that the new census model, coupled with a system that allows targeted digital ads, could save the 30-second spot. “If advertisers can get a more targeted way to reach consumers, the consumers will also benefit because they won’t be bothered by junk,” says Hanlon. “The mechanism of an advertising pod is woefully out of date. When ads become more targeted, they also become more interesting and informative.”
Another company hoping to exploit the effects of such tectonic shifts in advertising is ErinMedia. The Bradenton, Fla.-based audience-research company says it can mathematically calculate what type of person is sitting in front of the TV without the need for a diary. A patent-pending technology called IDM (Inverse Demographic Matrix) takes three data points—U.S. census data, the set-top–box data and a description of the program being watched—for a formula that figures out who is watching TV.
“We’ll take the reams and reams of set-top–box data and turn it into meaningful reports,” says CEO Frank Maggio. He believes that, by working with system operators on a local level, ErinMedia will eventually be able to draw data from 15%-20% of the country’s digital-cable set-top boxes, giving it up to 8 million “data points.” “We can take the data and apply it to whatever model or mathematical formula the MSO wants,” he says.
Maggio believes this new service will be enormously helpful to new digital networks: “When you consider that a given market might only have 200 Nielsen households, there’s a good chance that a digital-cable network might get a zero.” The flaws in the Nielsen model, he says, are exactly why Nielsen is scrambling to add more households, count DVR usage and even begin using set-top–box data. Despite his unkind words for Nielsen, Maggio acknowledges, “If there’s going to be a sea change, it’s going to require the cooperation of Nielsen.”
Still, an accurate analysis of TV viewing will be difficult even with Nielsen’s help. Bob Barocci, president/CEO of the Advertising Research Foundation, says that even a mix of Nielsen and set-top–box data doesn’t answer the biggest issue for advertisers. “Even if we could collect accurate data overnight from every TV household in America, all we have is a measure of media delivery,” he says. “The real issue advertisers are grappling with today is measuring media consumption.”