La TV en la Oscuridad

Este excelente artículo publicado en el CJR, explica como los cambios tecnológicos, acompañados de nuevos hábitos en el consumo de medios, están trastornando a todos los actores de la indutria, incluso a los que se supone saben mucho del tema. Al final una traducción del texto, sin editar.

Por Bryan Keefer
Since the beginning of TV time, ratings from Nielsen Media Research have been the arbiter of success on the tube. They determine, in large part, how the roughly $60 billion spent on television each year is allocated — and that $60 billion has a tremendous say in what stays on the air and what doesn’t. “Nielsen is the currency by which the television ad marketplace is traded,” says David Ernst, executive vice president for the media-buying firm, Initiative. But what if that currency slips? As any economist will tell you, a currency is only as good as the faith in it of the people who use it. And when the currency falls, there goes the neighborhood.
And faith in Nielsen is showing some slippage. The central problem is that what Nielsen was originally set up to do — measure programs broadcast by the big networks to viewers who watch in their homes — is increasingly no longer the norm. Just as the Internet has transformed print media, technology is radically transforming the way we watch TV. Consumers can watch television without their television sets, by using broadband Internet connections to stream video online, or even watching on their cell phones. Digital video recorders, now in about 6 percent of homes, have made what the industry calls “time-shifted viewing” much easier, even for those who used to leave their VCR’s blinking “12:00.” And video-on-demand, once the realm only of pay-per-view movies and sporting events, has begun allowing viewers to watch shows, including evening news broadcasts, that they once could only watch in real time. Television is, increasingly, an on-demand medium.
Consider, for example, ABC News Now, a 24-hour news channel launched about two years ago. It features programs with major ABC personalities like Sam Donaldson (viewers are invited to submit presidential trivia questions to “stump Sam”). But the only screens it’s currently available on are computer monitors and cell phone displays (ABC let a trial run on digital cable lapse earlier this year, but announced in April that the service would resume in July). And people are watching; according to the network, viewership peaked during the political conventions last year in the hundreds of thousands.
But those numbers are self-reported for a reason: the Nielsens can’t capture that audience. And if the Nielsens don’t capture it (and they don’t capture anything outside of in-home viewing on television screens), it doesn’t count, as far as ad buyers are concerned. “The advertising industry discounts the audience outside the home to zero,” says Initiative’s Ernst. “We basically don’t pay for those audiences.”
While the television industry has always griped about Nielsen’s various shortcomings, this is something entirely different, and unsettling to the television world.
Contrary to the impression one gets from the tidy tables of audience figures published each week by the Associated Press and USA Today, the Nielsens don’t capture every hour of television watched on every set. Rather, like a political poll, the ratings rely on sampling. For its national ratings, Nielsen tracks about 7,100 households using its “People Meters” — set-top boxes wired into television sets that record what is being watched, when, and by whom (family members press a button to tell the box who’s watching). It uses the same technique to capture information in five local markets (New York, Los Angeles, Chicago, Boston, and San Francisco; five others will be added later this year). In about fifty other local markets, Nielsen uses an automatic meter that doesn’t record demographic information — who is watching — combined with paper diaries asking viewers what they tuned into during the infamous “sweeps months”; in 150 of the smallest local markets, it uses only the paper diaries.
Like Tolstoy's proverbial family, everyone in the television industry is unhappy with the Nielsens in their own way.
Out of this come two critical computations: a program’s “rating,” the percentage of all televisions in the market tuned to a particular program; and the “share,” the percentage of televisions turned on at a given time which are watching that program. For example, for the week ending February 13, 2005, NBC Nightly News pulled in a 7.1 rating and a 13 share (each rating point is equal to 1.08 million homes, though homes often have multiple viewers, complicating the calculation). ABC World News Tonight came in second, tying NBC’s rating and share but with fewer total viewers, and CBS pulled in third, with a 5.3 rating and a 10 share. (Combined, the three newscasts had nearly 28 million viewers per evening, or roughly the same audience as American Idol, the top-rated program that week.)
Such ratings and shares, combined with demographic information, become the basis for decisions by advertising buyers, and the ratings do a good job of providing information about the bigger shows on the larger networks. But the ratings are only as good as what they can measure. Nielsen makes its money by charging clients — television channels and stations, advertisers, ad buyers, and others — for both collecting viewership data and for analyzing it. And like Tolstoy’s proverbial family, everyone in the television industry is unhappy with the Nielsens in their own way, mostly having to do with what the ratings fail to capture.
To begin with, the company only places its meters in residences. That means it doesn’t cover any viewing done outside the home — in offices, dorms, bars, or anywhere else. And while a few decades ago all or nearly all television viewing might have been done in the home, that’s no longer the case.
For example, for financial network CNBC, that translates into daytime ratings that the network feels don’t accurately reflect its audience. CNBC’s own research demonstrates that it has a 79 percent share of the cable viewing done in offices during the business day, and a very upscale demographic. Its promotional materials for advertisers encourage buyers to “Think outside the house.” (Evidently, this tack has been fairly successful; CNBC’s revenues were reportedly nearly $300 million last year.)
Another problem: because Nielsen’s ratings are based on a sample, it’s difficult for the ratings to accurately cover channels that don’t get substantial national distribution (20 to 30 million homes, according to those in the TV industry, though Nielsen suggests it might be less). The sampling methodology also makes some in the business feel as though the ratings can be arbitrary. “You get the sense that if two little old ladies in Nebraska aren’t watching your broadcast,” one news executive says, “your Nielsen numbers can drop half a million viewers.”
But far more important are the technological limitations of the ratings. “Nielsen has a huge challenge in trying to keep up” with new technology, says Alan Wurtzel, president of research and media development for NBC Universal. “My worry has always been that the technology will outstrip our ability to measure it. And I think we’re beginning to see evidence of that right now.”
Nielsen’s older meters, which the company is beginning to phase out, can’t collect information about time-shifted viewing via digital video recorders or about televisions with digital tuners, because its meters depend on being able to measure the analog signal from a television’s tuner. In the past, the company simply excluded households with such devices, which it labeled “technically difficult.” But with more and more digital sets in homes, and millions using digital video recorders, this is a market that Nielsen can no longer ignore.
In July, Nielsen began installing what the company calls an “active/passive” meter. The meter reads a code embedded in the audio signal of each program (inaudible to viewers), which the meter tabulates and sends back to Nielsen’s data processing center. There it is matched up and converted into viewing information. This system allows Nielsen to measure what’s being watched at any time — digital signals, and DVR playback.
The Portable People Meter is likely at least a few years away from being rolled out.
Still, Nielsen won’t begin reporting data on DVR viewership until January 2006. When they do, they will provide three cuts of the data: real-time viewing, time-shifted viewing done the same day the program airs, and a data set including all DVR viewing up to a week after the program airs. (Such data will give TV ad sellers maximum leverage with ad buyers.) According to David Poltrak, executive vice president for research and planning at CBS, it’s the networks and the largest shows that stand to benefit from the new data. “All the research we’ve seen so far suggests that the amount of viewing to the networks goes up” when DVR use is factored in, he says, “and the amount of viewing to the top twenty shows goes up substantially.”
While Nielsen is catching up with DVR use, the company’s slow pace of innovation remains a source of frustration for its clients. The company has long had a de facto monopoly on the television ratings. Various initiatives to launch competitors to Nielsen — most recently, a network-backed initiative called SMART, in the 1990s — have failed for two reasons: Nielsen’s aggressive response to competition, and the television industry’s unwillingness to give a competitor the long-term financial backing necessary to establish itself. “While I’d love to see competition, there isn’t any right now,” says NBC Universal’s Wurtzel, “They are the only game in town. It is really in everybody’s best interest for them to do as well as possible. I think they understand the urgency.”
Nielsen points out that, in some areas, it is caught between the interests of its two major groups of clients. Spokesman Jack Loftus says that advertising sellers — channels, networks, and system operators — want as much viewing counted as possible (DVR, outside the home, etc.), while advertising buyers prefer that the sample remain narrow. “It’s not that we can’t do it,” he says, “it’s that you can’t get agreement from the industry on doing it.
“I’ve often thought that if there were other ratings services, it would be good for us,” he adds, “because then people could compare, and see the value of our services.”
It’s possible that new technology will grant Loftus’s wish. While several firms have been doing propriety research for networks for decades, the data generated by new TV technology is opening the door to other companies. Comcast, the nation’s largest cable provider, now contracts with an outside company, Rentrak (best known for analyzing video rentals and movie box-office receipts) to evaluate video-on-demand usage. Other cable providers are also linking up with Rentrak.
Meanwhile, DVR service provider TiVo has quietly entered the fray. The company records data about what users are watching down to the second — of particular interest to advertisers who want to know if subscribers to the service are watching their ads, or fast-forwarding through them. The company also publishes its own weekly list of the top twenty-five programs, based on subscribers who have programmed their machines to record all episodes of various shows. (Lost and The O.C. both made the TiVo top twenty for the week ending March 27, though neither cracked the Nielsen list.)
The future, however, may lie in collecting actual viewing information from large numbers of people — not just a small sample — through the digital cable boxes already in place in twenty-four million homes. Those boxes are capable of capturing information about the programs being watched, though spokespeople for Comcast and Time Warner, the two largest cable providers, said their companies don’t do so right now. A Comcast spokesperson, however, did note that the company is considering various options for using that data.
Nielsen, however, is quick to point out that such data wouldn’t include the demographic and socioeconomic information that advertisers depend on, and that Nielsen provides. Loftus says that the company is in discussions with cable providers about using information from digital cable boxes to supplement its current sample. “Somebody is going to take that aggregated data and do something with it,” he says, “and we’d like it to be us.” Meanwhile, Nielsen, together with Arbitron, a radio ratings firm, is at work on a portable version of its People Meter. In March, however, Arbitron experienced some technical difficulties with the device, which it is replacing in trial homes*, and the device is likely at least a few years away from being rolled out.
"I call it the Tinkerbell phenomenon. If everyone believes Tinkerbell can fly, this whole media economy will work."
But even the portable People Meter won’t necessarily capture the ways in which viewer behavior may evolve. David Liroff, the chief technology officer at WGBH, the PBS station in Boston, says that, “While our focus for the moment is on someone sitting in their living room using a DVR on their television, we’re already seeing that the ways people access video content will become so numerous that they’re too complex to be tracked.” He adds, “Andy Grove, of Intel, has the idea of a strategic inflection point — the point at which old rules no longer apply, but new rules haven’t been written yet — and we’re certainly there.”
At the moment, for example, there is no effective way for a third party to provide ratings for wireless and broadband services. According to Bernie Gershon, the senior vice president and general manager of ABC News Digital Media Group, ABC News Now is able to monitor how many people are watching, but gets very little information about who they are or how closely they’re paying attention. Yet some advertisers, at least, are willing to take a chance on a service that doesn’t have — and may never have — Nielsen ratings. While Gershon notes that for now the channel’s primary source of revenue is through subscriptions via its broadband partners, “part of the hook is to be able to reach consumers in new and interesting ways. So there are a bunch of advertisers who are looking at broadband and wireless as the next wave of advertising opportunities.”
So far, the television industry has proven willing to adapt to the holes in Nielsen’s sample. But, as in any economy, television depends on faith in its currency, in this case, the Nielsens. If the community comes to believe that currency is unstable, it may well replace it with something else. “I call it the Tinkerbell phenomenon,” says Liroff. “If everyone believes Tinkerbell can fly, this whole media economy will work. But if there are too many doubters, and Tinkerbell has a hard time getting off the ground, the question becomes: what drives the media economy?”



Correction: The above has been corrected to more accurately characterize the status of the Portable People Meter.

Bryan Keefer is assistant managing editor of CJR Daily and co-author of the New York Times best-seller All the President’s Spin: George W. Bush, the Media, and the Truth.


TV IN THE DARK

By Bryan Keefer


Desde el comienzo de la televisión, los ratings de Nielsen Media Research han sido los árbitros del éxito. Deciden cómo el negocio de $60 billones cada año se distribuye por lo tanto la permanencia de los programas depende directamente de lo que diga este estudio.

Pero la fé en Nielsen tuvo un bajón. La forma en que se medía la cantidad de personas que ven televisión ya no es una norma.

La tecnología ha transformado la manera en que los consumidores ven programas de televisión. Los consumidores pueden ver programas sin utilizar su televisión, reemplazándola por Internet o incluso viendo las mismas transmisiones en el celular. Además los grabadores de video digitales, se están utilizando en el 6% de las casas. También el llamado video–on-demand o pay per view también han permitido que los telespectadores vean no sólo películas, sino sus programas de la tarde a la hora que quieran. La televisión se ha vuelto un medio “por demanda”.

Muchos clientes están insatisfechos:
Para comenzar, Nielsen sólo mide el rating de los hogares. Esto significa que no cubre nada en las oficinas, hostales, bares, etc. Y en actualidad no toda la televisión que uno ve es en casa.

Por ejemplo, la cadena de noticias financieras CNBC ha demostrado mediante estudios que un 79% de las personas que la ven lo hacen en el trabajo.

También hay una limitante tecnológica en los ratings. Nielsen sólo puede medir la señal análoga. Por esto se dice que la tecnología avanza más rápido que los métodos para medirla.

Nielsen acaba de instalar lo que llama un medidor “activo-pasivo”: instalaron un código en la señal de audio de cada programa 8imperceptible para los telespectadores) que registra que se está viendo a cualquier hora sea digital o no. Lo que registra va de regreso a Nielsen y se transforma en información legible.

Aún falta mucho-años- para que se cree el People Meter portable. Asimismo, la última tecnología “activa-pasiva” comenzará a funcionar en enero de 2006.

La lenta reacción de innovación tecnológica de Nielsen a frustrado a sus clientes. La compañía ha tenido un monopolio en los ratings. Las iniciativas de potenciales competidores para Nielsen- la más reciente fue en los 90 y se llamó Smart- no han resultado por dos motivos.

Una reacción agresiva de Nielsen ante los competidores y la poca voluntad de la industria de la televisión para dar un financiamiento de largo plazo para que se pueda establecer.

A su vez TiVo (proveedor de DVR) ha entrado lentamente ofreciendo un registro de lo que los usuarios están viendo cada segundo, lo que lo vuelve atractivo para los avisadores, que pueden saber si los suscritos a TiVo están viendo sus avisos, o sólo están adelantándolos.

Esta compañía también publica una vez a la semana los top 25 más grabados por los subscriptores a las semana.

Algunos empresas de cable digital, que colocan unas cajitas en cada televisor, también pueden captar qué programas están siendo vistos. Sin embargo, Nielsen argumenta que este sistema no puede identificar la demografía ni el grupo socioeconómico.

Hasta el momento la televisión se ha adaptado a las deficiencias de Nielsen. La televisión depende de la fe en que se le tenga a Nielsen. Pero, si esta fe empieza a decaer. Qué pasará con la televisión? Al final quién maneja la economía de los medios?