Nielsen's TV ratings mess could be tipping point in its dominance
Nielsen TV ratings are like the weather. People—in this case TV network executives—complain a lot, but are also resigned to the fact that they can’t do anything about it.
That truth has never been more apparent than amid the pandemic. TV network groups, led by the industry trade body the VAB, have protested loudly that Nielsen’s workarounds for maintaining its TV ratings panel during the pandemic have failed, shortchanging networks on viewership. On May 6—the same day the VAB finally collected enough data to suggest Nielsen was seriously undercounting TV viewership—Nielsen’s stock rose 2.5% thanks in part to healthy margins and renewals of multi-year network contracts with pricing escalators.
Fact is, despite efforts to move away from relying on Nielsen ratings as the currency for TV buying, the shift isn’t happening quickly enough, forcing TV networks to continue to buy Nielsen ratings.
But the latest network-Nielsen controversy is different from the long series of other such tussles. It includes a rare—the VAB calls it unprecedented—rebuke for Nielsen from the Media Rating Council, the industry body whose accreditation gives TV ratings their legitimacy as currency. Aside from acknowledging Nielsen did in fact undercount audiences—at least during the month of February—there’s also the thorny issue around its measurement of Black and Hispanic audiences. At a time of heightened sensitivity to racial justice issues, it’s clear that Nielsen’s recent TV measurement problems disproportionately impacted Black and Hispanic households.
All of this comes as a shift toward streaming deals and new competitors present a real threat to Nielsen’s standing as the prevailing leader.
Finding truth in the ‘truth sets’
The MRC, citing simulations Nielsen provided in response to network complaints, said on May 10 that Nielsen undercounted TV viewing by 2% to 6% and people watching TV by 1% to 5% in February. That only covers one month, but it supports the VAB’s contention that Nielsen has undercounted TV viewing since at least last summer, possibly costing media companies billions of dollars.
Workarounds Nielsen was making in the name of following COVID-19 safety protocols meant fewer panel households participating in data collection either because their equipment was on the fritz, they headed to live with in-laws, or they just stopped pressing buttons to indicate they’re watching.
Nielsen began showing a substantial and growing dip in total TV viewing and total people watching traditional TV starting in June. The VAB already had been arguing with Nielsen privately for months about the numbers and took its grievances public in April. For one thing, Nielsen’s decline in TV-viewing households wasn’t backed up by other data providers, such as Comscore and VideoAmp, which get their data from millions of set-top boxes and smart TVs.
Nielsen too has its own census viewing data from devices and smart TVs, which may be as big or bigger than those of competitors, and which it increasingly used alongside its panel data, says Mainak Mazumdar, Nielsen’s chief research officer.
Nielsen often refers to its panel as the nationally representative “truth set” that will keep the bigger census databases honest, since even those millions of set-top boxes and devices don’t cover everybody in the U.S. But during the pandemic, the bigger data set also had to serve as a sort of truth set for the panel.
Nielsen’s read of its own census data sources led it to conclude that the decline its panel showed in traditional TV viewership during the pandemic was real, Mazumdar says.
For the TV industry, Nielsen’s flub could serve as a much-needed explanation for why ratings have been on a freefall. But Nielsen in an April 9 blog post and white paper pointed to other issues, including a lack of sports, delayed series premieres and “programs produced in homes (and basements).” For example, October 2020 saw 13% fewer new episodes on traditional TV than October 2019, Nielsen said. It believes people shifting to non-ad supported streaming options like Netflix, Amazon Prime and Disney+ were in some cases cutting the cord or just not turning on old-fashioned “linear TV” at all.
Yet executives of Comscore and VideoAmp told Ad Age that their data showed no significant decline of households tuning into traditional TV over the course of the pandemic. Indeed, Comscore even shows sustained increases in traditional TV usage among Black and Hispanic households last year.
Then on May 6 came more evidence that Nielsen was undercounting TV audiences after all. The VAB had been pointing for weeks to 9,400 households that had stopped reporting TV watching or otherwise had spotty data reporting.
These so-called “impaired households” were leading to an undercount that Nielsen wasn’t catching because its field representatives had stopped going into homes during the pandemic to check on problems, says VAB CEO Sean Cunningham.
Some of those homes Nielsen had been servicing remotely, in some cases with field reps working from their cars, Mazumdar says.
But Nielsen resumed in-field maintenance of all panel homes on March 22. Within a month, Nielsen reps had visited about half those homes, fixing equipment in most, but temporarily or permanently suspending data collection from about 1,000, according to VAB. As soon as Nielsen resumed in-field maintenance, its data on total TV viewing began to rebound to levels not seen since June.
That reversal led the MRC to request the simulation on what effect Nielsen’s pandemic workarounds had on its data. Ultimately, that ended weeks of “denial, delay and deflection,” Cunningham says.
There remain unanswered questions, like why Nielsen showed declines even in TV streaming among adults ages 18-34 for the six months ended in February vs. the year-ago period. Cunningham says that’s “contrary to four other sources, including our member companies, who had some of the largest streaming services.”
The VAB says Nielsen’s monthly panel size went from 39,600 households in February 2020 to 29,500 in February 2021, and that latter number included the 9,400 “impaired” households Nielsen just began visiting in late March. Effectively, that could have cut the number of households actually providing reliable data almost in half -- at a time when more people than ever are needed to accurately count viewership in an increasingly fragmented TV market.
Nielsen cites somewhat different numbers, with around 47,200 homes in February 2020, down to 37,900 currently. And Mazumdar says that while 9,400 homes were identified as needing maintenance, Nielsen isn’t finding problems with all the homes.
Decline of Black and Hispanic homes
Of particular concern to networks has been the loss of Black and Hispanic homes from the Nielsen sample. The VAB says Black homes in Nielsen’s sample have declined 28%, while Hispanic homes have declined 23% and white homes are down 18%.
Nielsen weights data from remaining households to make up for such discrepancies so that it still reports numbers for a representative sample of the U.S., Mazumdar says. But Cummingham says that means unreasonably large weight given to each minority household, increasing the margin of error.
This is especially problematic since networks that specifically target these audiences are already coming from a place where they need to demonstrate the relevancy and power of the African American audience, says one network ad sales leader.
Nielsen’s issues with Black and Hispanic audiences isn’t a new one, says Louis Carr, president, media sales at BET Networks. “This has been an ongoing problem. There aren’t enough meters in homes, issues with upkeep and all being done in a timely manner. I have been having these conversations for decades.
“The problem is, when you have such a small sample and anything goes wrong, it is magnified,” Carr adds. Pre-COVID, Carr estimates that one Nielsen household had about a 2% impact on BET’s ratings. With about 9,400 homes “impaired” during COVID, Carr says each home had as much as a 25% impact.
Carr points to BET’s award shows, which often trend on Twitter for multiple hours, but that social chatter is not reflected in the ratings. “It is obviously not a one-to-one comparison, but you would expect it to be a lot different than some of the outcomes in the ratings,” he says.
“We still, when it comes to weighting, have a pretty set process,” Nielsen’s Mazumdar says. “That’s agreed upon with the MRC.”
Nielsen has been working with the MRC as COVID affects the mix of its sample too, he says. “If you take all this information collectively, we felt, and we continue to stand behind the data and the weighting that represents what’s happening in the marketplace. We are making every effort to make sure the sample is representative.”
Real change coming?
While sniping between networks and Nielsen over TV ratings quality has been common for decades, the intensity of the current complaints and the growth of alternatives are among reasons to believe that a real threat to Nielsen’s monopoly is emerging.
“We’ve never had an industry before that has so much growth and promise that you have in census-level data sets and panels in the millions, not the thousands,” the VAB’s Cunningham says. “There’s not a market or a sector that isn’t helped by healthy competition.”
Nielsen has its own play in cross-platform media measurement of video using a panel in the millions under the moniker Nielsen One, which CEO David Kenny said last week could likely become a key currency for ad deals across TV and streaming by next year. Certainly the valuation of Nielsen vs. one of those competitors – Comscore – suggests investors are betting on Nielsen.
But this latest stumble on legacy linear TV ratings may not inspire confidence in Nielsen’s ability to accurately measure an even broader and more complex video universe.
“The networks are all looking at alternatives,” says Jane Clarke, CEO of the Coalition for Innovation in Media Measurement, an industry group created 12 years ago to develop improvements to a Nielsen-dominated space. “Each of them is creating their own proprietary measurement approach. There are groups creating proprietary group approaches. It’s so fragmented that I think there will ultimately be some consortium among networks. And if anything, this thing with Nielsen is pushing them to accelerate that.”
She cites Comscore, VideoAmp and Oracle Data Cloud as leading third-party alternatives to Nielsen. Emerging panel providers, which could help provide person-level estimates within those big household data sets, are “definitely cheaper” than Nielsen, Clarke says, “but not perfect yet.”
Clarke says she was told by one of the big media companies more than a year ago that it took out only a three-year contract with Nielsen while planning to develop a new alternative to run in parallel for a year before trying to make a change. Virtually all major networks are going into the upfront saying they’ll have multiple ratings or currency alternatives, she says.
ViacomCBS plans to offer guarantees against Nielsen or Comscore data, John Halley, chief operating officer for ad revenue and exec VP of advanced marketing solutions, said at February’s CIMM Summit.
“The big problem with getting away from Nielsen is that it’s such a big number in the budget of every media company,” Clarke says. “And these new panels are having trouble getting funding in that interim period. You have to get an alternative panel up and running for a year or two—and keep paying Nielsen during that period—and it always falls apart because nobody has the money. But the big four networks are closest to doing that.”
She says similar spending hesitancy from marketers is making it hard for the Association of National Advertisers as it tries to assemble a new cross-media measurement pilot. “They all think the networks should pay,” Clarke says. The ANA declined to comment.
But certainly Nielsen competitors see an opening. “That decline [in total TV viewership] and then it coming back suddenly, that’s exactly what a currency should not be,” says Comscore Chief Revenue Officer Carol Hinnant. “Small panels are just no longer reliable.”
VideoAmp Chief Measurability Officer Josh Chasin is more charitable toward Nielsen, saying he believes the company has done the best it could to maintain panel quality during the pandemic. But he says the episode sheds light on the down side of panel measurement.
“Buyers and sellers are going to be very concerned about the entire currency that they’re relying on going away, being disrupted, and that clearly opens the door for other players, because the legacy advantage goes away,” Chasin says. “What we’re hearing from both the buy and sell side is that they want another data source in place.
“Everybody recognizes the ubiquity of data sets that can be used as currency means Nielsen’s monopoly is coming to an end,” he says.
Ultimately any measurement alternative networks embrace has to get consent from agencies and clients – and that will still be a hard sell.
“The agencies think the networks are posturing and whining, and they do this all the time and just want to undermine the numbers before they go into the upfronts,” Clarke says. “And there are some underlying trends that are real. Linear TV is down.”
“The networks are really posturing on this,” says Cathy Shaffner, chief investment officer of independent media shop Empower. “They’re using it to garner the double-digit increases they’re looking for in the upfront. There probably was a little bit less of a dip [than Nielsen estimated], but there really was a dip.”
Shaffner says she also doubts networks are ready to walk from Nielsen. “The fact that they’re able to complain and get this independent view and get the numbers changed in advance of the upfront says they’re still tied to Nielsen a little bit more than they want to be,” she says.
Still, Shaffner says with networks demanding that 20% to 30% of upfront buys this year be in digital properties, that may be “the shove that Nielsen needs to get their measurement” of connected TV alternatives in place, “or they’re going to be obsolete.”
Another media agency executive, who asked not to be identified, said many buyers remain reluctant to accept Nielsen currency alternatives, regardless of the current controversy. “Clients love their year-on-year metrics, and until they buy into a new view on effectiveness and outcomes, there’ll still be resistance from conservative voices from both the agency buy side and clients. There will be a CTV tipping point, but not yet, and Nielsen needs to try and manage that transition and keep their position.”
How is the TV industry responding to the streaming wars? On May 24 and May 25 hear from ad sales leaders, agency executives and top brands on the state of the TV ad marketplace and how streaming is poised to reinvent the $20 billion upfront marketplace. RSVP here.