Skip to main content

Making Multiplatform Viewing Count

As more young viewers access programming through a multitude of different screens, advertisers, networks and agencies have struggled with determining how to count them.

This may be the year they finally do it.

“I do think we will look back on 2015 as a defining year, simply because we really have reached a tipping point,” NBCUniversal research and media development president Alan Wurtzel said. “I think everybody understands [that] the technology that is out there and the resulting consumer behavior have become so disruptive, the changes have happened so quickly, that there is no turning back.”

Those disruptive behaviors include binge-watching, time-shifted viewing and using other platforms or screens, he said.

“All of that has gotten to the point to where it’s gotten everyone’s attention,” Wurtzel said. “This behavior is not just about younger consumers, millennials. It really is mainstreaming across all these different age groups. Folks who are over 50 are engaging in exactly the same behaviors.”

Programmers are experimenting with new metrics, and Nielsen is racing to keep up. Since 2006, the research giant has been tracking time-shifted viewing (for up to seven days after the original airing) via digital video recorders and on-demand services. Recently, some have advocated that the measurement window be expanded to seven days after air (so-called C7 measurement).


But with more viewers accessing programming through different devices and different services, Nielsen executive vice president of product leadership Megan Clarken said now may be the time to expand that window even further.

“Content is flowing to platforms that are not ad-supported, like Netflix,” Clarken said. “The C3 rating and the definition around the C3 rating are limiting the industry from telling the full story.”

But such a shift will require a collaborative effort between ad buyers, sellers and measurement companies. The capability to track different viewing methods has been available for years, Clarken said, and complaints that Nielsen can’t measure online, mobile or multiscreen viewing aren’t true.

RBC Capital Markets analyst David Bank agreed that more collaboration is necessary, adding that the push for C7 and beyond is a bit misleading. While he said longer periods would definitely lead to larger audiences, he isn’t sure that’s a metric advertisers would want to buy against.

“I think if you counted total views and you went out 30 days, I think you would have a remarkably healthier TV environment, but I don’t know if it’s monetizable,” Bank said, adding that those shows generally carry an identical ad load. “They [advertisers] would pay for C30 if they could be certain that they wouldn’t be advertising a sale that ended 14 days ago.”

Bank also took issue with the blame game that has cropped up as traditional TV ratings have plummeted. The onus can be placed equally on networks, distributors and ratings agencies, he said.

“It takes three to tango,” Bank said. “Investors in the space don’t care that it’s not Nielsen’s fault, or the advertisers’ fault or the operators’ fault. All of their fates are tied together on some level. The blame game isn’t really constructive for any of them. They all play a role.”

Today, networks are using a variety of customized layered data from several sources to sell advertising. While other measurement units are helpful, Bank said he believes that a single, unified currency will eventually rule the day.

Nielsen and others are working toward that goal.

Already Nielsen, Rentrak, comScore and others are tracking various viewership platforms, with Nielsen gathering data for its Total Audience Measurement initiative from distributors and networks across the country. Nielsen began the TAM initiative in 2014 and has signed up about eight distributors and 34 networks, according to Clarken, although she would not disclose the names of the participants.

“This year is a really pivotal year,” Clarken said, adding that Nielsen plans to have a Total Audience product in the marketplace by the end of 2015, depending on how many networks and distributors participate. “We’ll continue to support C3 ratings as part of that, but what we would hope will happen during the course of either [is that] while we’re building the capability, we can help shape the marketplace or help move the marketplace to a new currency metric, or as an alternative and another way for them to continue to have those negotiations, but also support C3 or C7.”

Still, it’s obvious that a change has to come and that it’s coming slower than some would like. And in some cases, the networks themselves are taking matters into their own hands.

When NBCU reports the next-day Nielsen rating for a program, Wurtzel said, it also includes an estimate for the three-day and seven-day ratings. NBCU also releases a report called the Total Audience Measurement index (TAMi), which the programmer began compiling in 2008. Wurtzel said the TAMi was not intended to be a currency, but rather a tool to show how the audience is shifting.

Viacom — which had a highly publicized dust-up with Nielsen in 2011, when kids’ network Nickelodeon lost its first monthly ratings crown to The Walt Disney Co.’s Disney Channel that November — is experimenting with ways to more accurately track multiscreen viewing.

“If it’s not showing up in the numbers we want from Nielsen, we find other ways to track it,” Viacom Media Networks executive vice president of data strategy and consumer intelligence Kern Schireson said at a recent Paley Center for Media conference. “We have premieres that are going out doing a number on Nielsen and doing triple that number in full-episode status in the same time period on a digital platform.”

According to people familiar with Viacom’s thinking, the programmer has teamed with a trio of ad agencies and measurement companies to capture and analyze set-top box, online and mobile data (Nielsen is not among the participants) to determine just who is watching its shows. The ad agencies involved represent about 2% of Viacom’s annual global ad sales, but the goal is to expand that business to include more ad volume in the future.

Outside sources can also help with hard-to-measure networks or dayparts. CNBC recently said it would track its daytime viewing exclusively via market-research firm Cogent Reports. With Nielsen, which will still measure CNBC’s primetime shows, the daytime audience sample was sometimes so small that the network occasionally generated no ratings at all for the time period, an obvious oversight, Wurtzel said.

Discovery Communications also has been experimenting with ways to use data to make smarter media buys for clients, forming partnerships with Rentrak, Clypd, Nielsen Catalina, NBI and Lake 5 Media.

For instance, Discovery will combine Rentrak’s set-top box data with Polk automotive data to provide more comprehensive measurement of ad campaigns across its programming portfolio.

The network will use the Clypd platform to streamline the buying process and allow select customers to trial audience optimization and deliver increased levels of marketing-relevant target impressions in concert with traditional age and gender benchmarks.

Nielsen Catalina and NBI provide single-source data, including viewing behaviors and purchasing information from individuals or households, while Lake 5 Media offers additional analytics to those new data sets.

“We are constantly looking for new ways to provide our advertisers with enhanced results by leveraging improvements in data and technology,” Discovery senior vice president of ad sales research Beth Rockwood said in a statement. “Offering our customers a seamless way to buy our networks will make us a better partner, while effectively managing the sale of our inventory.”


Networks are missing out on an increasingly large segment of the audience for its shows that aren’t being counted, Wurtzel said. For example, a typical episode of NBC’s hit thriller series The Blacklist generates a 2.7 rating in the 18-49 demographic the day after air, rising to a 4.2 rating three days later. At C7, the rating rises to a 4.6, as more Nielsen-measured time-shifted and VOD viewing is added in.

But using other data sources, the actual C7 rating for the show should be about 5.6, Wurtzel said, as there is another full ratings point Nielsen isn’t counting that includes viewers watching via PC or laptop, tablet, smartphone, game console, OTT device or out-of-home. That single rating point, he said, represents about 1.3 million additional viewers, or about 17% of the show’s total audience.

The gap grows even wider with shows that are skewed toward younger audiences. For example, Parks & Recreation snagged a 1.2 nextday rating in the 18-49 demo, rising to 1.6 on a C3 basis. The show’s C7 rating was 1.7, but another full ratings point was missed due to alternate viewing methods, which Wurtzel said represented 37% of the show’s audience.

Wurtzel was quick to add that Nielsen doesn’t warrant blame for all of TV’s measurement challenges; its currency will be used by ad buyers and networks for the foreseeable future.

However, “I do feel that if you are the currency of a $70 billion business that you do have an obligation to get it right and to do it as quickly as you can,” Wurtzel said.

Nielsen is doing just that with the Total Audience product, Clarken said.

With Total Audience, programmers can track viewership for both content and ads across platforms, regardless of the ad model, whether or not a program is C3-eligible or which device is used to view it.

“That’s a really, really important step for the industry to take, and one that is gaining momentum,” Clarken said.

Another possible solution could be to change the way data is captured, Wurtzel said. For the most part, he added, Nielsen concentrates on measurement behind the screen — codes embedded in shows by content providers and then captured by the ratings firm. The problem with that so-called watermarking, he said, is that sometimes the codes don’t get passed through.

A more “front-of-the-screen” approach could alleviate that problem, Wurtzel said. Technologies that could be used include “audio fingerprinting,” which measures who is watching a given show and where they’re watching it by employing a device such as a smartphone to match the audio it “hears” with a cloud-based audio bank that identifies the program.

“It requires the measurement company to put it in place,” Wurtzel said. “At the end of the day, it is the responsibility of the company that owns the currency, and right now that is Nielsen, to provide the industry with what the industry has been looking for.”


Wurtzel and others would also like Nielsen to reconsider the current sampling method to calculate ratings. Current sample sizes of 50,000 to 55,000 respondents worked well in the days of fewer channels, but as viewing becomes more fragmented, the sample size would need to reach at least 1 million to accurately measure viewing, some have argued.

“It could very well be that the single-source panel is becoming obsolete,” Wurtzel said.

Clarken disagreed, adding that the sample methodology is the “judge and jury” of measurement data. “With any research, you need something that sits there that is representative to make the final assessment,” she said. “That big data set has to be calibrated, weighted, balanced [and] corrected.”

Massive online data sets are important, she added, but they must be matched against other sources to ensure that they give the correct snapshot of a viewer. Nielsen, she said, uses census data to determine certain metrics; has a relationship with social-media giant Facebook; and in March acquired Exelate, which aggregates and distributes third-party online data from more than 200 providers.

But even that data has to be weighed and corrected for things like shared devices, Clarken said, so buyers can determine who in the household is in front of that device at a specific time.

Most important in Nielsen’s mission, said Clarken, “is giving advertisers the confidence that the numbers that they get back to justify that spend are numbers that a third party independently has verified and can truly tell them that they got what the paid for.”