When Netflix started reporting its own ratings — as it has for films like “Bird Box” and TV shows like “You” — the entertainment industry jumped on the numbers like a kid opening a gift that’s been wrapped and under the tree for months. Now, the lovingly appointed “mayor of television” is here to tell us why the prize inside may not be exactly what we thought.
Speaking at the FX TCA panel Monday morning, CEO John Landgraf said, “Netflix is not telling you the full story.” The key difference is the measurement metric used by Netflix isn’t the same as the traditional television ratings metric — average audience. Adopted by FX, along with most cable and broadcast networks, the average audience is calculated by adding up every minute viewed of an episode (or season) and dividing it by the total duration of the program.
But that’s not the figure Netflix reported.
“Many Silicon Valley companies have long used various metrics used on video views or video start-ups, which inflates their views and is not remotely accurate representation of a long-form program’s performance,” Landgraf said, arguing that the average audience measurement is “the methodology that best represents an apples to apples comparison of viewership — not just how many people are watching but how many people watch, how often, and for how long.”
Netflix counted each view based on when an account watched at least 70 percent of one episode. So, per Landgraf, the 40 million accounts that have viewed “You” equates to a much smaller audience who’ve kept watching. “Really, it’s an average audience of 8 million viewers,” Landgraf said, “which is very good” but doesn’t equate to the No. 1 show in America — as an apples to apples comparison of Netflix’s report with traditional metrics would indicate.
“An average audience of 40 million would not only make ‘You’ and ‘Sex Education’ the two most watched shows on television, it would give Netflix two original series hits to join their one true breakout, which is ‘Stranger Things,'” Landgraf said. “But 40 million people are not watching each episode of ‘You’ or ‘Sex Education.'”
To that end, he pointed to Netflix’s projection that “Sex Education” would hit 40 million views, as well. In average audience, Nielsen measured just 3 million U.S. viewers per episode.
Moreover, Landgraf accused Netflix of inflating perceptions of their viewership overall. He argued the streaming giant has been “reporting their singles like home runs” when “Stranger Things” remains their only true home run.
“When you take hundreds of at bats, you’re going to get some singles, some doubles, and even the occasional home run. ‘Stranger Things’ is absolutely a home run, but it’s the only one,” Landgraf said. “You’re [also] going to get a high number of strikeouts.”
Landgraf conceded Netflix “has some very good shows” and “a number of shows that are hits,” but said “their true batting average would be seen as unimpressive.” Similarly, Landgraf stressed the critical acclaim FX receives for most of their programs, as well as the lack of a comparable qualitative batting average for Netflix. Per FX, the streaming giant produced 530 shows in 2018, and only 62 were included on any critic’s year-end best list. In comparison, 13 of FX’s 14 original programs made at least one critic’s rankings, while HBO landed 20 of their 70 programs on a best-of list.
He also stressed the importance of the press in being diligent when reporting figures that aren’t supported by third-party ratings services. He then explained exactly how Nielsen measures ratings:
“What Nielsen does is it captures the audio signal from a program that’s airing inside the home,” Landgraf said. “Any time that program is played inside a Nielsen home, it’s capturing that.”
What it can’t do is tell the difference between a movie playing on a DVD player and a movie playing on Netflix. For example, if the service hears Disney’s “Moana” playing in the living room, it will count it toward the Netflix ratings’ tally even if it’s not playing through the service. “They’re gonna have to put a streaming meter in there to differentiate the source,” Landgraf said, before pointing out that the only way to play a Netflix original series is through the service, so there’s much less room for error there.
Nielsen also can’t measure views played outside the home — aka when someone watches Netflix on their phone or computer, either using headphones or when they’re away from the Nielsen meter. But Landgraf said, using data he’s looked at through Hulu, that would only account for 10-15 percent of data.
“I think the [Nielsen] data is very accurate. Netflix will probably tell you it’s not, but I think it is,” Landgraf said. […] “One way or another, the truth will come out, as it always does.”