What makes a TV show popular? What even counts as a TV show anymore?
The explosion in quantity, quality, and providers of television collectively known as “Peak TV” has effectively vaporized the meaning of what used to be basic descriptors. What’s a comedy? What’s a network? And what’s true for everything else in TV is especially true when it comes to ratings. In the age of DVR viewing, streaming services, and the steady erosion of individual shows’ audiences that only Empire and Game of Thrones can seemingly shake, it’s harder than ever to tell how many people are watching a show, or even where they’re doing it — let alone decide on a universal metric for what constitutes a hit, or a bomb, or just good enough to get renewed.
Once upon a time, Nielsen polled a carefully composed subset of American households, crunched the numbers, and handed down a set of statistics generally agreed upon by the people who make, showcase, and sell their wares on television. Networks knew what was working. (Or wasn’t.) Studios, producers, and actors knew how much they were worth. (Or weren’t.) Audiences knew what was popular. (Or wasn’t.) And advertisers knew where they could guarantee eyeballs for their commercials. (Or couldn’t.)
Now, Netflix and its accompanying market forces, like delayed viewing and independence from advertising, have thrown a toolbox worth of wrenches into that carefully calibrated system. For the production studios that make TV, the networks that air it, and the advertisers that want to profit from it, this leaves a mess of open questions, not all of which can be answered by Nielsen, the firm still synonymous with American ratings. For a small but growing number of companies, this presents an opportunity. And for the rest of us, it means TV’s major players are making a good-faith effort to figure out what we’re really watching, and hopefully make more of it.
One of those companies is Symphony Advanced Media, a five-year-old, San Francisco-based group the layperson might recognize as the reason we know Fuller House is a megahit, despite Netflix’s refusal to share statistics. Some of their numbers seem literally unbelievable — were more than 14 million people really willing to watch an Olsen-less ’90s reboot — and designed to take advantage of a gaping black hole in the ratings business. But those numbers are only part of the story, excluding both some of what Symphony does and how it does it. DVR’d episodes and Netflix shows account for an ever-larger part of our media diets, but they often don’t show up in Nielsen ratings. Symphony is trying to figure out where else people watch TV besides … actual TV.
That’s the “what.” The “how” is Symphony’s main selling point, and the reason it bills itself as a tech startup as much as a ratings company. (Well, that, and because calling itself a tech startup lets Symphony raise venture capital.) To get its numbers, Nielsen uses several methods to figure out what people are watching and when, including people meters (that thing attached to TV sets you generally think of when you think of Nielsen), set-top-box data, viewing diaries, and an audio recognition device called the Portable People Meter. Symphony uses just one: an app, downloaded to the cellphones of its more than 15,000 panelists. Audio recognition software then picks up whatever people are tuning into, wherever they’re tuning into it: their TV sets, their laptops, or their smartphones. “[It] measures everything you want to measure from one approach,” says Bill Harvey, a media research consultant who’s worked with Symphony as well as networks and advertisers.
VideoPulse, Symphony’s syndicated, cross-platform ratings report, has been up and running for paying clients since September 2015. (Symphony does not disclose specific customers.) The methodology is still a work in progress, and Symphony president and CEO Charles Buchwalter recognizes that paid smartphone panels tend to undersample older viewers, who don’t use technology as much, and wealthier viewers, who don’t need the money. But during a months-long pilot phase, Symphony offered the data for free to participants like NBC, Viacom, and Warner Bros., who then offered feedback — and compared Symphony to the competition.
Buchwalter reports that most of the pilot clients signed on when Symphony went commercial, which is why he’s confident in data that isn’t directly comparable to Nielsen’s, because Nielsen isn’t collecting it — like Netflix ratings, even though the company dismissed Symphony’s findings as “remarkably inaccurate” after NBC research head Alan Wurtzel rolled them out at January’s Television Critics Association press tour. Of course, it’s in Netflix’s interest to maintain maximum secrecy: Netflix has nothing to gain from advertisers it doesn’t have, and plenty to lose in negotiating with studios and talent currently missing a crucial metric of what they’re worth.
Symphony also offers a metric called “live plus thirty-five,” or L+35 for short: the number of people who watch a given show within 35 days after its initial air date. Nielsen’s longest-lead metric, L+7, cuts off at just a week after air — and it’s the number flagging shows will point to as a reason for renewal, despite poor same-day ratings. Even then, Symphony estimates this excludes up to a quarter of TV viewing overall, and possibly even half the viewership of specific, younger-skewing shows like Jane the Virgin or Bob’s Burgers. That recent but massive shift in viewing habits will only accelerate as those younger viewers grow up — but it’s a shift that Nielsen isn’t fully accounting for, and advertisers therefore aren’t paying for. “There’s this … industry behavior pattern in which, on one side of their brain, [advertisers] say, I’m gonna pay you and you’re gonna accept the payment based on what Nielsen says the audience is,” Harvey says. “And on the other side of the brain, they say, ‘Of course, we know this isn’t the whole audience.’ … [Networks] are not getting paid for those other impressions, ’cause they’re not being measured.”
It’s true that Symphony offers one of precious few glimpses into the notorious black box that is Netflix (and to a lesser extent, Amazon and Hulu), a fact it cannily takes advantage of by releasing its findings not just to paying clients, but to the press. If people are dying to know just how far an investment of at least $120 million goes and Netflix won’t say, Symphony is happy to fill that information gap — and get its name out in the process. (The answer is just over 3 million viewers ages 18 to 49, which is to say, not as far as Netflix might have hoped.)
Buchwalter believes Symphony’s biggest asset isn’t a single data point, but the context a critical mass of them can provide: a way to combine separate insights like L+35 numbers and how many people watch Transparent into a holistic portrait of the TV industry that’s more indicative of how people actually experience TV. They’ve found that people who watch Netflix’s Jessica Jones also watch ABC’s Agents of S.H.I.E.L.D., and about 10 percent of shows’ new viewers in a given season are viewers who’ve caught up with the show on a platform like Hulu. So-called “linear” TV — shows that air live and people watch week by week — isn’t separate from Symphony’s specialty, but in conversation with it. Symphony is essentially selling a transcript of that conversation, and the transcript shows an industry that isn’t fragmented into streaming and linear, or Netflix and network. TV’s factions aren’t so much competitive as symbiotic. Put more selfishly: The streaming bump is real, and there’s a chance your favorite show could end up on Amazon — or better yet, renewed — because of it.
Parrot Analytics, a four-year-old data science firm founded in New Zealand, is also attempting to fill in the widening cracks in traditional ratings — and coining new metrics of its own in the process. Parrot’s focus is more global than Symphony’s or even Nielsen’s, aiming to gauge what it calls “demand for content” country by country. It’s another way to fix the problem of studios basically flying blind when it comes to distributing — and racking up licensing fees for — their shows internationally. Under Parrot’s system, the company claims, there’s now a way to directly gauge which American TV shows are destined to be huge in Japan.
But Parrot’s technology gives it just as much insight into how many viewers are watching Stranger Things right here in the States as abroad. Parrot departs from Nielsen and Symphony alike in that it does not use carefully composed, statistically representative panels. Instead, Parrot devised a way to go straight to the source, tracking and weighting online behavior like social media engagement, illegal downloads, and binging to measure total “demand expressions” for a given show. They’re not perfect metrics; “active online” is already a skewed subset of the larger TV-viewing population. But Parrot aims to make up for in quantity what panel-based methods have in statistically calibrated quality.
“Even if you were to panel or survey 30,000, 40,000, 50,000 people in the U.S.,” says founder and CEO Wared Seger, “you’re still extrapolating to a population that is orders of magnitude larger than that number.” Parrot processes over a billion data points a day and compiles them into a country-specific “demand rating.” According to Parrot, there were about 39 million demand expressions for Stranger Things in the U.S. the week after its July release, compared to just over 22 million for The Get Down. It gives a number to what we already knew, if only in the abstract: Stranger Things is a hit, and The Get Down failed to hold our attention. That’s still more information than we had before — and more than Netflix was ever planning on giving.
Like Symphony, Parrot deals with data that didn’t exist 10 years ago about viewing platforms that didn’t exist then, either. The studios and distributors these companies work with aren’t just interested in dirt on the competition (though NBC’s Wurtzel was certainly happy to tell his TCA audience that most streaming shows fall short of broadcast numbers). Just as important, they’re looking for information required to do business. Wurtzel’s NBC, for example, owns Netflix original Unbreakable Kimmy Schmidt through its studio arm NBCUniversal Television, which means it’s regularly haggling over how much Netflix should pay for the privilege of airing its show. And those transactions are one way studios are making up for the revenue their parent companies are losing in ads.
The reason streaming services don’t provide viewership data, says Bruce Tuchman, former president of AMC Global and a current Parrot board member, is “not just because they’re not advertising-based. It gives them leverage in negotiation. … Before, when you had ratings data, you know how your show performs,” and could demand compensation accordingly. Now, says Tuchman, “the playing field’s not even.” Which is where companies like Symphony and Tuchman’s own Parrot come in: Nielsen isn’t providing streaming insight quickly enough, so smaller companies spring up to meet demand.
Despite Symphony and Parrot’s innovations, though, Nielsen remains the sprawling juggernaut in the room. Part of this has to do with the basic infrastructure of the business; Nielsen has staggered, eight-year contracts with many of its clients, meaning there’s no critical inflection point on the horizon and that a massive portion of companies’ research budget is booked for almost a decade at a time. There’s also the simple inertia and the institutional advantage that comes with incumbency. (You’re not the only one who thought the Fuller House numbers sounded made up.)
But a theme emerges from discussions with both companies: The goal isn’t to replace Nielsen — it’s to supplement it. Just as TV itself has splintered, ending the unquestioned dominance of broadcast networks and offering ever more options for viewers, so has the business of measuring TV. “You can’t just buy Nielsen,” Harvey says. “That doesn’t cover all your needs. You have to buy other things. You have to be realistic, and that is the truth.” What holds true for the rest of media holds true here: The singular information source is a thing of the past. Streaming isn’t going anywhere. It’s time to understand it and account for it — and most importantly, make money off of it.