If its ratings don’t really matter, why is Netflix suddenly canceling so many shows?

Reed Hastings tends to be sanguine when asked about the video streaming service Netflix’s notorious refusal to release its ratings.

In May 2017, speaking shortly after the cancellation of Baz Lurhmann’s long-awaited 1970s musical drama “The Get Down” after just one series, Hastings, Netflix’s co-founder, chairman and CEO, told CNBC’s “Squawk Alley” that he felt “our hit ratio is way too high right now. ... I’m always pushing the content team, ‘We have to take more risk. You have to try more crazy things.’ Because we should have a higher cancel rate overall.”

In the year since he made those remarks, Hastings has gotten his wish: 13 shows have been canceled over the course of the last 12 months. At the heart of the problem is the ratings issue.

Ever since the American-Norwegian crime drama “Lilyhammer” became Netflix’s first original series to premiere in 2012, the company has been committed to voraciously generating new content in addition to its huge library of existing TV shows and movies.

Netflix’s business model is centered around its 125 million subscribers, not advertisers; user base is the company’s focus, not viewership.

Netflix, with a current market cap of $145.8 billion, expects to spend as much as $8 billion on original and licensed content in 2018 and is aiming to release 700 original series in addition to 80 original movies in 2018.

The company’s disruption of the traditional TV business extends to its refusal to release ratings. Netflix’s business model is centered around its 125 million subscribers, not advertisers; user base is the company’s focus, not viewership.

And yet as its volume of original content has rapidly increased, numerous shows have misfired. While Netflix has enjoyed recent success with “Stranger Things” and “13 Reasons Why,” 2017-18 has also featured a litany of failures.

One industry source who works for a digital data company and spoke on condition of anonymity said he knew of three Netflix shows that failed to attract 1 million U.S. viewers in the first seven days of release: “Gypsy,” a drama starring Naomi Watts and Billy Crudup; “Girlboss,” a comedy inspired by Sophia Amoruso’s business memoir; and “Flaked,” a dramatic comedy starring Will Arnett about a troubled self-help guru.

“Gypsy,” produced by Universal Television, was ditched six weeks after its premiere. “Girlboss” was canceled after one season. The second season of “Flaked” was released earlier this month, but industry observers anticipate it will become Netflix’s 20th canceled original show.

Much-hyped shows that went on to be canceled, including “The Get Down” and Judd Apatow’s relationship drama “Love,” have experienced ratings declines over the course of their runs, the source added.

Ratings do count

Analysts and industry sources think that for all its public disdain for ratings, Netflix is more like a conventional television network than it cares to admit.

“If a Netflix show had a large viewership, it would make no sense to cancel the show,” Michael Pachter, a digital media analyst with Wedbush Securities told MarketWatch. He added: “The shows that have been canceled (“Marco Polo,” “Sense8,” “Bloodline,” The Get Down,” “Lady Dynamite”) had [varied] reviews so the only possible common link is their low viewership.”

But one producer on a Netflix original prestige period drama, who wished to remain anonymous, said he had “never had a conversation about ratings with Netflix. I’ve got no idea how many people watch the show, as I’m paid a premium based on a percentage of the original budget.”

Jeffrey Lodgson, a media analyst and managing director of JBL Advisors, said stockpiling series was essential to Netflix’s success. “Netflix appears to be almost exclusively judged and valued by traders and investors based on paid subscriber additions rather than more traditional valuation tools (EBITDA, earnings or cash flow) so adding an almost unprecedented number of new series each of the past few years to find a handful that build audience loyalty and attract new subscribers globally is the current success quotient for Netflix’s stock.

“A series that is not gaining audience stickiness (or seemingly has a pathway through story or talent additions) gets canceled even though some portion of the subscriber base may have connected with a show and are not pleased that it disappears.”

Despite the recent setbacks, Netflix’s cancellation rate is still notably lower than those of traditional television networks. “The failure rate for network TV is over 50 [percent],” said Michael Pachter at Wedbush. “I don’t for a second believe that Netflix is better at producing content than its competitors. On the contrary, I think it is about the same and that it continues to chase bad programming decisions after many of its shows have been canceled.”

The fact that an estimated 90% of Netflix originals are under license deals, rather than fully financed by the company, he said, “binds Netflix to producing two or more seasons, and they are forced by their own contracts to throw good money at a failing show.”

Pachter added that Netflix shows “earn a return after the third or fourth season of a series as it has built an audience and word of mouth begins to kick in. I do not believe they get a return on comedy specials and movies, since the content is relatively costly to produce.”

“I think it’s a fair inference that cancellations of Netflix content mean that content did not meet whatever ‘ratings’ or, more specifically, the viewership bar that Netflix sets,” said Joe Bonner, an analyst at Argus Research.

“But perhaps Netflix is not that concerned with ‘ratings’ or viewership but rather more about creating content that generates cultural ‘buzz.’ Think ‘House of Cards’ and ‘Orange Is the New Black’ and, more recently, ‘Stranger Things.’ Buzz might generate interest in subscribing, which is what management really cares about.”

Yet for all the good buzz does in terms of perception, it’s viewership that will reduce Netflix user’s dependency on licensed content. (According to a study by 7Park Data, 80% of Netflix U.S. viewing is of licensed content, with 20% going to original shows.)

Competing for eyes

Among the companies that have tried to track Netflix’s ratings, Nielsen has gotten the most traction. Since last October, Nielsen has been tracking how many people are watching Netflix shows by using what’s called Subscription Video on Demand (SVOD) Content Ratings. Nielsen now measures streaming content in comparable ways to how it tracks traditional TV networks offering audience demographic data, reach and frequency.

Nielsen estimated 15.8 million viewers watched the premiere episode of Netflix’s “Stranger Things 2” within three days of its launch last fall, 11 million watched the Will Smith movie “Bright” within the three-day post-release period, and most recently “Lost in Space” drew 6.3 million viewers within that initial three-day span.

“Netflix has changed how consumers are thinking about content — when people are watching things on Netflix, they are there for an extended period,” said Brian Fuhrer, Nielsen’s senior vice president of product leadership. “With Netflix, each program has its own niche viewing target. ... [T]hey are interested in retaining existing subscribers but also in reaching across to people that traditionally haven’t subscribed.”

“A good example of that was ‘The Crown,’ which was [attracting] an older viewing audience but one a little bit outside the traditional Netflix viewer.”

While Netflix has disputed Nielsen’s estimates, saying the ratings don’t take into account mobile-device viewing, Nielsen’s Fuhrer said the system is popular among conventional TV network clients: “If they’re noticing their programming is down this week, they’re able to look in our systems to see their program was competing with a very different landscape because of a Netflix original that had dropped.”

Glowing reviews

Netflix is looking to reduce its cancellation ratio with impending second series of the women’s wrestling comedy “GLOW” and Walt Disney Co.–owned Marvel’s “Luke Cage”as well as premiering “Simpsons” creator Matt Groening’s new series “Disenchantment” in August.

It is also increasingly adopting a global mind set. Seven new European shows will debut on Netflix in the second half of this year. “Our belief is that great storytelling transcends borders. ... [Netflix is] committed to being a voice for European entertainment,” Netflix Chief Content Officer Ted Sarandos said at a Rome event in April.

The global aspect of President Barack Obama and first lady Michelle Obama’s star power partly accounts for Netflix’s signing them up for a multiyear production deal.

Analysts doubt Netflix ratings will be released anytime soon. “That would not serve Netflix’s plan or perspective on the entertainment universe,” Logsdon of JBL Advisors said, adding that “talent may try to use viewership numbers to extract higher compensation than a more simplistic renewal process.”

But whatever the bruised creative feelings when shows get canceled, “the talent” seems quite happy with the ratings status quo.

“It was liberating to never know the ratings of the show,” said actor Norbert Leo Butz, who starred in the Florida-set thriller “Bloodline,” which was canceled in September 2016 after three seasons on Netflix. “Not having to work in that environment definitely reduced the pressure while we were making it.”

Netflix shares have gained 74% in 2018 so far, while the S&P 500 has gained 1.6% and the Dow Jones Industrial Average is flat.