MoviePass Wasn't a Naive Bet. Two CEOs Pleaded Guilty to Calling It One.

In the spring of 2018, MoviePass was handing theaters about $12.02 for a ticket — and had agreed to collect, from the person using that ticket, $9.95 for the entire month, no matter how many movies they saw.³ Picture the cash register: every time a subscriber walked into a screening, the company lost money the instant the lights went down, and the heaviest fans lost it the fastest. The more people loved the product, the more it cost. That is not a flaw in the model. For most businesses it would be a death sentence. For this one, it turned out to be the point.

The story everyone tells about MoviePass is a tragedy of ambition: two bold entrepreneurs slashed the price to $9.95, bet they'd make it back selling data and ad deals, and got the math heartbreakingly wrong. It's a generous story. It is also, by the admission of both men in federal court, false. They didn't get the math wrong. They knew exactly what the math was — and used it anyway, to do something the math was never meant to do.

What the $9.95 was actually for

On August 15, 2017, MoviePass cut its plan to $9.95 a month for unlimited movies. The day before, it had roughly 20,000 subscribers. Sixty days later it had blown past 600,000.¹ That spike is the whole story in one number — and it tells you what the price was designed to produce, which was not profit but momentum. The cut landed at the exact moment Helios and Matheson Analytics, a thinly-traded company on the Nasdaq, took its majority stake; within weeks HMNY was raising its own purchase price for the company and citing, by name, 'the significant and rapid increase in subscribers since MoviePass announced its new $9.95 per month subscription fee.'² The subscriber count wasn't a business metric. It was a press release for a stock.

Here's the part the naive-visionary story gets backwards. The thesis everyone repeats — that MoviePass would monetize its troves of viewing data with AI — wasn't a sincere plan that failed. The DOJ established that Farnsworth falsely claimed HMNY possessed AI and big-data capabilities it did not have, while telling investors the unlimited plan was 'tested, sustainable, and profitable or break-even.'⁵ Both pillars of the redemption narrative — the data play and the someday-it-pencils economics — were, in the government's words and the defendants' own pleas, lies told to move a share price.

The math was worse than 'a ticket costs more than $9.95'

The popular autopsy says the model was doomed because a movie ticket costs more than $9.95. That's the lazy version, and it's wrong in a way that matters. Through the third quarter of 2017 the average U.S. ticket ran about $8.60 — below the monthly fee.⁷ If every subscriber had gone to exactly one movie a month, the plan would have nearly broken even on tickets alone. The model didn't break on the first ticket. It broke on the second.

This is the structural trap of unlimited pricing, and it isn't subtle. A flat fee is a magnet for exactly the customers who destroy it. The light user who'd go once a month was a small win; the cinephile who'd go ten times was a catastrophe — and the $9.95 price was engineered to attract the cinephile. So the company did the only thing a fraud can do when usage threatens to expose the lie: it hid the usage. Prosecutors say executives directed employees to throttle high-usage subscribers — quietly blocking the heaviest fans — while telling the public that demand was 'naturally declining.'⁶ The throttling wasn't a desperate operational fix. It was a cover story.

But couldn't it have worked, like Costco or Amazon Prime?

The honest objection is that plenty of great businesses sell below cost on purpose. Costco's hot dog loses money so the membership can win; Prime's free shipping bleeds so the cart fills up. So why dismiss MoviePass as fraud rather than a loss-leader that simply ran out of runway? Because a real loss-leader is a subsidy with a destination — a second product, profitable enough to carry the first. MoviePass had no second product. The 'data monetization' that was supposed to be the profitable far side of the subsidy was, per the DOJ, a capability the company didn't possess and lied about possessing.⁵ A loss-leader without a profit center behind it isn't a strategy; it's just a leak. And the tell is in what the company chose to optimize: not retention, not the second product, but the subscriber count that goosed a share price — and when usage threatened that count, it hid the usage rather than fixing the economics.⁶ You don't quietly throttle your best customers if you're building a business. You do it if you're protecting a story.

The numbers caught up the way they always do. HMNY posted an operating loss of $126.6 million for a single quarter in 2018, and when it disclosed in July that it couldn't pay the merchants it owed, the stock lost 96% of its value in short order; shareholders sued, alleging the model was never sustainable.⁴ The service shut down in September 2019, and the parent company filed for Chapter 7 bankruptcy in January 2020.⁸ The lawsuits were right. The model was never sustainable — and the men running it knew.

The cruelest detail is how cheap the corpse turned out to be. After the bankruptcy, MoviePass's co-founder bought the brand back out of the wreckage for $140,000.⁸ The DOJ pegged investor harm from the scheme in the hundreds of millions. That is the real spread MoviePass produced: not $9.95 against a ticket, but a fortune in shareholder money against a name worth less than a studio apartment. The plan was never built to make the math work. It was built to make the math irrelevant for exactly as long as the stock kept climbing — and the moment the climbing stopped, there was nothing underneath it at all.