The two-week movie

Jeffrey Nemon goes to the cinema about once a week. He holds an AMC A-List subscription, follows the release calendar with the diligence of a man who reads the trades for pleasure, and pays attention, by his own account, to almost everything. In August he missed three films he wanted to see. The Rivals of Amziah King, a crime picture with Matthew McConaughey, played his local AMC for roughly a week. Teenage Sex and Death at Camp Miasma, a slasher that had been doing rather well, had been reduced to a single showing at ten o'clock at night. Tony, A24's Anthony Bourdain biopic, would be "practically gone after this weekend".

This is a peculiar sort of failure. Mr Nemon is not a lapsed customer to be recovered with a loyalty scheme. He is the loyalty scheme. The question his complaint provoked, among the producers and sales agents on the industry mailing list where he aired it, was whether cinema has contracted the disease of subscription television: whether, by cancelling things too quickly, the business has taught its most committed customers not to take a chance on anything unproven.

The analogy is to Netflix and its habit of killing series after a single season. The damage there was never any individual show. It was the aggregate lesson, that investing attention in a new programme is a poor bet, which made every subsequent launch harder, which in turn made cancellation likelier. Reputations of that kind outlive the behaviour that earns them.

Applied to the cinema, the charge is usually laid against the theatrical window, the stretch of exclusivity before a film can be rented at home. Squeeze it to a fortnight, the argument runs, and the rational moviegoer stops turning up. Why pay $13.17, the current American average, for something that will be on the living-room television in three weeks?

It is a good argument. It is also, at the moment, wrong, and wrong in an instructive way.

Windows are widening. They have been for two years. One analysis of 182 American releases through 2025, excluding Netflix titles and day-and-date experiments, put the average gap from cinema to digital at 32 or 33 days. Omdia, a research firm, tracked the average time to first paid rental at 39 days in 2025, up from 36 in 2024, with the ten biggest films averaging 51 days, up from 45. Twenty-two titles held out beyond 45 days, against 18 the year before. Then the commitments hardened. At CinemaCon in the spring every major studio recommitted to a 45-day floor. Universal, which had been triggering home rentals as early as the 17th day, moved to five weekends of exclusivity this year and seven from 2027. Disney is routinely past two months.

The film in Mr Nemon's complaint makes the point sharply. Tony opened on August 7th in six cinemas. A fortnight later A24 put it in 1,642, a jump of 568%. In its fourth weekend it reached its maximum of 1,907 and had taken $11m. No home date has been announced, and a 45-day window measured from the limited opening would not produce one before the last week of September. Whatever kept people away from Tony in August, it was not the prospect of watching Tony in August.

Which is the more interesting finding, because the belief persists even though the fact beneath it has expired. Between 2020 and 2024 the studios did train audiences to wait, and advertised the fact. Andrew Cripps, who runs distribution at Disney, has conceded as much; his studio now withholds all advertising for a film's streaming debut until immediately before it happens, which is an admission that the previous decade of marketing taught precisely the wrong lesson. The policy has changed. The reputation has not, and reputations decay far faster than they rebuild. In this narrow sense the comparison with Netflix is exact, and it flatters neither party.

And yet Mr Nemon's actual grievance had nothing to do with windows. He did not stay at home because a rental was imminent. He turned up to find the film had gone.

Consider Camp Miasma. MUBI, its distributor, opened it on the same day as Tony, in five cinemas, expanded it to 1,036 on August 21st, and watched it fall to 579 a week later. Nothing about its digital availability changed in those seven days. What changed is that 457 cinemas decided the screen was worth more to something else. The film was not windowed out from under its audience. It was booked out.

This is a different mechanism with a different remedy. A picture carrying 45 days of contractual exclusivity and a two-week practical life has an effective window of a fortnight, whatever the paperwork says. The exclusive period is still running. There is simply nowhere left to spend it. The customer in that situation has not chosen to wait; he was never offered a plausible chance to go. Supply-side scarcity is being read as demand-side apathy, and the distinction matters, because only one of the two can be fixed.

It matters for a second reason, which is that the aggregate numbers decline to co-operate with the pessimistic story. American box-office takings passed $7bn by the end of August, the summer alone accounting for $4.4bn, about 20% ahead of the same point last year though still some 7% shy of the 2019 pace. All this while the average window sits at roughly a month. Were quick home availability meaningfully suppressing attendance, this is not the shape the data would take.

What the survey evidence describes is not a population learning to wait but one that settled at home some years ago and stayed there. Polling by AP-NORC found that roughly three-quarters of American adults had watched a newly released film at home in the previous year, about 30% of them monthly. Two-thirds had seen a new release in a cinema at least once; only 16% did so monthly. Among the under-30s, nearly half stream new films every month and about a fifth attend that often. Cinema-going is not a default being eroded by streaming. It is an exception that must be earned, individually, every time.

Which is why withdrawing the alternative would not help much, and why marketing has quietly become the whole game. Buddy, a horror comedy no studio or streamer would buy out of the Sundance festival in January, was picked up in April by two small distributors, opened at the end of August to the best figures in Roadside Attractions' history and is tracking past $10m on a modest advertising spend. It won on attention, not on scarcity. Financiers have noticed. Agencies now advise clients to carry the marketing line inside the finance plan rather than hope a buyer supplies it later.

All of which bears on the announcement that set the mailing list off in the first place. AMC, the world's largest cinema chain, is starting a distribution label called Leawood Films, after the Kansas suburb where the company keeps its headquarters. It will take only films that arrive financed or finished. It will not develop or produce anything. Film-makers keep their intellectual property and their downstream revenue, on terms the company says may "differ markedly" from a studio deal. It honours a 45-day rental window and a 90-day streaming one and, contrary to the first assumption on the list, it will book across other circuits at home and abroad rather than confine itself to AMC's own screens. Adam Aron, the chief executive, says the aim is "to increase the number of small and mid-budget sized movies that get released theatrically". The first titles are unlikely before 2027.

The advisory bench repays a close reading, because it is a statement of intent. Toby Emmerich ran Warner Bros. Ricky Strauss ran marketing for Disney. Kyle Davies ran domestic distribution for Paramount. Craig Jacobson is an entertainment lawyer. Three of the four spent their careers releasing films. None of them spent it buying them. Leawood is not being staffed to acquire.

The obvious objection, raised within the hour, is that this describes a booking agent with better stationery. The producer still pays for prints and advertising, still sources his own ancillary sales, still carries the risk; AMC supplies access to infrastructure it happens to own already. "Gun for hire", as one participant put it, quoting the press release back at itself.

The objection is fair as far as it goes, but it misses the one thing AMC alone could sell. Opening-weekend screens are a commodity: any service company can buy them, and several do it well. What no service company can offer is a guarantee of hold, a contractual floor on playing time and, more to the point, on the hour of day a film is shown in its third and fourth weeks. No such product exists anywhere in the market. It is also the only thing that would have stopped Mr Nemon missing three films in a single month.

Which raises the question the mailing list asked first and never answered: how any of this differs from Open Road, and whether it matters if it does not.

It differs in exactly one respect, and the respect is the whole thing. Open Road was founded in March 2011 by Tom Ortenberg as a joint venture between AMC and Regal, then the two largest chains in America, on precisely the theory Mr Aron is now advancing: that exhibitors want more product on their screens and can put it there themselves. It released eight to ten films a year. It bought them, paid guarantees, financed the prints and advertising and kept the rights. In 2016 it won Best Picture with Spotlight. In August 2017 it was sold to Tang Media Partners, merged with IM Global, renamed Global Road and announced $1bn of production spending over three years. Thirteen months later its lenders seized control and it filed for Chapter 11. Raven Capital bought the wreckage for $87.5m.

Open Road is therefore not a precedent Leawood has overlooked. It is a precedent Leawood is designed around. Every element of the new label is the old one with the risk reversed: the film arrives financed rather than acquired, the producer carries the advertising rather than the distributor, the producer keeps the rights rather than surrendering them. AMC has examined what happens when an exhibitor tries to be a buyer and concluded, reasonably enough, that owning screens does not make a company good at valuing films, and that valuing films is the part that kills you.

So it matters that Leawood is different, and not in the producer's favour. It differs in precisely the way that protects AMC's balance sheet and in none of the ways that would relieve a film-maker's. Open Road's defect, seen from the seller's side of the table, was also its virtue: it wrote cheques. Leawood does not. What went into Chapter 11 in 2018 was the exhibitor-owned buyer, and nothing announced this month revives it.

Then there is Fathom, which complicates matters further, because AMC already co-owns a distributor. Fathom Entertainment, held jointly with Regal and Cinemark, grossed more than $145m in 2024, a rise of 45% on the year before, and dropped "Events" from its name in January 2025 to signal that it had outgrown the one-night engagement. It reviews some 3,000 submissions a year and releases fewer than 100. Classic re-releases have supplied about a fifth of its revenue for two decades; a 15-year-old Coraline took $34m. When The Blind, a faith-market biopic, sold out its seven-day run, Fathom extended it by six weeks.

Fathom is the most instructive of the three, because it sells the opposite of what everyone in this argument claims to want. It does not offer a long run. It offers a date. The engagement is short and, decisively, announced as short. Ray Nutt, the outgoing chief executive, describes the business without romance: "everything we do is incremental. We're bringing more bodies into movie theatres so they can sell more concessions."

For a producer choosing among the three, the distinctions are practical rather than philosophical. A finished film with an identifiable community behind it, a faith audience, an anime following, a horror fandom, a concert, a documentary with an organisation willing to fill the seats, belongs at Fathom, whose product is a date and an audience that assembles itself. A film that needs four weeks of word of mouth from a general audience is wrong for Fathom, and is the only real case for Leawood, on the condition that somebody extracts an answer about playing time before signing anything. And a producer who needs the distributor to supply the money is looking for Open Road, which no longer exists in that form and is not being rebuilt. That, rather than the label itself, is the news.

Whether AMC will sell a guarantee of hold is a separate question, since the company would be selling against itself. A label whose value rests on theatrical durability depends on the forbearance of the booking desk down the corridor, whose discipline about revenue per screen is what makes independent runs short in the first place. Both instincts are rational. Every cinema that dropped Camp Miasma was clearing its market correctly: a slasher taking a few hundred dollars a screen in its fourth week is not a good use of an auditorium. The cost of that decision simply appears on no individual manager's ledger.

Which returns us to Mr Nemon and his three missed films. The business has spent five years arguing about how long a picture should stay in cinemas, as though duration were the variable. Fathom has spent twenty years demonstrating that it is not. Its films play for a week, sometimes for a single night, and people turn up, because Fathom tells them exactly when the film goes away and sells that fact as the reason to come.

The two-week movie is not failing because two weeks is too short. It is failing because nobody says that it is two weeks. Mr Nemon did not decline to see Amziah King. He assumed, on the evidence of a nationwide release and a film star, that he had a month, and he was wrong by three weeks. A distributor who had told him he had nine days would have had his Saturday. AMC is now preparing to spend several years and a considerable sum establishing whether it can promise film-makers a longer run. The cheaper experiment would be to tell the audience how short the run already is.