The Binge Discount: Why Dropping a Whole Season at Once Can Cost a Platform 39% of Its Revenue

Based on the research ofZhao, Mehta and Shi, "An Empirical Investigation of a Digital Platform's Release Strategy for Serialized Content," Marketing Science, 2026

The Convenience Trap

The logic of the all-at-once drop feels unanswerable. Give people everything, let them consume it however fast they want, and you have maximized their satisfaction and your completion numbers in one move. Netflix built an entire brand on that premise starting in 2013, and its executives have defended it in almost identical language ever since. "There just is no data that weekly is better," Netflix's chief content officer Bela Bajaria told a UCLA law audience in 2023. "There's no data to support that weekly is better, and it's not as good of a consumer experience."

A new paper in Marketing Science says the data exist, and they say the opposite. Clarice Zhao, Nitin Mehta and Mengze Shi built a structural model of visits and purchases on a Chinese online book platform that holds a 70% share of the romance genre and sells its serialized novels chapter by chapter, using six months of 2017 transaction data covering 575 readers and 138 bestselling titles in the "Pure Love" category. The platform's actual practice was sequential release: chapters trickle out over time rather than arriving as a finished book. The authors then asked what would happen to revenue if that same platform switched to releasing every chapter the moment it was ready, mimicking a pure binge drop. The answer: a 39% fall in platform revenue, driven by a 60% collapse in the number of times readers came back to visit at all.

That number is large enough to be the whole argument. Giving readers everything up front did exactly what intuition predicts for the single book in front of them, it let them binge it. But a platform doesn't make money from one book. It makes money from a reader who keeps opening the app, and a reader with nothing new to look for stops opening it.

Why Returning Beats Devouring

The mechanism the paper isolates is not mysterious once stated plainly. Simultaneous release induces binge consumption: when more chapters are sitting there unread, readers burn through them faster, and faster consumption this period raises the odds of consumption next period, which raises the odds a reader finishes, and pays for, the whole book. Sequential release does something different: it forces readers to come back to get the next installment, and each return trip is a chance to browse other titles they are not yet following. In the data, a notification that new chapters are available for a book a reader already follows roughly doubles that reader's probability of visiting the platform that day. Binge consumption sells the book in front of you. Exploration sells every other book on the platform. A platform that only optimizes for the first effect is leaving the second one, which is the larger one, unharvested.

A randomized field experiment on a real video-on-demand platform reaches the same conclusion through a different door. Miguel Godinho de Matos, Samir Mamadehussene and Pedro Ferreira ran a five-week trial with a major multinational telecom provider's streaming service, randomly assigning 84,000 viewers to either a gradual "drip" release schedule or an all-at-once release of the same shows. Viewers in the drip group were 48% more likely to keep using the platform than viewers who got everything at once. At the end of the trial, drip-release viewers were 1.7 percentage points more likely to subscribe, a 3.48% baseline subscription rate for the all-at-once group versus roughly 5.18% for the drip group, a 48% relative lift. "The moment all-at-once viewers finish a fully released show, they often leave the platform," de Matos said when the study was published. "A drip schedule keeps viewers engaged for weeks, giving them time to search, browse, and find other shows they enjoy." Mamadehussene put the mechanism in almost the same words Zhao, Mehta and Shi use for a book platform: "Releasing episodes slowly creates natural touchpoints that bring viewers back each week. Those repeated visits dramatically expand content discovery and strengthen retention."

The Buzz That Doesn't Outlast the Weekend

The revenue argument matches what independent demand analysts have been documenting in television for years, measured a different way: not through a platform's internal visit logs, but through public search and viewing interest. Ampere Analysis compared the top Netflix shows, released as full seasons, against the top HBO shows, released weekly, and tracked each show's "Popularity Score" against its own premiere-month level. The Netflix titles fell to 80% of their premiere-month popularity within a single month; the HBO titles took four months to decay by the same amount. "A weekly release pattern more easily facilitates conversation around a show," Ampere's Rahul Patel explained. "Between episodes, viewers have ample time to discuss and rewatch episodes, which is less likely to be the case if an entire season is released together."

Parrot Analytics' longer-run demand data sharpens the same pattern at scale. Looking at the 50 best post-finale "decay rates" for a TV season, how slowly demand falls off after the last episode airs, weekly and periodic releases combined account for 48 of the top 50 since 2016; for the decay rate during the active release window itself, they account for 44 of the top 50. Of the top 50 seasons ranked by how many weeks they stayed above their own seasonal-average demand, weekly releases alone claim 43 spots. Warner Bros. Discovery's then-CEO David Zaslav drew the conclusion bluntly in February 2024:

"The best way to drive interest and engagement is not by dropping the entire season on a platform all at once, but by allowing the buzz and anticipation to build over time."

None of this erases binge release's real upside. Parrot Analytics credits the format with helping turn unfamiliar titles into fast global hits, Squid Game ranked 200th in worldwide demand on its premiere day and hit number one nine days later, a velocity a weekly rollout structurally cannot match, and Netflix's own Q3 2022 shareholder letter argued it is "hard to imagine how a Korean title like Squid Game would have become a mega hit globally without the momentum that came from people being able to binge it." The honest reading of the evidence isn't that binge is wrong. It's that binge is a hook, not a business model, and a platform that stops at the hook is giving up the months of return visits that a metered release would have captured instead.

The Hybrid Payoff, By the Numbers

This is where Zhao, Mehta and Shi's counterfactuals become genuinely actionable rather than merely cautionary. Relative to the book platform's actual sequential practice, a cleaner pure-sequential policy, releasing exactly one chapter per day for every title, would raise revenue by about 5%. But the revenue-maximizing policy in their model is neither pure strategy. It is a hybrid: release the first 40% of a book's chapters simultaneously, then release the remaining 60% sequentially, one chapter a day. That hybrid design generates 16% more platform revenue than the status quo, and far more than either pure extreme. The reason is that the hybrid turns the two effects from rivals into partners: the simultaneous first 40% is enough to hook a reader into the story, raising the odds they finish it, while the sequential remainder keeps forcing them back to the platform, raising the odds they explore everything else on it.

The television industry is visibly groping toward the same architecture, without yet landing on one number. Fabric Data's 2025 tracking shows Max cutting all-at-once releases to just 11% of its slate, down from 50% in prior periods, while running "The Pitt" entirely on a weekly schedule; Disney+ split its releases almost evenly across four distinct formats in the first quarter of 2025, including straight weekly drops and weekly releases with a multi-episode preview; and Prime Video applied a hybrid directly from Zhao, Mehta and Shi's playbook to "Reacher" season three, releasing the first three episodes together before switching to weekly installments. Luminate's 2026 release-strategy report shows the same split industry-wide: among the seven major U.S. streamers besides Netflix, the weekly share of original premieres rose from 57% in 2024 to 61% through mid-2026 while binge's share fell from 43% to 38%. Netflix itself is the outlier that proves the point is about trade-offs, not ideology: more than 80% of its original seasons from 2024 through the first half of 2026 still dropped as full-season binges, and Prime Video's own all-at-once share actually rose from 33% to 67% of its 2025 slate even as it ran hybrid experiments elsewhere. Nobody in this industry has settled on a single answer, because the right split is an empirical question specific to a title and a platform's economics, exactly the kind of question Zhao, Mehta and Shi's model is built to answer, and exactly why their 40% figure is a finding from one book platform's romance catalog in 2017, not a universal constant to import unchanged.

The Rule for Any Serial Platform

The operator's real decision was never binge versus weekly. It was always how much to give away up front to create a hook, and how much to meter afterward to create a reason to return, and the data, from a Chinese romance-reading app to an 84,000-viewer streaming experiment to a decade of global demand tracking, agree that the two halves of that decision are not in tension when you size them correctly. They are complementary. A platform that releases everything maximizes one visit and forfeits the next twenty. A platform that releases nothing until next week maximizes the twenty visits but risks losing the viewer before the hook lands at all. The 39% figure is the cost of picking only the first half of that trade; the 16% figure is the reward for building a release schedule that, like the shows themselves, has both an opening and a middle.