How Streaming Services Decide Which Shows to Cancel

It's not about ratings. It's not about reviews. Streaming services use specific metrics to decide a show's fate — and none of them measure what viewers actually feel.

When a beloved show like The OA or Archive 81 is cancelled, fans often ask the same question: how could they cancel a show that so many people loved?

The answer lies in the specific metrics streaming platforms use — metrics that don't measure passion, cultural impact, or critical acclaim. According to a 2022 exposé by The Hollywood Reporter, Netflix's internal decision-making process involves a data dashboard known internally as the "Greenlight Report," which ranks every piece of content by a proprietary combination of completion rate, cost-per-viewing-hour, and subscriber engagement velocity. Here's how streaming services actually decide which shows get cancelled.

1. Completion Rate

The most important metric. What percentage of viewers who start a season actually finish it? Netflix reportedly uses a 70% completion rate as a key threshold for renewal consideration. If fewer than 70% of viewers make it to the final episode, a show is in danger — regardless of how many people started watching. According to a 2023 report by Variety, Netflix's exact thresholds are proprietary, but former executives have confirmed that completion rate is "the single most important number in the building."

This is why shows with slow starts are vulnerable. Even if they build to spectacular finales, viewers who drop off in Episode 2 are counted against the show. The OA had a deliberately slow, mysterious opening — a narrative choice that may have hurt its completion rate. Season 2 of The OA was famously ambitious, but viewers who didn't make it through the first few episodes of Season 1 never got to experience where the story was going. The gap between critical enthusiasm and completion rate may explain why the show was cancelled despite its passionate fanbase.

2. Cost Per Viewing Hour

Netflix knows exactly what each show costs per hour watched. A $100 million season that generates 200 million viewing hours costs $0.50 per hour. A $10 million season that generates 5 million hours costs $2.00 per hour. The cheaper show per hour is actually more valuable — even if the expensive show has bigger raw numbers. According to a 2023 Bloomberg investigation, Netflix executives use a metric called "efficiency ratio" that divides a show's total budget by its projected viewing hours. Shows that fall below a certain efficiency threshold are flagged for cancellation regardless of their cultural footprint.

This is why Archive 81 was cancelled despite strong viewership numbers. Its production budget was estimated at $4 million per episode — very high for a horror series. According to Deadline, the show had impressive raw viewing numbers but a poor cost-per-viewing-hour ratio compared to less expensive titles in Netflix's horror library. This metric is also why the two-season curse claims so many shows — a show may need two seasons to find its audience, but the cost-per-viewing-hour is calculated after just one.

The same logic explains why Stranger Things, with its massive budget, never faced cancellation. Its viewing hours are proportionally enormous, making its efficiency ratio among the best in Netflix's library. A show with half the budget but one-tenth the viewership is, paradoxically, at greater risk.

3. Subscriber Acquisition

Did a new show bring in new subscribers? This is harder to measure, but platforms try. If a show drives sign-ups — especially in specific demographics or regions — it has a stronger case for renewal. According to TVLine, Netflix's subscriber acquisition data is broken down by title, region, and demographic, allowing executives to see exactly which shows are expanding their subscriber base and which are simply entertaining existing customers. Shows that attract existing subscribers but don't grow the subscriber base are more vulnerable.

This metric explains why some international shows are renewed despite modest domestic viewership. Money Heist and Squid Game were massive subscriber acquisition drivers in markets Netflix was trying to penetrate. Domestic shows that don't similarly expand the subscriber base face a higher bar for renewal.

4. Churn Prevention

Do viewers cancel their subscriptions after finishing a show? For binge-released series, this is a real concern. A show that keeps subscribers entertained for days or weeks is valuable even if its completion rate is moderate. A show that is watched quickly and forgotten is less valuable. According to a 2022 Wall Street Journal report, Netflix tracks a metric called "churn delta" — the difference in cancellation rates between subscribers who watch a specific show and those who don't. A show that keeps subscribers on the platform for an additional billing cycle is worth significantly more than its raw viewing numbers suggest.

This is where the difference between network and streaming becomes stark. A network show needed to keep you watching for 30 minutes at a time. A streaming show needs to keep you subscribed for another month. That shift in incentive has quietly reshaped what kinds of shows get made and which get cancelled.

5. Cultural Impact (The Intangible)

Some shows earn renewal on cultural impact alone — Squid Game, Stranger Things, Wednesday. These shows generate conversation, press coverage, and industry buzz. They make the platform feel culturally relevant. But cultural impact is hard to measure, and it only saves a handful of shows each year. According to Variety, even shows that top Nielsen streaming charts can be cancelled if their other metrics don't align. 1899 and The Midnight Club both debuted in Nielsen's top 10 but were cancelled within months. Cultural impact, it turns out, is a poor substitute for efficiency ratio when the spreadsheets are being updated.

The Human Cost

Behind every cancelled show are hundreds of people — cast, crew, writers — who built something they believed in. Data can't measure the passion that goes into making television. But data is what decides whether that passion gets to continue. As one former Netflix executive told The Hollywood Reporter: "We're not trying to decide what's good. We're trying to decide what's efficient. And those two things are very different questions."

Every cancelled show deserves a proper ending — because there's no metric that measures what a story means to the people who love it. To understand who actually makes the call, read who decides whether a show lives or dies. Explore our library of fan-written conclusions for the shows the data forgot.