Many executives use Netflix Inc. (NASDAQ: NFLX) as an example of how artificial intelligence can change the bottom line. The company has long said that its recommender system has a measurable effect on business.
Two high-ranking Netflix executives wrote in a peer-reviewed article that personalization and recommendations “save us more than $1B per year,” mostly by keeping people interested and lowering cancellations.
The same paper says that recommendations affect choice for about 80% of the hours streamed on the service. This shows that discovery is not just a side feature at Netflix.
That framing is important because investors often shorten the story to “AI dollars.” Netflix’s own description is more accurate.
The value is clear because there is less churn, a higher lifetime value, and fewer replacement purchases, all of which help subscription revenue.
Better matches keep people watching, and happy subscribers are less likely to leave, which means more steady cash flows. Retention is the key for a subscription platform that can raise prices or add premium features over time.
The monetization flywheel is now bigger than it was when Netflix put out that academic work. Advertising is another way for the company to turn engagement into money.
Netflix’s Upfront presentation said that as of May 2025, the ad-supported tier had 94 million monthly active users around the world.
That scale turns viewing time into impressions that can be sold, which in theory increases the effective yield on engagement created by the recommender system.
It’s also a design statement that about four out of five hours of viewing are guided by suggestions made by machines.
Netflix tunes not only rows and rankings, but also artwork, evidence panels, and search.
The goal is to make it easier to get to the play button after opening the app. This increases the number of hours watched and makes it less tempting to leave during slow content cycles.
These mechanics are now standard in both streaming and commerce. Personalization is changing more and more what people see on their phones and TVs, from catalogue orders to thumbnail art.
As platforms improve their discovery tools, like Apple launching an AI search engine for Siri and Safari, the same competitive race happens in other places.
More Read
The $1 billion number is an estimate from Netflix’s top executives that was published in an academic journal. It has been around for a few years and is not listed as a line item in financial statements.
As Netflix adds live events and builds its own ad-tech stack, personalized ranking decides not only what plays next, but also which ads are seen and how often.
That means that the recommendation engine is not just a way to keep customers, but also a part of the revenue architecture.
There will still be a lot of talk about the bigger effects of AI on jobs and productivity, such as whether to be cautious or hopeful about it taking over jobs and the economy.
