Netflix Overhauls Platform and Metrics

The new viewership metric, effective since June 2023, is "Views," calculated by dividing the total hours watched by a title's runtime. This replaces the previous primary metric of "hours viewed," which inherently favored longer series and films. For instance, under the new system, the shorter series *Wednesday* surpassed the longer *Stranger Things* Season 4 as the most popular English-language TV title, despite the latter having more total hours viewed. This change in the primary success metric directly alters the objective function for Netflix's content recommendation and personalization algorithms. The system now optimizes for a more normalized measure of engagement completion rather than raw watch time. The qualifying period for a title to make the "Most Popular" lists has also been extended from 28 to 91 days, acknowledging that shows can gain traction over a longer period. The sunsetting of older devices targets hardware that cannot support newer security features, video compression formats, or the processing demands of the evolving user interface. This includes devices like the PlayStation 3 and smart TVs from over a decade ago, whose limited hardware and outdated operating systems create technical debt and prevent the deployment of new features. The mobile app redesign incorporates a vertical, swipeable short-form video feed, a feature tested since at least May 2025. This is not intended to make Netflix a user-generated content platform like TikTok, but to leverage a familiar UI for content discovery, using clips from its own library to drive viewers toward full series and films. This provides Netflix with a new stream of high-fidelity user interaction data—such as dwell time, skips, and shares—to further refine its recommendation models. All these significant product changes are validated through Netflix's extensive internal experimentation platform, which allows for rigorous A/B testing before a full rollout. Engineering and product teams create different "cells," or user groups, to test variations of features and algorithms. Key metrics like streaming hours and long-term user retention are tracked to determine which changes lead to a measurably better user experience and become the new default.