Netflix Product Improvement: Reduce the Friction of Rating Titles

Netflix's rating page is annoying its users. As a product manager at Netflix, how would you fix it? Explain how you would find out what makes it annoying, which users you would focus on, what you would change, and how you would know the change worked. ```hint Two goals share the screen The viewer wants to keep watching, while Netflix uses ratings to improve recommendations. Look for a fix that removes the annoyance without quietly weakening that signal. ``` ### Clarifying Questions - Which rating surface is meant: the prompt shown when a movie or season ends, the rating controls on a title's page, or the taste picker shown to new members? - How do we know it is annoying (complaints, surveys, low response rates, sessions ending at the prompt), and has it always been this way or did it start recently? - Is the problem worse on some devices, such as TVs controlled with a remote, than on phones and the web? - Does Netflix need to keep collecting the same volume of explicit ratings, or can it rely more on viewing behavior? ### What a Strong Answer Covers - Why Netflix asks for ratings and what the product would lose without them, leading to a goal that balances viewer experience and signal quality. - Segments by viewing behavior and device, with a reasoned choice of whom to fix it for first. - Specific causes of the annoyance (timing, frequency, input, perceived value), separated from one another. - Prioritized solutions, each with its effect on rating volume and rating bias. - An experiment with a primary metric and guardrails on recommendation quality and retention. ### Follow-up Questions - If you pre-filled the rating based on viewing behavior, how would that affect the data the recommendation system learns from? - After your change, ratings collected per member drop sharply but satisfaction with the experience rises. Do you ship it? - How would the design differ for profiles shared by several people, or for kids' profiles?