Netflix browse homepage: choosing its goal, north-star metric and guardrails

You are the product manager for the Netflix homepage: the screen a member sees after picking a profile, with rows of titles to browse. How would you set goals for this page, and which metrics would tell you it is succeeding? ```hint Define a good visit Members do not open the homepage for its own sake. Decide what a successful visit looks like from the member's side, and which event in the product marks it, before you list anything to count. ``` ### Clarifying Questions - Does the homepage include the hero preview at the top, the Continue Watching row and the entry points to search and My List, or only the rows of recommendations? - Which devices are in scope? TV apps, phones and the web differ in session length and in autoplay previews. - Is this an existing page with baselines to improve on, or are we setting goals for the first time? - Beyond helping members, should the page serve business goals such as promoting new originals, and how should those rank against member satisfaction? ### What a Strong Answer Covers - A goal for the page tied to Netflix's subscription model and to the member's job of finding something worth watching - A north-star metric that captures a successful visit rather than clicks or time spent on the page, with its definition made precise - Funnel metrics from landing to play to sustained viewing, and guardrails against misleading artwork, endless browsing and narrow recommendations - How ambiguous signals such as browsing time are interpreted, and how targets are set from baselines, segments and seasonality - How the metrics connect to retention and are used in experiments ### Follow-up Questions - After a redesign, plays started from the homepage went up, but viewing hours per member stayed flat. What might be happening, and what would you check? - A member leaves the homepage without playing anything but adds two titles to My List. Should that visit count as a success or a failure? - Retention is the outcome Netflix cares about, but it moves slowly. How would you show that your homepage metric predicts it, so experiments can be judged within weeks?