Reduce the Wait Users Feel on YouTube: Speed Improvement Plan and Metrics

You are a product manager on YouTube, and you are asked how you would make the product faster. Decide what "faster" should mean for the people who use YouTube, identify where they lose time today, and propose the improvements you would prioritize, together with how you would measure whether they worked. ```hint Faster for whom, and at which step? Walk through the path from opening YouTube to watching the next video, and ask where the wait comes from the system being slow and where it comes from the user being left with nothing to look at. ``` ### Clarifying Questions - Does "faster" mean technical performance such as app launch, video start and buffering, how quickly people find something worth watching, or how quickly creators' uploads become available? - Which surfaces matter most: the mobile app, the website or the TV app? - Is a specific complaint or metric behind the request, such as slow starts in certain countries, on certain networks or on certain devices? - What, if anything, may we trade away for speed: viewers' data usage, video quality or delivery cost? ### What a Strong Answer Covers - An explicit definition of "faster" and a justified choice of which users and surfaces to focus on. - A step-by-step map of where time is lost in the YouTube experience, separating real delay from delay the user only perceives. - Prioritized improvements specific to YouTube, each with its cost and trade-offs. - Speed metrics that look beyond the average and link to engagement outcomes, with guardrails and a plan to test the changes. ### Follow-up Questions - Preloading the next likely video would cut start times sharply but increase delivery cost and viewers' data use. How would you decide whether, and for whom, to ship it? - Median time to first frame improved, yet watch time did not move. What could explain that, and what would you do next? - How would your plan differ for YouTube on smart TVs?