Sizing daily YouTube video views across short clips and long-form

How many YouTube videos get watched in a single day? Build an estimate you can defend: decide what counts as a watched video, set up an equation, put round numbers on each input and check the result. ```hint Not all viewing is alike Viewers range from someone who opens the app for a few minutes to someone who keeps it playing for hours, and a short vertical clip lasts a fraction of a long upload. Think about which of these differences your model has to capture. ``` ### Clarifying Questions - Is the estimate worldwide or for one country? - Does it count plays (every time any video is watched) or distinct videos that get at least one view that day? - Do short vertical clips count, and does a play that is swiped away after a couple of seconds count as watched? - Which surfaces are in scope: the mobile app, the website, TV apps, and players embedded on other sites? ### What a Strong Answer Covers - A definition of the unit (plays versus distinct videos, and how short clips and very brief plays count) and of the scope (geography and devices). - An equation from viewers to daily watch time to videos, with round, labeled assumptions. - Segmentation by how much people watch and by format, since short clips and long videos differ in length by an order of magnitude or more. - Correct arithmetic, a final number with a range, and a sanity check from a second angle such as total daily watch time. - Which inputs drive the answer and how the estimate could be checked. ### Follow-up Questions - How would the answer change if a view only counted after 30 seconds of playback? - How would you estimate the number of distinct videos that get at least one view on a given day? - Viewing on TV screens is growing. How would that shift your count even if total watch time stayed flat?