Classic estimation question and I still fumbled the setup.
Break the problem into a top-down estimation: start with Instagram's user base, estimate daily active users, then estimate average likes per active user per day by considering content consumption and engagement patterns. Use round numbers and state assumptions clearly, then sanity-check the final figure against known benchmarks.
Pro tip: Show your work and narrate your assumptions; interviewers care more about your structured thinking than the exact number. Also, mention that you'd validate the estimate with internal data (e.g., likes per DAU) if available.
Confirm whether the question refers to global Instagram likes per day, including all types of posts (photos, videos, reels, stories). Ask if you should include likes on comments or only on posts.
Start with world population (~7.8B), estimate internet penetration (~60%), and social media adoption. Assume Instagram has ~2B monthly active users (MAU) and ~1B daily active users (DAU).
Consider that not all DAU like content. Assume 50% of DAU like at least one post per day. For those who like, estimate an average of 5 likes per day (some like more, some less). This yields 0.5 * 5 = 2.5 likes per DAU per day.
Multiply DAU (1B) by average likes per DAU per day (2.5) to get 2.5B likes per day. Sanity-check: Instagram has ~2B MAU, so 2.5B likes/day implies ~1.25 likes per MAU per day, which seems plausible.
Compare with any known public data (e.g., Instagram reported 4.2B likes per day in 2019). Adjust assumptions if needed, and discuss potential sources of error (e.g., bots, power users).
AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.