Estimation question

Estimate the amount of storage needed to store all the photos and videos taken on Pixel phones.

Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.

What this question tests

Whether you can build a structured bottom-up estimate combining an active user base, capture rate, and file size assumptions.

Clarify scope: total new photo and video storage generated by active Pixel users in a year, not historical cumulative storage.
Estimate the active Pixel user base, a modest slice of Android's roughly 3 billion devices, say tens of millions of active Pixel users.
Estimate capture behavior: average photos per user per day and the smaller share of time spent shooting video versus photos.
State size assumptions: roughly 3 to 5 MB per photo, and a few hundred MB per minute of video at typical phone resolution.
Multiply users by daily captures by average file size, then scale to a year, keeping photo and video storage as separate line items.
Sum the two totals into an annual storage estimate, stated as a range, and flag which assumption drives the number most.

Breaks the estimate into photo and video separately, since their file sizes differ by orders of magnitude.
States a specific active Pixel user base assumption instead of using Android's total install base as a proxy.
Uses labeled per-item size assumptions, like 4 MB per photo, rather than guessing an aggregate number directly.
Flags video as the dominant driver of total storage despite being a smaller share of captures, and explains why.

Treating photos and videos as the same size, which badly understates total storage.
Using Android's entire user base instead of scoping to actual Pixel-specific users.

How would cloud backup versus on-device-only storage change this estimate?
Which single assumption would you want real data on first to sharpen this number?