Product design question

How would you improve image search?

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

Product sense for a core search product feature, requiring a specific, technically grounded improvement.

Pick a specific user intent, like visual shopping searches, where current image search often returns visually similar but not purchasable results.
Identify the pain point: users can't easily filter image results by where to buy or how similar visually to their reference image.
Prioritize improving shoppable image search over general image search quality, since it's a clear, high-intent use case.
Propose a solution: visual similarity search combined with direct purchase links and price comparison for shoppable results.
Define success metrics: click-through rate to purchase and search refinement rate (fewer repeated searches for the same item).

Names a specific, high-value intent (shopping) rather than a vague 'better image search' answer.
Proposes a concrete mechanism, visual similarity matching plus purchase links, that's technically groundable.
Gives an illustrative metric, e.g. raising click-through to purchase by 25% for shoppable queries, marked as an assumption.
Considers monetization alignment, since shoppable search links naturally to Google's ad and shopping business model.

Staying too general with 'improve AI accuracy' instead of naming a specific use case and mechanism.
Ignoring how this improvement could align with Google's existing shopping ads business.

Chapters of the AI PM course that teach what this question tests.