Strategy question

Design a discovery process for Claude Science that would help you uncover a non-obvious, high-value use case with academic researchers and industry R&D teams. How would you separate real workflow pain from interesting-but-niche requests, and what criteria would you use to decide whether the opportunity merits engineering investment?

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

Tests design of a discovery process that separates genuine high value scientific workflow pain from interesting but low priority requests.

Start with structured interviews across both academic researchers and industry R&D teams, since their incentives and constraints differ enough to reveal different pain points.
Probe for frequency and cost of the pain, asking how often the workflow happens and how much time or budget it currently consumes, not just whether it sounds interesting.
Observe actual workflows where possible, such as shadowing a researcher through a real analysis task, since stated pain often differs from actual behavior.
Score candidate use cases on frequency, cost, and whether Anthropic's specific model strengths meaningfully improve the outcome versus existing tools.
Set investment criteria up front, such as requiring evidence from multiple independent researchers before committing engineering resources to a single suggested use case.

Combines interviews with direct observation, since researchers may not accurately self-report where their real time is lost.
Uses frequency and cost as the filter for real pain versus a one-off request, rather than trusting enthusiasm alone.
Requires the use case to specifically benefit from Anthropic's model strengths, ruling out cases better served by existing specialized tools.
Sets a pre-commitment threshold, such as convergent signal from multiple independent researchers, before greenlighting engineering investment.

Relying only on self-reported interview pain without observing actual workflows, which can overstate niche requests.
Greenlighting a use case based on one enthusiastic researcher instead of requiring convergent evidence.

How would you distinguish a workflow pain point specific to Anthropic's strengths from a generic tooling gap?
What would you do if academic and industry researchers pointed to different top use cases?