Strategy question

Researchers, data ops, and external vendors each bring different pain points: researchers ask for custom workflows, vendors complain about confusing UI steps, and engineering warns against supporting too many variants. How would you turn this mix of one-off requests, qualitative feedback, and constraints into a clear 6-month roadmap for the platform?

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 synthesis of mixed qualitative signals from three different stakeholder groups into a coherent, prioritized platform roadmap.

Categorize each piece of feedback by root cause, for example researcher custom workflow requests likely reflect gaps in flexible task configuration, not one off feature gaps.
Weigh vendor UI complaints heavily, since confusing steps directly hurt throughput and data quality at scale across every task researchers run.
Test engineering's variant sprawl concern against real usage data, for example how many distinct task configurations exist today and how much they actually differ.
Propose a roadmap built around a smaller number of flexible, configurable primitives rather than either unlimited custom variants or one rigid workflow.
Sequence the six months with vendor UI simplification first, since it is lower risk and improves every existing workflow, then invest in configurable primitives for researcher requests.

Reframes scattered researcher requests as a shared need for configurability rather than treating each as an independent feature request.
Validates engineering's too many variants concern with actual data on configuration sprawl instead of dismissing or accepting the concern at face value.
Sequences vendor facing fixes early since they compound across every task, before investing in more flexible researcher facing tooling.
Proposes a small number of reusable primitives as the roadmap's backbone, balancing researcher flexibility against engineering's maintainability concern.

Treating each stakeholder's feedback as a separate backlog item instead of looking for the shared underlying need.
Ignoring engineering's variant sprawl warning and promising unlimited customization that becomes unmaintainable.

How would you decide which researcher requests are worth building configurable support for?
What would you cut from the six month roadmap if the team lost half its capacity?