Metrics question

How would you evaluate a dataset we're going to purchase that has competitor data on rides (in the U.S)?

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 rigorous framework for evaluating a data asset's quality, coverage, and business value before a purchase decision.

Clarify the use case first: what decision this competitor ride data would actually inform, like pricing or market share estimation.
Assess data quality: sample size, geographic and time coverage, and whether the methodology is disclosed and verifiable.
Check for representativeness: whether the data covers all competitor markets and ride types Uber cares about, or just a subset.
Estimate the business value: how much better a decision, like a pricing change, would be with this data versus without it.
Weigh cost against that value, including legal and compliance review, since competitor data can carry contractual or legal risk.
Define an acceptance criterion, such as data must cover a stated percent of relevant markets at an acceptable freshness, before recommending purchase.

Ties the evaluation to a specific downstream decision instead of assessing the dataset for its own sake.
Explicitly checks methodology and sample representativeness rather than trusting the vendor's summary claims.
Weighs cost against a concrete estimate of decision value, not just data quality in isolation.
Flags legal and compliance risk around acquiring competitor data as a real part of the evaluation, not an afterthought.

Evaluating data quality in the abstract without tying it to a specific decision it would inform.
Ignoring the legal and compliance risk of purchasing a dataset containing competitor information.

What would make you walk away from this dataset even if the quality looked good?
How would you validate the vendor's data against Uber's own internal numbers?