Example questions
The companies, situations and case numbers are fictional. These are our own practice questions, with no claim that an employer uses them.
Product sense for a booking problem
Morrow Workshops lets community groups book meeting rooms. Organizers say it takes too long to find a suitable room, but no usage figures or confirmed cause are available. Decide which user problem to investigate and the first test to run.
How would you clarify the problem before choosing a feature?
An example approach
I identify organizer segments and observe their booking attempts because a search complaint could reflect missing availability, unclear capacity or approval delays. I define the task and outcome rather than assuming a new filter solves it.
How might the problem differ for a first-time organizer and a frequent one?
A search metric that can mislead
Harbor Catalog reports more searches per user after a redesign, and the team calls that increased engagement. The product helps users find an item and submit an order. Decide how to test whether the redesign helps that task or adds friction.
What could an increase in searches per user mean?
An example approach
It could reflect useful exploration or repeated failed searches, so I define the user task and distinguish sessions, users and search attempts. I check instrumentation and traffic mix because a metric increase has no fixed direction of value.
How could changing the search-trigger event inflate the metric without a behavior change?
Prioritization with uncertain estimates
Vale Products can fund one change: a reliability repair, simpler onboarding or an export requested by a customer. Impact and effort estimates are uncertain, and the goal has not yet been agreed. Explain how to reach a decision the team can defend.
What decision criteria would you establish first?
An example approach
I clarify the product goal, commitments and material risks because impact means different things for reliability and acquisition. I identify any safety or contractual requirement before comparing discretionary work.
Can a mandatory reliability fix be treated as just another growth-score item?
Estimate a service’s capacity need
Cedar Community has 200 organizers in this planning case. Assume each hosts two events a month, 30 people attend each event and one quarter of attendance instances generate a reminder request. Estimate monthly requests and decide which assumptions need checking.
Build a driver tree and state the unit of the answer.
An example approach
I multiply organizers by events per organizer, attendees per event and reminder-request share because each factor converts the previous unit. The result is reminder requests per month, not distinct people, since one person may attend several events.
Which factor would you need to estimate distinct monthly people?
A release with a broken core journey
Juniper Booking is releasing a redesigned checkout to ten percent of eligible users. In this case, completion falls from 80 of 100 attempts in control to 55 of 100 in the release group, and payment-support tickets rise. The launch owner wants full rollout today; decide how to lead execution.
Which release decision needs attention first?
An example approach
I pause expansion and assess the core checkout failure because completion falls from 80 percent to 55 percent and support harm rises. I check that attempts and exposure are defined consistently before calling the redesign the cause, while retaining a safe containment path.
What would you do if revenue rises because the release group happens to contain larger orders?