Free tool

Find out exactly why your PM interview answer isn't landing.

Paste a question and your answer. You get scores on the five dimensions real interviewers use — plus the one thing to fix that'll move your score the most.

How it works

1

Paste the question and your answer

Any PM question type: product sense, behavioral, metrics, prioritization, estimation. Written answers work; rough notes work.

2

Get scored on five dimensions

Structure, Specificity, Reasoning, Decision Quality, Delivery. Each scored 1–10 with a one-line note on why.

3

See your single most impactful fix

One concrete change that would move your score the most. Not a list of notes — the one thing.

What this tool doesn't do

This grades written answers once. The full product does the whole prep loop.

Real-time coaching as you speak

Inside the product, the AI listens to your spoken answer and gives live feedback mid-session — the same way a real interviewer would probe.

Company-specific playbooks

Meta asks different things than Google. Your prep is calibrated to the actual company, role level, and question themes in your target job.

Tracks your improvement over time

Every session is scored and saved. You can see exactly which dimensions are improving and which still need work across all your mock rounds.

See what's inside

Questions

Is this really free?

The first grade is free with no email. Two more are free after you drop your email. After that, live practice is inside the 7-day trial.

What question types does this work for?

Product sense, behavioral, metrics, prioritization, and estimation questions. The tool detects the type automatically and applies the right scoring rubric.

How is this different from asking ChatGPT to critique my answer?

ChatGPT will give you general feedback. This grades you on the five specific dimensions PM interviewers actually use — structure, specificity, reasoning, decision quality, and delivery — and tells you exactly which one to fix first.

Will my answer be used to train AI?

No. Your input is sent to our model to generate the grade, then dropped. We don't store the text of your answers.