Calibrating a Win-Loss Interview Script for Leading Questions

Marketing Academy · Field Work●Product Marketing

MiniAudit· 25 minutes

Chewy

Objective: Given a draft 10-question win-loss interview script, identify every leading question and rewrite it as an open question that won't just confirm existing assumptions.

You're the product marketer on Chewy's Autoship retention team. A colleague on the retention squad drafted a churn-interview script and asked you to review it before the first calls go out.

Open questions surface the real answer; closed or leading questions just confirm what the interviewer already believes. Mark each draft question and rewrite the leading ones.

Before you start

What you'll need

Free path (everything below is enough to finish)

Google Docs

FreeMark up and rewrite the draft interview script

Comments make it easy to route the calibrated script back to the interviewer

The process

1 step

Step 01 of 01

Distinguishing open questions from leading questions in a research script

Do not ask leading questions like 'Did our pricing seem competitive?' Ask instead: 'How did you think about the pricing across the options you evaluated?' Open questions surface the real answer; closed questions confirm assumptions.

The draft script asks: 'Was the Autoship discount not generous enough for you?' Is this open or leading, and what would you ask instead?

Google Docs— Open the draft script in Google Docs and mark each question inline.

Procedure

Read each of the 10 draft questions and label it OPEN or LEADING
For every LEADING question, note which assumption it bakes in
Rewrite each LEADING question as an open version that doesn't presuppose the answer

Sample output

Q3 (draft): 'Was the Autoship discount not generous enough for you?'
Label: LEADING (assumes discount size was the problem)
Rewrite: 'How did the Autoship discount compare to what you expected when you signed up?'

Healthy

A rewritten question that lets the customer name the actual reason, even if it has nothing to do with the discount at all.

Unhealthy

A question that already contains the answer the interviewer expects, so every response just confirms the existing theory about why customers churned.

What this means

If a question can be answered with a simple 'yes, that was it' without the customer adding new information, it's leading.

So what do I do about it?

Symptom	Action	Effort
Every churn interview comes back blaming the same single factor the team already suspected	Re-audit the script for leading questions before running the next batch of interviews	30 min

YouYou can do this yourself, no engineering access required.

Final deliverable

An annotated version of the 10-question script with every question labeled OPEN or LEADING, and a rewritten open version for each leading question.

See a reference example

Sample output

Mailchimp churn-interview script, calibration (excerpt)

Q2 (draft): 'Did the automation limits feel too restrictive?'
Label: LEADING
Rewrite: 'Walk me through a time the automation limits changed what you were able to do.'

Success criteria

You're done when you can:

Correctly labels at least 7 of 10 questions as open or leading
Every leading question gets a genuinely open rewrite, not just a softened version of the same assumption