THE COVEN GIRL GANG: CUSTOMER DISCOVERY SCRIPT AND EVIDENCE LOG
Author: Stuart Kerr Spindlow
Template version: 28 September 2026
Guide: https://thecovengirlgang.com/start/test-your-business-idea#discovery-log

Use a plain text editor. Duplicate the blank entry for each conversation.
This is an original suggested structure, not a validated survey. Interview
counts do not establish market size or predictable sales. No paid tool needed.

BEFORE YOU START
Audience and problem:
Riskiest assumption:
How participants will be selected (include alternatives and contrary views):
What evidence would justify continuing, changing or stopping:
Decision/review date:
Who can access notes, why they are needed, and deletion/review date:

OPENING (adapt accurately, do not promise what you cannot do)
I am researching [task/problem]. This will take about [realistic duration].
I will use the notes for [purpose], share them with [people, if any], and
keep identifiable notes until [date/review rule]. Taking part is optional;
you can skip questions or stop. Are you happy to continue?
Ask separately before recording. Notes may be enough. Do not request account
details, confidential client work or unnecessary sensitive information.

QUESTIONS
1. Tell me about the last time you needed to [complete the task].
2. What happened first, and what did you do next?
3. Which part, if any, was difficult?
4. What did you use or who did you ask for help?
5. What time or money did that take? Label estimates as estimates.
6. What happened when it was delayed or left undone?
7. How did you decide whether to pay for help, and who was involved?
8. What have I missed about how you handle this?
Follow-up: Can you give me an example?
Avoid suggesting the answer. Record reactions to a later pitch separately.

BLANK ENTRY
Participant code (keep contact details separately):
Date and reason this person fits the audience:
Participation agreement and separate recording decision, if relevant:
Most recent example and sequence of events:
Current alternative:
Exact quote OR clearly labelled paraphrase:
Reported time/money cost (actual or estimated; unknown is acceptable):
Consequence/priority:
Evidence against our assumption:
My interpretation (separate from what the participant said):
Unanswered question:
Offer shown afterwards, if any:
Next step offered / agreed / actually completed (distinguish these):
Identifiable notes deletion/review date:

ROUND REVIEW
Pattern supported by concrete examples:
Contrary evidence and missing perspectives:
What remains unknown:
Continue / change / stop, and reason:
Smallest next honest test:
Owner and review date:

FICTIONAL EXAMPLE, NOT OBSERVED RESEARCH
Five freelance designers discuss their most recent invoice. Three describe
missing purchase-order details; one reports no difficulty; one cannot recall
a recent example. Two of the three already use invoicing software.
Interpretation: try a client-information checklist before another invoicing
app. Next test: invite suitable people to use a checklist on a real invoice
and report what still goes wrong. These invented counts are not demand,
market share or a conversion forecast.

PRIVACY
A code does not make notes anonymous if you can reconnect them to a person.
Collect only what the decision needs, restrict access, explain retention and
delete identifiable notes when no longer needed. Research participation does
not automatically permit marketing emails. Check the rules for your activity.

PRIMARY GUIDANCE CHECKED 28 SEPTEMBER 2026
https://www.gov.uk/service-manual/user-research/using-in-depth-interviews
https://www.gov.uk/service-manual/user-research/getting-users-consent-for-research
https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/data-protection-principles/a-guide-to-the-data-protection-principles/data-minimisation/