# Customer discovery questions: what to ask before building

Customer discovery is more useful when you ask about what someone has actually done than whether they like your idea. Start with a recent instance of the problem, follow the current workaround, and ask what made the situation difficult. Keep enthusiasm separate from evidence of willingness to change.

## Ask about the last time
For a fictional scheduling product, ask “Tell me about the last time you arranged a meeting across timezones.” Follow with “What did you use?” and “Where did you check the timezone?” These questions invite a concrete story. “Would an amazing timezone tool help?” invites agreement with your pitch.

## Use a conversation sheet
| Question | What to listen for |
| --- | --- |
| When did this last happen? | A specific situation rather than a general opinion |
| Walk me through the steps | Actual workflow and tools |
| What went wrong or took effort? | Friction in the person's own words |
| What did you do next? | Existing workaround |
| Who chooses and pays for the tool? | Decision process |

Do not turn the conversation into an interrogation. Ask permission, let the person finish, and follow the interesting details. Keep notes factual and collect only the personal information you need.

## Use AI after the conversation
```text
Organize these redacted discovery notes.
Separate observed behavior, participant statements, my interpretation,
and unanswered questions.
Identify current workarounds and evidence of actual effort or spending.
Do not invent demand or treat polite interest as purchase intent.
Notes: [paste]
Idea being considered: [context]
```

## End with a real next step
If the person is interested, a follow-up task, prototype session, or introduction may provide more evidence than a hypothetical rating. State honestly what exists and what does not. Do not accept payment for a product you cannot deliver without making the terms clear.

Look across conversations for patterns, but keep the sample and recruitment limits visible. Five similar conversations can suggest a useful direction; they do not establish the size of a market. The result should help you choose a smaller, better-defined problem to investigate or solve.

## Follow a real example beyond the first answer
Here is a fictional exchange for someone considering a scheduling service:

Participant: “Scheduling across timezones is annoying.”

Interviewer: “Can you walk me through the last meeting you arranged?”

Participant: “I sent three times in a group message on Monday. One person thought I meant their local time, so we moved the meeting.”

Useful follow-ups are “What did the message say?”, “How did you notice the mismatch?”, and “What did you do for the next invitation?” You are trying to understand the sequence. Do not interrupt with a demonstration of your preferred solution before the person explains their workaround.

## Separate the notes from the business conclusion
| Note from the fictional interview | What it supports | What remains unknown |
| --- | --- | --- |
| A group message listed three times | One reported scheduling method | How often this method is used |
| One person read the timezone differently | One reported misunderstanding | Whether a timezone label would have prevented it |
| The meeting was moved | A reported consequence | Time lost, cost, and willingness to pay |

“Everyone needs timezone software” does not follow from this conversation. A more useful next step is to interview other people who arrange similar meetings, or—with permission—observe how this participant writes the next invitation.

## Recruit for the situation you are studying
Look for people who have done the task recently. Tell them the topic and expected length of the conversation. You can ask where they already discuss the work, but do not scrape private groups or imply that a research conversation is a paid job offer.

If you mostly interview friends who want to encourage you, record that limitation. If people already use a workaround successfully, learn why they keep it. A product must fit the circumstances in which someone would actually change their behavior, not just sound attractive in a pitch.

---
SkillStall · 2026-10-03