How do I find out what customers want before I build?
Updated
Start with the customer's workflow, not your feature list. Run 20–30 problem interviews, map what buyers are trying to get done, and rank their pains by how often they happen and what they cost. Test a concept with the segment that hurts most. CorZen turns that into weekly tasks with an owner on each, on the timeline you choose.
- Two or three candidate segments, no pitch
- Sorted by how often and how costly
- The problem you build for
What does a customer discovery plan look like?
This is a sample 90-day roadmap for the Assess Product-Market Fit category with the Discovery Engine approach. CorZen also builds 30-day, 60-day, custom, and ASAP timelines. Your own plan changes with your market, your tools, and how many hours your team has each week.
Listen
Weeks 1–4
- Week 1Define 2–3 candidate segments and an interview guide with no product pitch.
- Weeks 2–4Run 25 interviews. Record the exact phrases buyers use.
Make sense of it
Weeks 5–8
- Week 5Cluster pains by how often they happen and what they cost.
- Week 6Pick one segment and one job to own.
- Weeks 7–8Draft two solution concepts. Review them with 8 interviewees.
Test the concept
Weeks 9–13
- Weeks 9–10Build a clickable prototype or a manual (concierge) version.
- Weeks 11–12Put it in front of 5–10 buyers. Measure completion and repeat use.
- Week 13Write the customer profile and positioning statement for the go-to-market plan.
How do you run a problem interview without pitching?
Ask about what buyers already did, not what they might do. People are poor at predicting what they would buy and good at describing what happened last week.
Ask about
Walk me through the last time this happened.
What did you do about it?
What did that cost you in time or money?
What have you already tried, and why did you stop?
Who else was involved in the decision?
Avoid
Would you use a product that…?
How much would you pay for…?
- Describing your idea before the last few minutes
- Interviewing friends or existing fans
- Counting compliments as demand
Save your idea for the end, if you mention it at all. Once you pitch, the interview turns into a sales call.
How do you decide which customer problem to solve?
Rate every pain on two things: how often it happens and what it costs when it does. The problem worth building for sits in the top right.
Niche
Rare but costly. Can work if the buyers are easy to reach.
Build here
Frequent and costly. Buyers already spend time or money on workarounds.
Ignore
Rare and cheap. Nobody will pay to fix it.
Annoyance
Frequent but cheap. People complain but rarely pay.
Write each pain down in the buyer’s own words. Those phrases become your positioning and landing page copy later.
How does the plan change on a 30- or 60-day timeline?
The order stays the same on every timeline: listen, make sense of it, test a concept. Shorter timelines mean fewer interviews, and the 30-day plan replaces the working prototype with a concept review. Custom and ASAP timelines keep the same order.
| Timeline | Interviews | Making sense of it | Concept test |
|---|---|---|---|
| 90 days | 25, weeks 1–4 | Weeks 5–8 | Prototype or concierge version with 5–10 buyers, weeks 9–13 |
| 60 days | 15–20, weeks 1–3 | Weeks 4–5 | Prototype or concierge version with 3–5 buyers, weeks 6–9 |
| 30 days | 10–12, weeks 1–2 | Week 3 | Concept review with 5 interviewees, week 4 |
A 30-day plan tells you which problem to solve. It can’t yet tell you whether buyers will use your solution.
What’s the difference between customer discovery and validation?
Discovery comes first. It finds the problem worth solving. Validation comes next and tests whether buyers will pay for your answer to it.
Discovery
Finds the problem worth solving.
You leave with a customer profile, a ranked list of pains, and the words buyers use.
Validation
Tests whether buyers will pay for your answer.
You leave with pre-orders, deposits, or paid pilots. See the validation template
Run discovery when you know the market but not the problem. Run validation when you already know the problem and need proof people will pay.
How does CorZen build this plan for you?
Tell CorZen your goal. Pick Assess Product-Market Fit, then the Discovery Engine approach, then a timeline: 30, 60, or 90 days, a custom length, or ASAP. A short AI interview asks about your niche, what makes your idea different, the tools you already use, and how much time your team has each week.
CorZen then builds a roadmap for that timeline, broken into weekly tasks. Every task has an owner, and a task can have more than one. On Team and Business plans, AI agents can draft deliverables for individual tasks, such as a customer research brief.
Which template should you look at next?
Frequently asked questions
How many customer interviews do you need for discovery?
Plan for 20–30, spread across two or three candidate segments. Stop a segment early when the last five interviews repeat what you already heard.
Where do you find people to interview?
Start where they already talk about the problem: online communities, industry groups, and people who have posted about their workaround. Introductions from people you know also work. Avoid friends and existing fans, who will answer kindly instead of accurately.
What does "jobs to be done" mean?
It's the progress a buyer is trying to make, such as getting paid faster or onboarding a new hire, rather than the product they use to get there. Mapping jobs keeps you focused on the outcome buyers want instead of on features.
Can you do customer discovery before you have a product?
Yes. Discovery comes before the product. You need a guess at who the buyer is and what problem they have, nothing more.