CustDev (customer development) means talking to real or future customers so that, before you launch ads or build a product, you understand what problem a person is solving and what they are willing to pay for. Below: when interviews are needed, whom to invite, how to run the conversation and how to turn your notes into decisions.
KelyanMedia editorial team · Updated November 13, 2026

An in-depth interview lasts 30 to 60 minutes and is held one on one. You do not sell or show a presentation. Instead, you ask how the person has already tried to solve their problem: what they tried, how much they paid, what annoyed them.
The result of this work is not statistics but a list of tested hypotheses: which pains are real, what words the customer uses to describe them and where they look for a solution.
In short: what you need to know
- TargetTo check whether people actually have the problem you want to solve before you spend money on the product and ads.
- FormatA one-on-one conversation about the person’s past experience, not about whether they would buy your product.
- SummaryConfirmed and rejected hypotheses, customer quotes and decisions: what to change in the product, offer and ads.
How CustDev differs from a survey and a focus group
An interview answers the question “why”, a survey answers “how many”. It usually makes sense to start with interviews and then use a survey to test your findings on a larger sample.
| Method | What it gives | When to choose | Limitation |
|---|---|---|---|
| In-depth interview | Motives, context, the customer’s own words | A new product, and it is unclear why people do not buy | Few people, so results cannot be read as percentages |
| Questionnaire, survey | Shares and frequency of answers | You already have hypotheses and need to understand the scale | People answer only the options given |
| Focus group | The group’s reaction to an idea or a mockup | Compare names, packaging, ad ideas | Participants follow the loudest person |
| Website analytics and CRM | What people actually do | The product is already selling | Does not explain the reasons behind behavior |
← Swipe to scroll the table →
If the interviews are one part of a larger market study, see the article on how to conduct marketing research.
How to choose interviewees
Talk to people who have already had the problem and tried to solve it. Friends and colleagues will give you polite answers, not honest ones.
Define 2 or 3 criteria for the right person, for example “a coffee shop owner in Tashkent who changed their coffee bean supplier in the past year”. A criterion based on an action is more reliable than age or income.
Where to look: your current customer base in CRM, people who submitted a request but did not buy, niche Telegram chats, referrals from people you have already interviewed. Former customers are especially useful: they know exactly what was missing.
How many you need: run interviews in rounds of 5 people per segment. If a new round brings almost nothing new, the segment is covered. If answers contradict each other, you have most likely mixed two different segments.
A 40-minute interview script
Ask about past actions, not future intentions. “Would you buy it?” almost always gets a “yes” that means nothing.
| Block | Time | Sample question |
|---|---|---|
| Awareness | 5 minutes | Tell me what you do and what a typical working day looks like for you. |
| The last case | 10 minutes | When was the last time you faced this task? What happened? |
| Attempts to solve it | 10 minutes | What have you tried? How much did it cost in money and time? |
| Choice and doubts | 10 minutes | Why did you choose this particular option? What almost stopped you? |
| Wrap-up | 5 minutes | What didn’t I ask that I should have? Who else should I talk to? |
← Swipe to scroll the table →
- record the conversation with the person’s consent so you are not distracted by note-taking;
- talk less than the other person and do not fill the pauses;
- when you get a general answer, ask for an example: “how did it go the last time?”;
- do not talk about your product until the end of the conversation, or from then on you will only hear compliments.
How to turn notes into decisions
Review interviews on the same day, while you still remember the intonation. A week later, only your own guesses will remain.
- From each transcript, write down the facts: events, amounts, timelines, competitor names.
- Separately save verbatim phrases about the problem: they will become website headlines and ad copy.
- Put the facts into a table: “hypothesis / how many confirmed / how many rejected”.
- For each confirmed hypothesis, write down an action: what changes in the product, the offer, the promotion channel.
A hypothetical example: you thought customers choose by price, but in conversations they keep mentioning delivery time. So the delivery time, not the discount, goes on the website’s first screen and into the ads.
Learn about online marketing for your audience →Important
Interviews reveal motives, but not market size. Check conclusions from a dozen conversations with sales, test ads or a survey on a larger sample before investing a large budget.
FAQ
Can CustDev be done online?
Should you pay respondents?
Who should run the interviews?
Next steps
You need to turn interview insights into a promotion strategy: see full-service digital marketing.
You need to test hypotheses with quick ads: learn about search ads.
You need to rewrite your website in the customer’s words: see website development.
Still have questions: call +998 33 339 33 34.
Discuss promotion for your audience →
