A customer conversation studio

Start with a decision a team needs to make

A small product team can have a precise question and very little time to investigate it. Perhaps trial users disappear after the first session, or customers keep asking for an integration the team has not planned. A customer conversation studio could help the team hear what actually happened before it commits to a solution. The first offer should be a bounded research package, with a decision at the centre and a deliverable the team can use.

For example, a founder could offer five facilitated interviews about the first week of using a product. The client would receive a recruitment screen, an agreed interview guide and a short synthesis meeting. Conversable.com could become the home for that offer. The concept is illustrative; no interview service is currently being sold through this website.

Define what the package can answer

Before recruiting anyone, ask the client to name the decision the interviews will inform. “Improve onboarding” is broad. “Understand what recent trial users expected to accomplish in their first session” gives the work a usable focus. Write the question in ordinary language, then list the types of evidence that would be useful. Recent examples of attempted tasks will usually be more helpful than opinions about an imagined feature.

Set limits with the client. Five interviews can reveal experiences worth investigating, but they do not describe every customer or establish how common a problem is. The deliverable should preserve those limits. A studio that clearly separates observation, interpretation and an open question gives the team a better chance of using a small study sensibly.

Recruit people with recent experience

The recruitment screen should select for the experience under examination. For a first-week study, ask when the person signed up, what they tried to do and how recently they used the product. Do not recruit only enthusiastic customers suggested by the sales team. If the question concerns people who stopped, the sample needs people who stopped, subject to the client's access and consent arrangements.

Agree who will contact participants, who can see their details and whether an incentive is appropriate. Explain the session length, the topic and any recording request before the person agrees. Keep scheduling simple. A short invitation with one booking link is easier to manage than a chain of messages where the participant has to guess what they are joining.

Make the session feel manageable

The interviewer should begin with a plain explanation of the purpose. State that the participant is being asked about their experience and that they can skip a question. Confirm recording permission separately. If a client observer attends, introduce that person and explain their role. A participant should never discover an unseen audience halfway through describing a difficult experience.

Build the guide around a recent episode. Ask what prompted the person to try the product, what they expected, what they did first and what happened next. Follow the sequence before asking them to summarise it. “Can you walk me through that?” is often a better next question than a suggestion about why they became confused. Leave space for unexpected detail.

Deliver evidence in a useful form

A compact synthesis can include the question studied, the participant criteria, recurring observations, contrasting experiences and unresolved questions. Attach examples to each finding so the team can inspect the reasoning. A quotation can illustrate a point, but it should not stand in for the whole account. Explain what the participant was doing when the quoted statement arose.

Then connect the findings to the client's decision. If three participants misunderstood the same label, the team might test clearer wording. If the interviews reveal two different jobs, the team may need another study before changing the product. The studio should offer a reasoned next step, with the level of confidence made explicit, rather than presenting every observation as a design instruction.

Find the first clients through a narrow channel

Early distribution could focus on small software teams in a founder community or an accelerator. A clear sample deliverable can show what the client would receive. It should use an invented demonstration case or material shared with permission, not somebody else's private interview. Explain the scope, responsibilities and exclusions beside the example.

A referral from a product adviser may also be useful because the adviser can recognise when a team has an unanswered customer question. The studio still needs to qualify the work. If the client wants a survey, usability test or market estimate, explain that those require a different method. A focused initial service is easier to deliver consistently than a promise to handle every kind of research.

Run one five-interview pilot

The first pilot should test the operating process as well as the interview guide. Record how long recruitment takes, where consent or scheduling becomes awkward and whether the client can use the synthesis. Ask the client which decision became clearer and which questions remain. Ask participants whether the session description matched the experience. These observations can guide changes to the package.

Before offering another round, review the actual work involved. Recruitment, preparation, interviewing and synthesis all need time. A viable package must account for each step, including failed scheduling attempts. The aim of the pilot is to discover whether the service can produce a useful, responsibly scoped result for a particular client. A recognisable domain can support the introduction; the quality of the research supplies the reason to return.