Ten Customer Interviews Beat Any Brand Workshop
Positioning invented in a room full of your own colleagues is a guess with a nice deck attached. Ten conversations with real buyers produce the sentences you should have been using all along.
The workshop problem
A brand workshop gathers the people who already know the product, asks them what makes it special, and writes down what they say. The output is internally coherent and externally meaningless, because everyone in the room shares the assumptions the exercise was supposed to test.
You end up with "we are the trusted partner for growing businesses", which describes every firm in your category and no buyer's actual problem.
Ten conversations with people who paid you money fix this, and they cost nothing except the asking.
Why ten is enough
You are not measuring anything, so you do not need a representative sample. You are looking for repeated language, and repetition shows up fast.
Nielsen Norman Group's long-standing finding in usability research is that five users surface around 85% of the problems because the same issues recur almost immediately. Message research behaves the same way. By interview six you will hear the fourth person describe the problem with the same odd phrase, and that phrase is your headline.
Ten gives you margin, and a second segment if your buyers split into two types.
Who to call
Four recent customers. They still remember the decision. Memory decays fast and rationalises as it goes.
Three people who nearly bought and did not. The most valuable and least conducted interviews in business. They will tell you what the objection actually was, which is rarely what your team believes it was.
Two long-term customers. They can tell you why they stayed, which is different from why they arrived and is the material for retention work.
One customer who left. Uncomfortable, and worth more than the other nine combined if you can get it.
The questions
Twenty minutes, recorded with permission, no slides. Rob Fitzpatrick's The Mom Test is the short book on this, and its central rule is the one to hold onto: ask about their life and their past behaviour, never about your idea.
- Take me back to the week before you contacted us. What was happening? This is the category entry point in their own words, and the material behind mental availability.
- What had you already tried? Names your real competition, which is usually "doing nothing" or "the intern".
- How did you find us? Free-text, unprompted. Expect surprises and write them down verbatim.
- What almost stopped you? The objection your sales page should be answering in the first screen.
- What did you expect that did not happen? Product and delivery gold.
- How would you describe what we do to a colleague? The single most useful sentence in the entire interview, because it is copy you did not have to write.
- What would make you recommend us? Then stop talking and let the silence do its work.
The questions that poison the data
"Would you use a feature that…" invites politeness. "Do you like our new positioning?" invites agreement. "How important is quality to you?" has one socially acceptable answer.
Anything hypothetical produces the answer the person thinks you want. Anything about the past produces evidence. When someone starts speculating, steer back: "when did that last happen?"
Turning transcripts into copy
Do not summarise into themes and lose the words. The words are the product.
Build one document of verbatim quotes, grouped by the trigger, the objection, the alternative considered, and the outcome described in their language. Then read your home page next to it. Most companies discover their site is written in the vocabulary of the industry while their buyers use the vocabulary of the problem.
Rewrite the headline using their phrasing. Rewrite the first objection-handling section using their objection. Pull the strongest quotes into case studies that survive a sceptic. Use the "week before" answers as your publishing calendar, because those are the moments people go looking.
This is also the cheapest way to make a small site convert better when you do not have the traffic to A/B test anything.
Frequently asked
Will customers agree to this? Most do, if you ask for twenty minutes, say it is not a sales call, and mean it. Offer nothing in return; payment changes the answers.
What if we have no customers yet? Then interview ten people who have the problem. It is pre-fit research, and it is more useful before you have built the messaging than after.
Can Fixora run them for us? Yes, and buyers often speak more freely to a third party. Ten interviews, a verbatim quote bank, and a rewritten home page. Ask for it and get the scope within 48 hours.
Have a project that needs this?
Tell us what you are building. We reply within 24 hours.
The service behind this
AI Marketing Services
A full marketing function without the headcount: strategy, content, campaigns, and outbound, run as one system and reported on the numbers that decide whether it paid.
