Every founder knows that early customer conversations shape the story you'll tell. But most treat them as validation checkpoints — a box to tick — rather than as the raw material for positioning. The truth emerges quietly in those first twenty conversations: patterns in the language your customers use, the problems they describe (not the ones you assumed), and the gaps between what they say they want and what they actually do. If you listen for it, the positioning writes itself.
Why twenty is not arbitrary
The number twenty sticks because it is honest about how sense-making works. By roughly twelve conversations, you will have heard the same three to four problems described in similar language — your brain begins to recognise the shape of your market. By twenty, those patterns are stable enough to build on. Before twelve, you are still chasing novelty. After twenty, you are confirming what you already think you know.
This is not a rigorous statistical claim. It is an operating heuristic. Researchers call this "saturation" — the point at which new conversations no longer surface fundamentally new themes. You hit diminishing returns. The work shifts from listening to counting.
The conversation structure: problem-first, not solution-first
The conversations that matter are structured backwards from how founders naturally want to run them.
Founders want to start with context: "Here's our product idea." Customers want to know the problem you are solving before they commit cognitive load to understanding your solution. So the conversation begins with their problem — not yours.
The question that opens is brutally simple: "Tell me about the last time you struggled with [domain]." Not: "Have you ever considered automating X?" Not: "Would you use a tool that did Y?" The domain is as broad as it needs to be. For a logistics startup, it might be: "Tell me about the last time you struggled with getting deliveries to the right place on time." You stop talking. You listen.
What follows is a series of refinements that all point in the same direction: the real problem, not the polished version you drafted in your strategy document.
What to listen for: words, not paraphrases
The gap between "what customers say" and "what you think they said" is where most positioning fails. It happens in real time, in the conversation itself.
A customer tells you: "We lose half a day every time our inventory system goes down." Your mind hears: "Downtime is a critical problem." But the positioning insight is not in "downtime." It is in "lose half a day" — the quantified, lived experience of the impact.
Later, another customer says: "When our stock runs out without warning, we have to scramble." Same problem, different emotional signature. The first is about disruption to operational rhythm. The second is about being caught unprepared.
The work is to preserve their language, not to normalize it into your vocabulary. Tag the transcript for the verbs they used: "lose," "scramble," "scramble," "blindsided." Tag the nouns: "half a day," "warning," "inventory." Over twenty conversations, you will find that three to four metaphors or concrete images recur unprompted.
Counting signals: the pattern recognition phase
Once you have the transcripts, the work becomes forensic. You are looking for three types of recurrence:
- Linguistic recurrence — the same word or metaphor appears in at least three conversations, unprompted by you. If three different customers describe the problem as "flying blind," that phrase has gravity. It is a signal.
- Emotional signature recurrence — the same emotional tone or frustration appears even when the language differs. One customer is angry; another is resigned. If that same tone shows up four times, it is real.
- Objection or scepticism recurrence — customers bring the same concern or hesitation to the conversation. "We've tried tools like this before and they didn't stick." If you hear that four times, it tells you what your positioning needs to address head-on.
Write these down. Create a spreadsheet. Tally them. The positioning emerges from the tallies, not from your interpretation.
The "say-do gap": the insight most teams miss
Every customer will tell you what they want. Few will do it. The gap between the two is where real positioning lives.
A customer says: "We need a solution that integrates with our existing stack." In their current operation, they are using five different tools, none of which talk to each other, and they are not actively working to change that. The stated desire (integration) masks an operating reality (we have legacy systems and changing them is hard). The positioning insight is not about integration. It is about reducing friction in the current, messy system — not forcing a clean one.
Another customer says: "We want real-time visibility." In practice, they check their system once a day and have no process to act on what they find. The stated desire masks a different reality: they want to feel like they could know, not that they need to know constantly. The positioning is about control and peace of mind, not about refresh rates.
Listen for where words become disconnected from behaviour. Ask follow-up questions that bring the two into focus: "You mentioned wanting real-time visibility — how often do you actually check that today?" "You said integration is critical — tell me how you've handled this with your current tools." The answers will reposition your understanding.
Three patterns recur — and why they mislead
Most teams conducting early customer interviews fall into the same three traps, over and over.
Leading questions that confirm prior belief — You ask: "One challenge we've identified is that most teams lack visibility. Is that something you've experienced?" The customer answers yes, because you have just described a real problem and a real frustration. But you have not learned anything. You have validated yourself. The test is whether the customer names the problem first, in their own words, before you introduce it.
Sampling only the people who already say yes — You talk to power users or early believers. They are enthusiastic, articulate, and easy to recruit for interviews. But they are not representative of your market. The positioning that emerges from conversations with "people who already love this category" will not work for the sceptics you will actually need to convert. Seek at least three conversations with people who are ambivalent or actively avoiding the space.
Hearing solutions where customers offered problems — A customer describes a struggle. Your brain immediately translates it into a feature request. "We struggle with prioritising our backlog" becomes "they need backlog prioritisation software." But they may actually need a framework to think through what matters; the software is incidental. You have shifted from listening to assuming.
What This Looks Like in Practice
Antidote Africa is a strong example of positioning built from early customer research rather than assumed market theory. The founding team conducted extended conversations with health workers and community health volunteers in rural clinic settings. The early interviews revealed not that "there is a gap in diagnostics software" (the founder's prior) but that the real friction was trust: health workers doubted algorithmic recommendations and feared they would be blamed if the algorithm was wrong. The positioning emerged from that conversation — not as a diagnostic tool, but as a tool that amplifies human judgment and keeps the clinician's authority intact. That insight changed everything: the interface, the language used, the feature roadmap. It would not have emerged from a list of "problems to solve" drafted in an office.
Indicators that you have genuine positioning material
After twenty conversations, you should be able to point to clear signals that you have learned something true about your market. These are observable conditions that suggest you are ready to move from research to positioning work:
- Three or more customers have used the same metaphor, word, or phrase unprompted — and it is not a phrase you introduced.
- You can describe your customer's problem in their language, not your category language, and they recognise themselves in that description.
- You have discovered a gap between what customers say they want and what they actually do — and you understand why the gap exists.
- You can name the scepticism or objection you will face, and you have heard it from at least three customers in different contexts.
If three or more of these are true, you have genuine raw material. The positioning work can begin.
When the conversation reveals the wrong audience
Sometimes the conversations reveal a harder truth: you are solving a real problem, but for the wrong audience. A team founded to sell scheduling software to freelancers discovers, through customer conversations, that the real problem is not scheduling — freelancers already have workarounds — but payment timing and cashflow uncertainty. Their positioning was addressed to the wrong customer.
The instinct is to refine the messaging. The actual move is to repoint the company. If your first twenty conversations suggest that a different audience has a more acute version of the problem you are solving, it is worth exploring whether that audience is your true market. This is not failure. It is course correction based on evidence.
Closing
The positioning work that agencies charge thirty thousand pounds for often begins with this: twenty conversations, an honest transcript, and the discipline to listen to what customers actually say rather than to what you expected to hear. Founders who do this work themselves, before hiring expensive help, arrive at that first agency meeting with genuine market insight instead of guesses. The agency's job becomes about translating that insight into a coherent brand story — not about inventing positioning from scratch.
If you are facing the task of articulating your market position and want to ground it in customer signal rather than assumption, we are happy to talk through how to structure that research and what to do with what you learn.