I don't understand why the customer isn't buying.

What to do when you don't understand why the customer isn't buying

Austin, two thousand and thirteen. Inside Outbox, the engineering team has built something genuinely clever. An elegant operational system for intercepting physical postal mail before it arrives at the customer's home, scanning it, and delivering a digital version to their phone.

The technology is impressive. They have solved hard logistics problems around intercepting mail at scale. The legal and privacy infrastructure works. The scanning quality is good. They have won design awards. The press has been warm.

What they have not won is paying customers.

Not for lack of trying. The product launches. Customers sign up. They use the service for a while. They tell friends about it. And then, when the bill arrives, they cancel.

Through two thousand and thirteen, the founders work through what should be the standard playbook. They tune pricing. They run promotions. They iterate on the onboarding. None of it moves the conversion. The retention numbers stay where they are: customers like the idea of the product, customers don't pay enough for it to cover the operational cost of running it.

What they discover, late, is something the conversion data had been telling them for months without their noticing. People who still receive physical mail have a relationship with it that is closer to tolerance than to frustration. They aren't suffering from mail. They are sort of living with it. The amount they are willing to pay to stop living with it is well below what the operation costs to run.

In their post-mortem, the founders are unusually honest. Every one of their customer conversations had been with people who liked the idea of the product. None had been with people who desperately needed it.

When the customer isn't buying, the question every team eventually has to face honestly is whether liking and needing have been treated as the same signal — or whether the team has been listening to the friendlier version of the conversation, and the quieter version is the one that pays.

So let's go to the office and work through it.

First decide whether the value is there at all

"I don't understand why the customer isn't buying."

The feeling is confused.

The product works. The pitch makes sense. The conversations have been polite. And nothing converts. The numbers say what they say, and the team's mental model of the customer doesn't match what the numbers describe.

Two choices. They look like the same conversion problem. The first move is different.

When you may be selling a problem nobody has

Choice one: the value isn't there. You've been selling something the customer doesn't actually want. The sales conversations have been polite but fundamentally disconnected from anything real in the customer's day. The pleasant-tone of the meetings has been masking the absence of a problem the customer is paying to solve.

If that's the read, stop selling for three days. Go and sit in the customer's working environment without pitching anything.

Steve Blank, formalising what he calls customer development, has written that the foundational mistake teams make is conducting customer research as a sales activity. The question you need to answer when sitting in a customer's environment isn't what do you think of our product. It's what are you actually doing all day, and where is the pain that would make you change anything. If you can't find a real pain after three days of watching, the product isn't the problem. The proposition is. And no amount of sales effort will fix it.

When the value is real but described badly

Choice two: the value is there and you're describing it badly. The product genuinely does something useful. The customer would want it if they understood it. But the pitch is about features when it should be about their problem — and the customer hears the pitch as a list of things you've built rather than as a way of doing what they were already trying to do.

If that's the read, rewrite every customer-facing piece of copy to talk exclusively about their workflow. Remove every mention of your features, your technology, your clever architecture.

April Dunford, working on positioning, has been clear that the test is whether a customer reading your rewritten page would recognise themselves in it before they recognise you. If they recognise you first, you've written a brochure. If they recognise themselves first, you've written positioning. Most teams have written brochures, and the brochure is the thing the customer is bouncing off.

How to diagnose the gap between selling and buying

Value-isn't-there, or describing-badly. Same conversion problem. Two different first moves.

Three tools. The discipline is to diagnose the gap between what you think you're selling and what the customer thinks they're buying.

Hear what the customer would say if you weren't there

The first is

Problem Interviews.

Problem Interviews come out of the customer-development tradition formalised by Steve Blank and Eric Ries through the late two-thousands and early twenty-tens — The Four Steps to the Epiphany (Blank, two thousand and five) and The Lean Startup (Ries, two thousand and eleven) are the canonical references. The deeper discipline traces through ethnographic interviewing in social science research and the contextual inquiry tradition that emerged from human-computer interaction work in the nineteen-eighties and nineteen-nineties.

The reason Problem Interviews matter when the customer isn't buying is that the team has been talking to customers and not listening to the quiet half of what they have been saying. The polite half — the half about the product — is loud. The diagnostic half — about the customer's actual workflow — is quiet, and only surfaces when the conversation has stopped being about the product.

The unique insight is that the point of the conversation is to hear what the customer would say if you weren't in the room. That sentence has to be carried as the discipline of the interview. The customer is being polite to you because you are there. The version of the truth you need is the version they would tell another customer, in a hallway, when you were not.

What you get when the interview is done well is a different sentence to act on. Not the product is fine, the customer just doesn't get it, which is consoling and false. The customer's actual problem is X, and our product addresses Y, and Y is not X, which is structural and acts as the basis for the next decision.

So. How to run it.

Pick. Real ones. Ideally a mix — some who tried and didn't buy, some who tried and bought, some who never tried. Compensate them. Schedule them inside a week.

Withhold. The conversation is about their day, their work, their current frustrations. The product comes up only if the customer surfaces it.

Ask. What did you try before? What are you using instead? What is currently frustrating about it? The workaround inventory is the diagnostic — the gap between their workarounds and their stated need is where the value is, or isn't.

Listen. Real pain has a specific texture. I lose two hours a week to this. I have to do this on Sunday evenings. We had to hire someone just to handle it. Generic yeah, that's annoying is not pain. Specific cost or specific workaround is.

Aggregate. After five interviews, the same problems will keep surfacing. Or they won't. Both are information.

Test what you built against the pains they actually have

The second is

the Value Proposition Canvas.

The Value Proposition Canvas was developed by Alex Osterwalder, Yves Pigneur, and Greg Bernarda in two thousand and fourteen as a companion to the Business Model Canvas, formalised in their book Value Proposition Design. The discipline traces through the broader Strategyzer toolkit and the jobs-to-be-done tradition we covered in scenario twenty-one. The Canvas itself is a two-circle diagram: the customer-profile circle on one side, the value-map circle on the other.

The reason the Canvas matters when the customer isn't buying is that the gap between the two circles is usually larger than the team thinks — and the gap is shaped like an assumption someone made in a workshop two years ago that nobody has revisited.

The unique insight is the two-circle structure. Most teams have a value map (what we built) without an explicit customer profile (the pains and gains the customer actually has). When forced to write the customer profile on its own and then check it against real customer evidence, teams discover the customer profile they have been operating on is largely fabricated — a workshop artefact rather than a description of real customers.

What you get is a structural test for the value gap. Not I think we should add a feature, which is opinion. Our value map addresses pain X; our customer profile, written from real interviews, names pain Y; X is not Y, which is structural.

So. How to run it.

Profile. On the right circle: customer jobs (what they're trying to get done), pains (frustrations and obstacles), gains (positive outcomes they want). Each item from the Problem Interviews, not the team's imagination.

Map. On the left circle: products and services (what you offer), pain relievers (how your product reduces specific pains), gain creators (how your product produces specific gains).

Match. Draw lines between specific items in the value map and specific items in the customer profile. The unmatched items are where the work is.

Diagnose. Customer pains with no matching pain reliever are gaps in the value map. Pain relievers that match no customer pain are gaps in your understanding of the customer — features you built that solve problems your customers don't have.

The Canvas is doing the editorial work of splitting what we built from what they need, in a structure that makes the mismatch visible rather than hidden inside the team's collective optimism.

Check whether the conversion gap is really a model gap

The third is

the Business Model Canvas.

The Business Model Canvas was developed by Alex Osterwalder in his two thousand and four doctoral thesis and formalised with Yves Pigneur in the two thousand and ten book Business Model Generation. The discipline is structural — a single page divided into nine boxes that together describe how a business creates, delivers, and captures value. It is now the canonical strategy tool for early-stage companies and is widely used in corporate-innovation contexts.

The reason the Business Model Canvas matters when the customer isn't buying is that the conversion problem is sometimes a business-model problem in disguise. The product is fine; the customer is real; the gap is that closing the gap is not actually profitable for you, or that the customer segment with the pain is the wrong segment to be chasing at all.

The unique insight is the nine-box structural test. The Canvas forces explicit description of customer segments, channels, revenue streams, cost structure, and key resources. The mismatch — the place where the business model breaks — is usually visible in the diagram once you fill it in honestly. Teams that fail to convert are often using a customer segment that is real but cannot be reached profitably through any channel they can afford.

What you get when you fill it in is a structural diagnosis. Not the customer doesn't want it, which is the wrong frame. The customer wants it, and the cost of acquiring this customer through the channels available to us is higher than the lifetime value, which is structural.

So. How to use it as a conversion diagnostic.

Segment. From the Problem Interviews and the Value Proposition Canvas. Be specific. Not small businesses; small business owners with five-to-twenty employees who currently use a generic accounting tool and lose at least three hours a week to manual reconciliation.

Channels. How do you reach this customer? Be honest. Most teams discover their channels are aspirational rather than operational — they assume a channel will work without having tested it.

Money. What does this customer pay you? What does it cost you to acquire and serve them? The arithmetic either works or it doesn't.

Break. The cell where the model breaks is usually one of: customer segment too narrow to scale, channels too expensive for the lifetime value, cost structure incompatible with the price the segment will pay. The broken cell is the conversation.

The Canvas is what stops the team from spending another quarter optimising a part of the business model that isn't the part that's broken.

A precedent: deciding which customer you are actually selling to

That's the toolkit. One more story before we close.

The Outbox story we opened with showed a team that had treated liking and needing as the same signal — and discovered that they were not, expensively, late. The story we close with shows the inverse — a founder who noticed the conversion data was telling her something the team had been refusing to hear, and acted on it before the runway expired.

New York, two thousand and twenty. Leigh Sevin co-founded what is now called Endear in two thousand and eighteen, under its original name Arthur — a generic SaaS retention tool with horizontal ambitions. Two years in, the conversion data is telling the team something they have been reluctant to hear.

Customers who try the product in its generic form almost never convert to paying. The few who do are all from a single niche: retail clienteling. Sales associates managing personal relationships with high-value retail customers, sending follow-up messages, building shared customer notebooks across the team.

The temptation, as Sevin describes in a two thousand and twenty-four Temy Co interview, is to treat the niche as a useful marketing wedge while keeping the underlying product horizontal. Keep the general-purpose retention tool. Sell harder into retail. Use retail as the entry point and broaden later.

Sevin and her co-founder make a different call.

They collapse the product into the niche. They rebrand the firm from Arthur to Endear. They rewrite the product around clienteling-specific workflows — shared customer notebooks between sales associates, messaging integrations for in-store staff, post-sale follow-up templates built on retail cadence rather than SaaS convention. They stop building features from the assumptions the team had been carrying into meetings about what a retention product ought to do, and start building features from real user feedback. The horizontal product is decommissioned. Retention customers outside retail are let go.

When a customer isn't buying, the honest answer is usually not that they don't understand the product. It is that the product wasn't built for them. Sevin's move wasn't to add a retail module to the general product. It was to notice that retail clienteling wasn't a module — it was the product, and everything else was noise the team had been carrying because they had been reluctant to narrow.

The question why isn't the customer buying? almost always deserves the answer: because you haven't yet decided which customer you're actually selling to.

So. Your Next Move from this playbook.

Which customer pain on your value map could you actually quote a real customer describing in their own words — and which ones did you make up in a workshop?