The two fixes are almost opposites, and guessing wrong in one direction costs months. Four tests that settle it, usually in an afternoon.
People arrive, look, and leave. Or they sign up and never come back. Working out whether you have a positioning problem or a product problem is the fork everything else depends on, and you cannot tell from the outside whether you built the wrong thing or are describing the right thing badly — and the two fixes are almost opposites.
If even a handful of people genuinely rely on the product, it is almost never a product problem.
Founders reach for the product explanation because it is the one they can act on alone. Rewriting a landing page feels like admitting the hard part is selling.
A positioning problem means the right people are not recognising themselves in what you say. A product problem means the right people recognise themselves, try it, and it does not do the job.
The distinction is in what happens after someone engages, not before. Traffic and signups tell you almost nothing on their own — the diagnostic signal is in the behaviour of people who got far enough to judge.
Run these in order. The first two usually settle it.
Test 1 — Does anyone stay? If even three or four customers use the product regularly and would be annoyed if it disappeared, the product does its job for someone. That makes it a positioning problem: you have built something real for a group you are not reaching.
Test 2 — Do your happiest customers have something in common? Look at your retained users. If they share a trait — same role, same size, same trigger event — you have found a segment, and your messaging is aimed wider than your product's actual fit.
Test 3 — Where do people drop? Leaving before signup points at positioning. Signing up and never activating points at onboarding. Activating and then fading points at product. Those are three different pages of work — the onboarding framework covers the middle one.
Test 4 — What do people say when they leave? "I did not understand what it did" is positioning. "It did not do what I needed" is product. "I forgot about it" is onboarding wearing a product costume.
| Signal | Positioning | Product |
|---|---|---|
| Traffic converts to signup | Poorly | Normally |
| Signups reach first value | Normally | Poorly |
| Existing customers | Stay and are happy | Drift away within weeks |
| Common objection | "I'm not sure this is for me" | "It doesn't do X" |
| Best customers | Share a clear trait | Look random |
| Cost to fix | Days | Months |
The bottom row is why the default matters. Getting this diagnosis wrong in the direction of "product" costs months. Getting it wrong toward "positioning" costs a week and produces useful information either way.
Sometimes the answer is that you built a reasonable product for a segment that does not have the problem badly enough. That feels like both and is really a third thing: a market selection problem.
The tell is that people understand what it does, agree it is useful, and still do not pay. Nothing is broken and nobody cares enough — which is the product-market fit question rather than either of the two on this page.
Ask three people who did not convert to try it while you watch, without helping. Twenty minutes of this distinguishes positioning from product more reliably than a month of analytics, because you see exactly where their understanding breaks.
If it is positioning: name the alternative your customers came from, then write everything toward that. Not what your product is — what it replaces. The positioning framework works through the five components, and the first one is the alternative.
If it is product: resist adding features. Find the one moment where the product fails to deliver its core job and fix only that. A product problem is almost never a missing feature — it is usually that the main thing works badly under real conditions.
If it is market selection: narrow to the segment that showed the most urgency and re-test. That is cheaper than either alternative and often reveals the product was fine all along.
Three responses that feel productive and delay the diagnosis.
A/B testing headlines. Useful once you know your positioning is right and you are optimising expression. Useless while the underlying claim is wrong — you are testing variations of a message aimed at the wrong person.
Lowering the price. If people are not converting because they do not understand the value, a lower number does not make the value clearer.
Adding features. The most expensive way to avoid the question. If signups are not activating, more surface area makes activation harder, not easier — see why signups do not convert for the fuller diagnosis.
If people who try it stay but not enough people try it, that is positioning. If people try it and leave — including the ones who should love it — that is product. The distinction lives in what happens after someone engages, not in traffic or signup numbers.
Look at whether anyone stays. If even three or four customers use it regularly and would be annoyed if it disappeared, the product does its job for someone, which makes it a positioning problem. Then check whether those customers share a trait.
Positioning, by a wide margin. It is also far cheaper to fix — days rather than months. Getting the diagnosis wrong toward 'product' costs you months; getting it wrong toward 'positioning' costs a week and produces useful information either way.
That is usually neither — it is a market selection problem. Nothing is broken and nobody cares enough, which means you built a reasonable product for a segment that does not have the problem badly enough. Narrow to the segment that showed most urgency.
No. Headline testing is useful once your positioning is right and you are optimising expression. While the underlying claim is aimed at the wrong person, you are testing variations of a message that cannot work.
Bring your signup, activation and retention numbers. Marcus names it as positioning, product or segment — and gives you one action.
Try GhostCoach free →14-day free trial · cancel anytime · 30-day money-back on Lifetime