People found you, understood enough to hand over an email, and then stopped. Four places that breaks, in the order worth checking them.
Lots of signups but no paying customers is a different problem from a quiet launch, and it needs a different diagnosis. People found you, understood enough to hand over an email, and then stopped. The gap between those two facts is where the answer is.
Check them in that order. Most founders assume the third and it is usually the first or second.
If almost nobody signed up in the first place, this is the wrong page — the flopped launch diagnosis covers that case, and the causes are completely different.
Launch traffic is the most misleading traffic you will ever get. A Product Hunt or Show HN audience contains a large share of people who sign up because the product is interesting, not because they have the problem.
The test is source-based. Split your signups by where they came from and compare conversion. If launch-day signups convert at 1% and organic search signups convert at 12%, the product is fine and your sample is contaminated.
This is the most common cause after a launch spike, and it is genuinely good news: nothing is broken. Your real conversion rate is the non-launch number, and you have far less data than you thought — see the conversion benchmarks by traffic and trial model for what to compare against.
Count how many signups reached your activation moment — the first point where the product visibly did its job once. Not logged in. Not completed onboarding. Did the thing.
| Activated | Reading | Fix |
|---|---|---|
| Under 20% | This is your problem. Nothing else matters yet. | Remove steps before first value |
| 20–50% | Meaningful leak, but not the only issue | Fix, then re-check conversion |
| Over 50% | Activation is fine | Move to break 3 |
If you cannot measure this, that is the finding — and instrumenting one event is a smaller job than most founders assume. The onboarding framework covers naming the activation moment and closing the gap to it.
If people activate and still do not pay, stop optimising and answer one question:
What does your product replace for the people who paid — and why did the others not need that replacement badly enough?
Everyone who pays for software is replacing something: a spreadsheet, a manual routine, a freelancer, an hour every Thursday. Your paying customers replaced something specific. Your non-converters either had nothing to replace, or what they had was good enough.
With four or five paying customers this is a twenty-minute exercise. Ask each of them what they did before, and how long it took. If the answers cluster, you have found your segment, and the answer is usually that you were marketing to a broader group than the one that actually has the pain.
That makes it a positioning problem rather than a conversion problem — the positioning framework starts by naming the alternative you replace.
The least glamorous and most fixable. Check whether a decision point exists at all.
Free products with no deadline convert badly because nothing forces a decision. Freemium versus free trial covers the model choice, and trial length covers where the deadline should sit.
Before changing anything, check how many of your non-converters ever came back a second time. A signup who never returned is an activation or traffic problem. A signup who returned four times and did not pay is a pricing or value problem. Those need opposite fixes.
| Finding | Do this | Not this |
|---|---|---|
| Launch traffic converts far worse than organic | Ignore the launch cohort. Get more organic. | Redesign the product |
| Under 20% activate | Remove one step before first value | Lower the price |
| They activate, do not pay | Interview your paying customers | A/B test the headline again |
| No deadline exists | Add one | Add features |
The right-hand column matters. Lowering price and testing headlines are the two most common responses to poor conversion, and both are wrong for three of the four causes above.
They are worth one email, not a campaign. A single message asking one question — what were you hoping this would do for you? — typically gets a low response rate and unusually honest answers.
Do not try to reactivate them with a discount. You do not yet know why they left, and a discount to someone who never activated confirms nothing except that they still will not use it. For the fuller sequence — segmenting the list, the message itself, and when to stop emailing someone entirely — see what to do with the ones who already went cold.
If this was a vibe-coded build that went from launch to stall quickly, the gap is usually the business layer rather than the product — what comes after vibe coding covers that transition, and trial-to-paid conversion covers the mechanics of the fix.
One of four things: the wrong people arrived, they never activated, they activated but did not need the product badly enough, or you never gave them a reason to decide. Check them in that order — most founders assume the third and it is usually the first or second.
Poorly, as a rule. Launch audiences contain a large share of people signing up because the product is interesting rather than because they have the problem. Split conversion by traffic source before drawing any conclusion — your real rate is the non-launch number.
Above 50% of signups reaching first value means activation is fine and the problem is elsewhere. Below 20% means activation is the problem and nothing else matters yet — no pricing or messaging change will compensate for people never seeing the product work.
Rarely, and not before diagnosing. Price is the wrong lever for three of the four common causes. If people never activated, they never got far enough to judge the price at all.
What does your product replace for the people who paid, and why did the others not need that replacement badly enough? Everyone who pays for software is replacing something specific, and your non-converters either had nothing to replace or what they had was good enough.
One email asking a single question — what were you hoping this would do for you — is worth sending. A reactivation campaign or a discount is not, because you do not yet know why they left and a discount to someone who never activated confirms nothing.
Bring your signup count, activation rate and paying customers. Marcus names the break and gives you one action for this week.
Try GhostCoach free →14-day free trial · cancel anytime · 30-day money-back on Lifetime