/unstuck — "Signups tripled but revenue is flat"

/unstuck diagnosis

The question

We switched to freemium 8 months ago. Signups tripled but revenue is completely flat. We thought more users in the door would mean more conversions. What are we doing wrong?

The reframe

You don't have a conversion problem. You have an ICP problem — your free tier is attracting users who were never going to pay. The signups tripled because free removed the filter that was keeping non-buyers out. Before optimizing the paywall, you need to know who's signing up and whether they match your paying customer profile.

Zoom PLG Model Canva Freemium Dan Olsen PMF Pyramid Pricing Roadmap Lenny Rachitsky Product Analytics

The Diagnosis

The freemium switch did exactly what freemium does: it removed friction at the top of the funnel. Signups tripled because you eliminated the #1 barrier to entry — price. But here's what the data is actually telling you: you tripled the number of people entering your building, but the new visitors are walking past the cash register and sitting in the lobby.

The question everyone asks at this stage is "how do we convert more free users?" That's the wrong question. The right question is: are these free users convertible at all?

Why this happens (Zoom PLG Model)

Zoom's freemium worked because their free tier had a natural conversion trigger — the 40-minute limit on group calls. The limit wasn't arbitrary. It hit at exactly the moment when the user was getting value and their meeting participants were engaged. The pain of hitting the wall exceeded the pain of paying.

Your free tier needs to pass this test: does the user hit a natural wall at the moment they're getting the most value? If your free tier is generous enough that users never hit the wall, you haven't built a conversion engine — you've built a free product with an optional donation button.

The ICP split (Dan Olsen PMF Pyramid)

Before freemium, your paywall acted as a filter. Only people who valued the product enough to pay would sign up. After freemium, that filter is gone. You now have two populations in your user base:

Population A: Users who match your paying customer profile — they have the problem, they have budget, they'd pay to solve it. These existed before freemium and still convert at roughly the same rate.
Population B: Users who signed up because it's free — they're curious, they're testing, they're using it for a lightweight use case that will never justify payment. These are your tripled signups.

The ratio between Population A and Population B is the number that matters. If Population A is 15% of your free users and Population B is 85%, then your "3x signups" only produced the same number of potential converters — they're just buried in a much larger haystack.

Diagnostic questions

What percentage of free users match your paying customer profile?

Threshold: above 30% = conversion problem. Below 15% = ICP problem.

If below 15%, optimizing the paywall won't help. You're trying to convert people who don't have the problem your product solves. The fix is upstream — change what your free tier attracts, not how it converts.

What's the activation rate for free users vs. your pre-freemium users?

Threshold: within 20% of each other = healthy. More than 50% gap = wrong users.

If free users aren't reaching the "aha moment" at anywhere near the rate paid users did, the new users aren't just non-payers — they're non-users. They signed up and left. That's a vanity signup, not a lead.

Which specific feature or limit would your best customers hit on the free tier?

Threshold: if you can't name it in 5 seconds, you don't have a conversion trigger.

Zoom had the 40-minute wall. Canva had the brand kit. Figma had the team collaboration limit. If your best users can get full value on free, there's no economic reason to upgrade — just an ethical one. Ethics don't scale.

Conditional Paths

If it's an ICP problem (Population B > 70%):

Don't remove freemium — redesign what free means. Add qualification friction back at the right point. Make the free tier serve a specific use case that naturally leads to the paid use case. Example: free for individual use, paid the moment a second team member is involved. This filters for teams (who pay) without filtering out individuals (who discover).

If it's a conversion trigger problem (Population A is healthy but not converting):

Your free tier is too generous. Find the moment of highest engagement and put the wall there. Analyze where your best paying customers were when they decided to upgrade. That behavioral moment — not a feature checklist — is where your limit should live. The wall should feel like an interruption of momentum, not a locked door on an empty room.

If it's a pricing problem (users hit the wall but don't pay):

The gap between free and paid is too wide. You might need a middle tier. If free is $0 and paid is $49/month, the decision isn't "should I upgrade?" — it's "should I 49x my spend?" A $12/month tier that unlocks the one thing power free users want most can bridge the gap and create a second conversion event from $12 to $49 later.

Critical Path

Segment your free users this week. Split them into Population A (matches paying profile) and Population B (everyone else). Use company size, use case, or engagement pattern — whatever your best customers have in common.

Measure Population A's conversion rate separately. If it's close to your pre-freemium conversion rate, freemium isn't broken — your top of funnel is just attracting more noise. That's fixable.

Identify your conversion trigger moment. Interview 5 customers who upgraded from free. Ask: "What were you doing in the product the moment you decided to pay?" That moment is your wall.

Redesign the free tier around the trigger. Everything before the trigger moment: free. Everything after: paid. Don't add features to free — remove the ones that let people get full value without converting.

Give it 60 days and re-measure. Signups will drop — that's the point. You're reinstalling the filter. Revenue should inflect within 6-8 weeks if the ICP is right.

Built with /unstuck — 155 business frameworks. One diagnosis.