Table of Contents
104
4 min
03.09.2026
What free users really cost: the economics of freemium
Oleksiy Bazyura
Financial Expert at Finmap
«We were proud of 40,000 free users. Then we counted what they cost in servers and support, and how many of them convert to paying. It turned out the paying ones barely covered the free ones. We were scaling not the business but the costs.»
This is a common trap in freemium SaaS. A large base of free users looks like success and is pleasant to tell investors about. But «free» is only for the user. For you each of them has a price: servers, support, onboarding, and the money spent to acquire them. The question isn't how many free users you have, but whether the paying ones cover them.
Why free users aren't free
Every active user, paying or not, consumes resource. Servers and traffic, support tickets, space in the database, sometimes the team's onboarding time. Plus you acquired some of them with paid ads: the CAC is spent, and there's no revenue from them and never will be unless they convert. In total, the free tier is a real monthly cost — just smeared out and invisible until you count it.
The key number: how many free users one paying user carries
The whole economics of freemium rests on conversion — the share of free users who become paying. If conversion is 3%, then each paying user accounts for about 32 free ones. That means the revenue from one paying user has to cover serving them plus thirty-two free users. While it covers, the model is healthy. The moment conversion drops or the cost of serving free users rises, the paying ones can't carry it — and the more free users you have, the deeper the hole.
| Movement for the month | Amount |
|---|---|
| Free users | 10,000 |
| Cost to serve (≈$0.50/mo each) | −$5,000 |
| Paying (3% conversion) | 300 |
| Margin from paying ($30/mo each) | +$9,000 |
| Net contribution | +$4,000 |
The model is in the black while conversion holds at 3%. But let it drop to 1.5% — paying users fall to 150, margin to $4,500, while serving the free ones is still $5,000, and the net contribution goes negative. The same free tier that brought $4,000 starts costing you $500 a month. And advertising it even harder only deepens the loss.
«A free base isn't an asset by itself. It becomes one only when you know exactly what share converts and what it costs to keep the rest. Otherwise it's just an expensive shop window.»
When freemium works and when it bleeds money
Freemium is healthy when the cost of serving one free user is low (the product scales without people) and conversion is steady and high enough for the paying ones to cover everyone. It bleeds money when free users need expensive support or heavy servers and conversion is low. Then a free trial or tighter free-tier limits are often healthier than «forever free for everyone».
What it looks like in real life
You hear the problem in typical phrases. «We have 50,000 users!» — while a couple of percent pay. «Ads bring lots of signups» — while revenue doesn't grow with them. «Servers get more expensive faster than revenue.» «Support is swamped, and most tickets are from people who'll never pay.» Each line is about a base people are proud of without counting what it costs.
How to see it for yourself
To understand whether the free tier pays for itself, you need to put two things side by side: the cost to serve (servers, support, acquisition) and the margin from paying users. In Finmap you track subscription revenue and the direct costs of infrastructure and support separately — and see whether the paying segment covers the cost of the whole base or the free users quietly eat the profit. Then decisions about free-tier limits and acquisition budget are made on numbers, not on pride in a big base.
Going deeper — SaaS unit economics in plain words and MRR and churn in real money.
A few closing tips
- Count the cost to serve one free user — servers, support, acquisition. Without it the free tier looks free to you too.
- Watch conversion: it determines how many free users one paying user can carry.
- Limit the free tier so it delivers value but doesn't fully replace the paid one.
- Don't scale free acquisition until you're sure the paying users cover it — otherwise you scale a loss.
- Compare freemium with a trial: if serving is expensive and conversion low, a limited trial is often healthier.
A large free base is a good story but not always a good business. It becomes an asset only when you know the cost to serve it and the conversion to paying. The moment those numbers sit side by side, freemium stops being a growth bet and becomes a manageable model.
Money Doesn't Disappear. You Just Don't See It.
Want to see whether your free tier pays for itself? Book a Finmap demo — in 30 minutes we'll show how to line up the cost to serve against the margin from paying users in your SaaS.
Share valuable content — become a source of insights
Frequently asked questions
What freemium conversion is considered normal?
For most SaaS it's 2–5% from free to paying. But what matters isn't the percentage itself but whether it's enough for the margin from paying users to cover the cost of serving the whole free base.
How do I calculate the cost to serve a free user?
Add the direct costs of infrastructure and support and divide by the number of active users. Add part of CAC if you acquire free users with paid ads. That's the price of one «free» user per month.
When is it better to replace freemium with a trial?
When serving free users is expensive (live support, heavy servers) and conversion is low. A time-limited trial reduces the number of «forever free» users you have to carry at your own cost.
Should I scale free-user acquisition?
Only when you're sure the paying users cover them at the current conversion. Otherwise each new free user is an added cost with no revenue, and scaling only deepens the loss.
Any questions left?
We are ready to answer them.
Finmap support
