Freemium looks obvious: make the product free, lots of people come in, and some end up paying. In practice it is the most demanding model of the lot, and it eliminates without mercy anyone who adopts it without checking the arithmetic.
The question is never "is freemium good?" but "do I have the volume it requires?".
Definition
Freemium offers a permanently free version of a product, with no time limit, and charges a subscription for advanced features or higher volumes. A minority of paying users funds the whole thing.
It differs from a free trial on one decisive point: a trial ends, freemium does not. A free user can stay ten years, consuming your resources every day.
The arithmetic that decides
Free-to-paid conversion generally sits between 1 and 5 percent, and the top of that range is rare. That number governs everything else, because it sets what a free user is worth, and therefore what you are allowed to spend on them.
Every free user stays profitable. Bringing more people in improves the result.
The threshold is price times conversion rate: the average monthly value of a free signup once spread across everyone. It is the real ceiling on what the free service can cost you, hosting and support included.
Move the sliders above. What appears is the number that decides everything: at 3 percent conversion and €19 a month, a free signup is worth an average of 57 cents a month. That is the absolute ceiling on what the service you provide them can cost you, hosting and support included.
That ceiling is low, and crossing it cannot be fixed with volume. Quite the opposite: if each free user costs more than they bring, attracting more people deepens the loss instead of closing it. That is the exact inverse of the intuition that pushes people to open a free tier to "get volume".
This is what disqualifies freemium on products with high marginal cost. A service burning compute or storage on every use passes 57 cents without effort, and every new free signup then becomes a net liability.
Where to draw the line
This is the model's trickiest decision, and there are only three families of answer.
| Type of limit | Works well when |
|---|---|
| By volume (number of items) | Value grows with usage |
| By feature | Advanced needs are clear and separable |
| By collaboration (number of seats) | The product spreads inside a team |
The classic mistake is making the free tier too generous out of fear of disappointing. A satisfied free user is not a future customer, it is a permanent cost. The limit has to collide with real usage, not stay theoretical.
What free earns you anyway
The model is not all cost, provided you go after its benefits deliberately:
- Distribution. A free product gets recommended without friction, which lowers CAC (customer acquisition cost).
- Social proof. Credible usage numbers reassure paying buyers.
- Information. You observe real usage without having to ask.
If none of those three is actively exploited, the free tier is spending with nothing in return.
Frequently asked questions
Freemium or free trial?
A trial suits products where value shows quickly and each user is expensive. Freemium suits products with near-zero marginal cost that benefit from being widespread. For a solo Micro-SaaS, a trial is almost always the safer choice.
What conversion rate should I target?
Between 2 and 4 percent is a decent result for a consumer product. Above 5 percent, check instead that your free tier has not become a disguised trial, which is not a problem but changes the model.
Can I reverse the decision?
Barely. Removing an existing free tier causes disproportionate anger and a wave of departures. If you hesitate, start paid: opening a free tier later is painless, the reverse is not.
How do I convert free users?
At the moment they hit the limit, never before. A message arriving when the user actually wants to do more converts several times better than a calendar-based nudge, and it does not damage the relationship.