The Freemium Gate
A permanent free tier whose edges are drawn where paying starts to make sense.
The problem it solves
A free tier that gives too little never grows an audience. One that gives too much never produces revenue. The gate has to sit on a line that moves with how people use the product.
freemium: A permanently free tier designed to convert some users to paid. is a segmentation exercise wearing a pricing costume. The free tier is not a discount. It is a definition of which users you are choosing not to charge, and the boundary has to track something that grows as a user gets more serious.
What the gate can be drawn on
- Capacity
- Storage, records, projects, minutes. Scales with commitment and is easy to explain. Risk: heavy users on the free tier can cost real money to serve.
- Seats
- Free for one person, paid for a team. The cleanest line in collaboration software, because the moment a second person arrives the value jumps.
- Capability
- Advanced features held back. Easy to build, easy to resent. Works when the held-back features serve a genuinely different job, such as audit logs or SSO.
- Support and control
- Free gets docs, paid gets a human and an admin panel. Invisible to individuals, decisive for companies.
The free tier should be good enough to use for real work and shaped so that real work eventually crosses the line. If nobody ever crosses it, the line is in the wrong place. If everybody crosses it in the first week, you have built a trial and mislabelled it.
A gate that breeds resentment
- Exporting your own data requires an upgrade
- The free tier deletes work after 30 days
- A core action is capped at a number nobody can predict
A gate that reads as fair
- Export is always free, volume is what costs
- Old work stays readable, new work needs room
- The cap is a round number shown in the interface from day one
Two costs get underestimated. The first is serving cost: free users consume support, storage, and bandwidth, and at scale they set the floor of your margin. The second is positioning cost: a generous free tier anchors the perceived value of the whole product, and it is very hard to raise that anchor later.
Worked example
A hypothetical seat line
A note-taking tool is free for one person and charges per editor above that. A team on the free tier hits the line the day a second person needs to edit. The upgrade conversation happens between colleagues rather than between the user and a modal, which is usually a better conversation. The risk is obvious: people share one login, so the line needs to be enforced by the product rather than by good faith.
When it fits
- Marginal cost per free user is low enough to absorb for years.
- There is a natural dimension, such as seats or volume, that grows with seriousness.
- Distribution benefits from free users, because they invite colleagues or produce public content.
- The paid tier serves a different job rather than removing an artificial handicap.
When it backfires
- The free tier is complete, so the only reason to pay is guilt.
- The gate blocks data export or account deletion, which turns a pricing choice into hostage-taking.
- Free users cost more to serve than paid users generate, and the gap widens with growth.
- The paid features are the ones the product is actually for, so free users never do anything worth describing to a colleague.
Products using it
Slack
Has drawn its free tier around message history and integrations rather than around the number of people who can chat.
Dropbox
Uses storage capacity as the line, with referrals granting extra space, so growth and the gate use the same currency.
GitHub
Made private repositories free and moved the paid line towards team administration, security, and support.
The psychology under it
The takeaway
Draw the free tier on a dimension that grows with commitment, then leave it alone long enough to learn.
Finished the teardown? Bank it and the day counts toward your run.