Why it matters
A trial is the last step before a customer pays. This rate shows whether the people who start a trial find enough value to stay, so it sits close to revenue and moves it quickly.
It is a cohort measure. Trials that are still running have not had the chance to convert, so counting them lowers the rate for no good reason. Group trials by the week they started, wait until every trial in the group has ended, and then count how many became paying customers.
The formula
In words: the paying customers that came from a start cohort, divided by the number of that cohort’s trials that have ended, then multiplied by 100.
In symbols:
trial_to_paid_rate = paying_customers_from_cohort / ended_trials_in_cohort x 100- The cohort is every trial that started in one period, such as one week.
- Only ended trials go in the denominator. A trial that is still running is neither a win nor a loss yet.
- A customer who pays once and cancels later still counts as converted. The later loss belongs to churn.
A worked example
Example numbers (made up for illustration): 600 trials started over three weeks. All of them have ended, and some of those people became paying customers.
| Start week | Ended trials | Paying customers | Rate |
|---|---|---|---|
| Week 1 | 200 | 30 | 15.0% |
| Week 2 | 200 | 36 | 18.0% |
| Week 3 | 200 | 30 | 15.0% |
| Whole cohort | 600 | 96 | 16.0% |
The cohort produced 96 paying customers from 600 ended trials, a trial to paid rate of 16.0%. Each start week keeps its own rate, so a strong week and a weak week stay visible.
What published data says
Published figures give a loose reference point, and they differ with how a trial is set up. Treat them as context for your own cohorts, not as a target.
- Median free-to-paid conversion rate across 200 products: 8%. ChartMogul, The Conversion Report (Kyle Poyar, 4 Feb 2026), January 2026 study of 200 products. Read on 18 Oct 2026. Source.
- Free-to-paid conversion of trials that require a credit card: 30%. Same ChartMogul report: trials requiring a credit card converted at 30% in the January 2026 study. Read on 18 Oct 2026. Source.
How to improve it
Most of the gain comes from getting a new customer to a useful result while the trial is still running.
- Find the first result that new customers value and point your onboarding messages at it.
- Before the trial ends, say what will be billed and on what date, so the first payment comes as no surprise.
- Ask the trials that did not convert why they left. The same few reasons usually repeat.
- Compare trial designs, such as length or card up front, on separate cohorts and not on one mixed group.
Common mistakes
- Dividing by trials started instead of trials ended, which scores unfinished trials as losses.
- Blending cohorts, so a good week and a bad week hide each other.
- Comparing a card-required trial with a no-card trial as though they were one kind of trial.
- Counting a customer as converted when they only entered a card, before any payment was collected.
Work it out with your own numbers
The trial conversion rate calculator takes your trial and paying counts and shows the rate for a cohort. Open the trial conversion rate calculator and enter your own figures. It runs in your browser and shows each step of the sum.
Read next
- Visitor to trial conversion rate: the step before the trial starts.
- Freemium conversion rate: the same idea for products that start with a free plan.
- Customer acquisition cost: how the conversion rate feeds the cost of winning a customer.
Questions
Which trials belong in the denominator?
Only trials from the cohort that have ended. A trial that is still running cannot yet be counted as converted or as lost.
When is a trial customer counted as paid?
When the first payment is collected. Entering a card is a step on the way, not a payment.
Does requiring a card change the rate?
It changes who starts a trial, so the two kinds of trial give rates that are not directly comparable. Keep them in separate cohorts when you track them.
How big should a cohort be?
Large enough that one customer more or less does not move the rate noticeably. If a week is too small, group two or four weeks together and keep the grouping fixed.