- 1 Measure time to value
- 2 Pick two or three trial variants
- 3 Size the test with a calculator
- 4 Split new sign-ups at random
- 5 Wait until every trial has ended
- 6 Compare paid customers per sign-up
Most companies copy the trial length they saw elsewhere. The right length depends on how long it takes your users to reach real value. A trial that is too short ends before they get there. One that is too long removes urgency and delays revenue. Testing it is a pricing decision you can make with data.
Why it works
A trial ending is a deadline, and deadlines drive action. Many users leave a decision until the last reminder, so a long trial mostly delays the moment they choose. A trial that is too short has the opposite problem: it ends before users have seen the result they came for.
The best length sits just past the time most users need to reach value. That point differs for every product, which is why copying another company’s number rarely works and testing does.
How to do it
- Measure time to value first. Check how many days it usually takes new users to complete your activation action. Your trial should comfortably cover that window.
- Choose two or three variants. For example, a shorter trial, your current trial and a reverse trial. Keep everything else the same, including price, plans and onboarding.
- Pick one success metric. Use paying customers per sign-up, measured after every group has had time to decide. Trial-to-paid rate alone can mislead if one variant also changes how many people start a trial.
- Size the test in advance. Use an A/B test calculator to work out how many sign-ups each group needs. Small businesses often need to accept larger differences or run the test longer.
- Split users at sign-up. Assign each new account to one variant and keep them there. Tell sales and support which variant each user is in so their messaging matches.
- Check retention, not just conversion. Look at whether customers from each variant are still paying a few months later. A variant that converts more but churns faster is not a win.
Example
A B2B reporting tool offers a 30-day trial. Its data shows that most users who activate do so within the first week, and many trial users go quiet until the reminder email near the end. The team tests a 14-day trial against a reverse trial where users get the full plan for 14 days and then move to a limited free plan. They let each group run until all trials have ended, then compare paying customers per sign-up and early retention. Whatever wins, they now have a reason for their trial length that holds up in a board meeting.
Metrics to track
- Paying customers per sign-up. The main metric for each variant, read only after every trial has ended. Higher is better.
- Activation rate during the trial. The share of trial users who reach the activation action. A shorter trial should not push this down.
- Time to activation. If it gets close to the trial end, the trial is too short.
- Retention after a few months. Compare paying customers from each variant. The winner should keep its lead.
When it backfires
- Calling it too early. Comparing groups before every trial has ended favours the shorter variant because those users have already decided.
- Changing other things mid-test. A new onboarding flow or price change during the test makes the result impossible to read.
- Free plan cannibalisation. A reverse trial can leave some users happy on the free plan forever. Make sure the paid features are ones active users will miss.
- Too little traffic. With only a handful of sign-ups a week, the test may never reach a clear answer. Make a judgment call based on time to value instead.
Growth Marketing Academy
Learn the framework behind this tactic
These lessons explain the math and the mental model, so you can adapt the tactic instead of copying it.
- Track 5 · Experimentation A/B Testing Statistics That Hold Up Size tests correctly, understand power and significance, avoid peeking, catch sample ratio mismatch, handle novelty and multiple comparisons, and know when not to test.
- Track 6 · Monetization Paywall Design: Freemium, Trials and the Moment You Ask Choose between freemium, free trial and reverse trial, place the paywall at the right moment, and pick hard or soft gates. D2C memberships included.
Questions people ask
What is a reverse trial?
New users get the full paid plan for a set period. When it ends, they drop to a free plan instead of losing access. They keep using the product and can upgrade when they miss the paid features.
How long should a trial length test run?
Long enough for every group to finish its trial and reach a payment decision, plus enough sign-ups for a reliable result. Use a sample size calculator before you start so you know how many users each group needs.
Should I ask for a credit card at the start of the trial?
Treat that as a separate test. Asking for a card tends to reduce trial starts and raise the share who convert, so judge it by paying customers per visitor.