Trial-to-paid conversion rate: calculate it by cohort
Calculate free trial conversion rate on finished trial cohorts, compare it by campaign, and tie each trial to its first Stripe payment.
By DATALYR · · 4 min read
Contents
The trial conversion rate formula
Trial-to-paid conversion rate is the number of trials that became paying customers divided by the number of trials started, measured on a group of trials that have all finished. If 400 trials started in October and 72 of them paid, the rate is 72 ÷ 400 = 18%.
The important word is finished. A trial that started yesterday has not had a chance to convert, so including it lowers the rate. Group trials by the date they started, wait until the whole cohort's trial period has ended, then count payments.
Count a conversion at the first successful payment, not when the trial ends or when an invoice is created. An invoice that fails is not a customer.
Avoid the calendar-month shortcut
A common shortcut divides this month's new paying customers by this month's trial starts. It mixes cohorts: some of this month's payments come from last month's trials, and some of this month's trials will pay next month.
In a hypothetical November, 80 trials convert to paid and 380 trials start. The shortcut reports 80 ÷ 380 ≈ 21.1%. But if 30 of those 80 came from September and October trials, the November figure says nothing reliable about November's campaigns.
The shortcut is acceptable only when trial volume is flat for months. When spend or pricing is changing, use cohorts.
Example: compare campaigns by cohort
This is a hypothetical October cohort of 400 trials on a 14-day trial. Campaign A produced 250 trials and 30 paid customers, a 12% rate. Campaign B produced 150 trials and 42 paid customers, a 28% rate. Together that is 72 of 400, or 18%.
Campaign A spent $5,000, so it cost $5,000 ÷ 250 = $20 per trial and $5,000 ÷ 30 ≈ $166.67 per paying customer. Campaign B spent $6,000, which is $40 per trial and $6,000 ÷ 42 ≈ $142.86 per paying customer.
Campaign A looks cheaper on trials and more expensive on customers. Trial-to-paid by campaign is what reveals that, and it is the reason to report the rate per campaign instead of only for the whole business.
Same cohort, different quality
Campaign A
12% → $166.67 per customer
250 trials, 30 paid, $5,000 spend, $20 per trial.
Campaign B
28% → $142.86 per customer
150 trials, 42 paid, $6,000 spend, $40 per trial.
How Stripe marks a trial and a conversion
In Stripe, a subscription on a free trial has the status trialing. Stripe sends customer.subscription.trial_will_end before the trial ends, by default 3 days before, so you can remind the customer. What happens at the end depends on the payment method: Stripe can create an invoice, cancel the subscription or pause it when no payment method is on file.
The conversion is the first successful payment after the trial. In DATALYR, a trial start records access with no money moved, and a trial conversion is the first paid period, so it counts as the acquisition. Later renewals are credited back to the campaign that acquired the customer, and they never count as new customers.
Send trial starts and first payments to ad platforms as separate conversions. Meta's StartTrial and Subscribe events, and LinkedIn's Start trial and Subscribe conversion types, keep the two apart.
Connect each trial to its campaign
A per-campaign rate needs each trial tied to the visit that produced it and to the payment that follows. The visit happens on your website, the trial in your app and the payment in Stripe, often days apart and sometimes on different devices.
DATALYR's Web SDK passes the visitor ID into Stripe Checkout and Payment Links, and the identify call on signup ties the app user to the visit. The first payment then carries the campaign of the visit. When your server creates subscriptions directly, add the visitor ID as metadata.visitor_id.
Check one journey before trusting the numbers: start a trial from a tagged URL, convert it with a test card, and confirm both events sit on one user.
Use the rate to steer spend
Optimizing ads on trial starts is reasonable while paid conversions are too few to learn from, but watch trial-to-paid by campaign as you do it. If a campaign's trials convert at half the average, its cheap trials are not cheap customers.
Pair the rate with activation. A trial that never reaches the product's first meaningful action rarely converts, and activation shows up days before payment, which makes it a useful early read on a new campaign.
Compare against your own earlier cohorts rather than published benchmarks. Trial length, card-required or not, and price all move the rate.
Related articles.
Use cases ·
Web app funnel tracking: connect ad clicks to paid accounts
Map a browser-based app journey from campaign visit to signup, activation, trial, and payment without confusing events with customers.
Use cases ·
Subscription app attribution: connect campaigns to paid conversions and renewals
Separate mobile trials, paying subscribers, renewals, and refunds while verifying the identity handoff through RevenueCat or Superwall.
Measurement ·
CAC vs CPA: distinguish an action from a new customer
Use a worked acquisition funnel to compare cost per action with customer acquisition cost, including repeat purchases, sales costs, and conversion lag.
