Timing is part of the offer
A free trial with a card has five moments where money and the buyer meet: the card entry at sign-up, the days of the trial, the reminder, the first charge and the line on the card statement. Each has a date, and the buyer either knows it or is surprised by it. Most payment timing advice online argues about the psychology of those moments. Before getting to that, it helps to see what is fixed by the payment system, because the mechanics limit what the psychology can do.
| Moment | What Stripe documents | Source |
|---|---|---|
| Card entry | A free trial can start with or without a payment method. Without one, the trial can cancel, pause or invoice when it ends. | Free trial documentation |
| Reminder | With customer emails on, a reminder goes out 7 days before the trial ends. A trial shorter than 7 days sends it at the start. | Trial reminder documentation |
| Reminder event | The customer.subscription.trial_will_end event fires 3 days before the end by default. | Trial offer documentation |
| First charge | An invoice is created when the trial ends and Stripe attempts the charge about an hour later. | Free trial documentation |
| Statement line | Trial wording is added to the statement descriptor, and text after the 10th character can be overwritten by it. | Trial reminder documentation |
The first charge is therefore not a surprise for the system. The date is set the moment the trial starts. Whether it is a surprise for the buyer depends on whether you showed it to them, reminded them and made the statement line recognizable.
What the behavioral research says, and what it does not
The Nobel committee published a popular summary of the work of Richard Thaler, which is a convenient, authoritative statement of several ideas people borrow for pricing. It covers mental accounting, reference points, loss aversion, the endowment effect and the tension between a planning self and a doing self. It is written about household money, savings and markets. It does not mention subscriptions or trials, and neither do the other sources used here report a measured effect of any of these ideas on trial conversion.
That does not make the ideas useless. It means the honest way to use them is as sources of hypotheses with a plan for testing each one. The table keeps what the summary says apart from what it might mean for a trial.
| Idea | What the summary says | A hypothesis for trials | How to check it |
|---|---|---|---|
| Reference points | People judge a price against a reference such as the price they last saw or the lowest they found. | The full monthly price shown beside the trial price becomes the reference for the first charge. | Compare cohorts that saw the later price before the card step and after it. |
| Loss aversion and ownership | Giving up something already owned feels worse than gaining the same thing. In a 1990 experiment, people given a mug valued it more than people without one. | A buyer who has set something up in the product feels cancelling as giving it up. | Compare cancellation among trials that finished setup and those that did not, and treat it as correlation. |
| Present and future costs | Costs and gains close in time take more of our awareness, and a thousand krona next year feels like less than today. | A charge 14 days away feels small at sign-up and large on the day it lands. | Track refund requests and complaints by days since the charge, with and without a reminder. |
| Mental accounts | People split money into separate accounts in their minds and judge each decision against its own account. | A small monthly charge sits in a different account from a large yearly one. | Offer the yearly price beside the monthly one and compare the mix chosen. |
| Defaults | The default option decides the outcome when people do not actively choose something else. | The plan preselected at the card step is the one most people keep. | Change which plan is preselected for half of new sign-ups and compare. |
Asking for the card moves the decision earlier
One timing choice has published numbers behind it, though they are survey averages and not a controlled test. For every 1,000 website visitors, the ChartMogul report lists 45 trial sign-ups and 3.6 paying customers when no card is asked for, against 35 sign-ups and 10.5 paying customers when the card comes first. Its conversion rate for card-required trials is 30%, and 20% of the free trial products in the study ask for a card up front.
Read that as a statement about timing. Asking for the card at sign-up moves the buying decision to the start of the trial, so fewer people enter and a larger share of those who enter pay. Asking at the end spreads the decision across the whole trial and loses people at the moment they are asked. The report also warns that a card requirement can cut sign-ups unless the product warrants the friction, which is why the choice depends on how serious the buyer is.
Where the card network limits the timing
Some products place a hold on the card at sign-up to check that it can cover the price. Stripe documents how long a hold lasts. For card payments made online the hold is typically 7 days. For a Visa payment started by the merchant it is 5 days, with an exact window of 4 days and 18 hours. A trial longer than one window cannot be covered by a single hold, so the hold has to be renewed or the card checked again before the charge.
| Trial length | Windows spanned at 7 days | Windows spanned at 4 days 18 hours |
|---|---|---|
| 7 days | 1 | 2 |
| 14 days | 2 | 3 |
| 30 days | 5 | 7 |
The 14 day trial, which 62% of the products in the ChartMogul report use, spans at least two windows. That is a mechanical fact about card networks and is independent of any psychology. TrialValid handles it by placing a fresh hold before an old one lapses, as the trial holds page describes.
The reminder and the date
The best-supported piece of payment timing advice is also the plainest: tell the buyer the date and remind them before it. Stripe sends a reminder 7 days before a trial ends when customer emails are on, and it exposes an event three days out for those who send their own. The ChartMogul report points to a 30 day trial from a large design tool that promises a reminder 7 days before the end, and its author writes that companies without such a reminder risk fewer sign-ups, weaker retention in months one to three and more refund requests. That is the author's reading, not a measured result, so treat it as a reason to test a reminder, not as a number to expect.
Timing choices worth testing
- When the card is asked for: at sign-up, mid-trial, or at the end. Measure sign-ups and payers per 1,000 visitors together, as the report does.
- Where the post-trial price appears: before or after the card step.
- Whether a reminder goes out and how many days ahead. Measure cancellations in the days after it and refund requests after the charge.
- How the charge appears on the statement. Measure the share of disputes and questions that mention an unrecognized charge.
- Which plan is preselected at the card step.
Test one at a time and size the test first, as the pricing test guide describes. Use the trial conversion rate calculator to turn each result into customers and revenue.
Claims to leave alone
Be wary of articles that give an exact day or hour as the best moment to charge, or a percentage by which a timing trick raises revenue. None of the sources here reports one, and a specific number with no study behind it is a guess. The same goes for claims that a buyer's brain does something specific at a certain moment. The research available supports general ideas about how people weigh costs and gains. It does not give a schedule.
Questions about payment timing
Is it better to charge at the start or the end of a trial?
The published figures favor asking for the card first on average, with fewer sign-ups and more payers per 1,000 visitors. They are survey averages across many products, so test it on yours.
Does a hold on the card keep the money reserved for the whole trial?
No. A hold lasts a number of days that depends on the card network, so a 14 day trial spans more than one window.
How far ahead should a reminder go out?
Stripe defaults to 7 days before a trial ends for its hosted emails. No source here compares other lead times, so treat 7 days as a starting point.
Do loss aversion and mental accounting apply to my trial?
They are general findings. No source here measures them on trials, so form a hypothesis and test it.
Should the price be shown before the card is entered?
Showing the later price early is the safest default, because it gives the buyer the real reference point. Test it against showing it after.
Sources
- Nobel Prize committee, Popular information on the 2017 economics prize (Richard Thaler): https://www.nobelprize.org/prizes/economic-sciences/2017/popular-information/, read 19 Oct 2026.
- Stripe documentation, Place a hold on a payment method: https://docs.stripe.com/payments/place-a-hold-on-a-payment-method, read 19 Oct 2026.
- Stripe documentation, Configure trial offers on subscriptions: https://docs.stripe.com/billing/subscriptions/trials, read 19 Oct 2026.
- Stripe documentation, Use free trial periods on subscriptions: https://docs.stripe.com/billing/subscriptions/trials/free-trials, read 19 Oct 2026.
- Stripe documentation, Manage compliance requirements for trials and promotions: https://docs.stripe.com/billing/subscriptions/trials/manage-trial-compliance, read 19 Oct 2026.
- Stripe documentation, Automate customer emails: https://docs.stripe.com/billing/revenue-recovery/customer-emails, read 19 Oct 2026.
- ChartMogul, The SaaS Conversion Report (Kyle Poyar, 4 Feb 2026): https://chartmogul.com/reports/saas-conversion-report/, read 19 Oct 2026.