The Seven-Day Trial: Should You Ask for a Card Up Front?

In short

• Card up front gets you far fewer trials and a much higher share of them converting. Pick it when your scarce resource is your own attention.
• No card gets you volume and information. Pick it when you still need to learn who your product is for.
• Seven days is long enough for most tools. Longer trials usually extend the decision, not the evaluation.
• Whatever you pick, say on the pricing page whether a card is needed and when the charge lands.

Ask for the card when you are one person and every trial costs you support time, because everyone in the trial has already decided the price is acceptable. Do not ask for it while you are still working out who the product is for, because you need the volume more than the filtering. The cost of taking the card is real and specific: nobody can look at your product before deciding, so your landing page and your docs have to do the job the trial used to do.

I changed this on Lighthouse recently. Until this month you could build a waitlist page without an account and share it, then pay if you wanted to keep it. That entry is gone: the only way in now is a seven day trial with a card taken at checkout, charged on day 8 unless you cancel.

This is the reasoning, what it costs, and the parts that have to work before day 8. If you have not settled on the number yet, that comes first in how to price your first indie SaaS.

Table of contents

• The three models, and who each one fits
• Why I took the card
• How long the trial should run
• What has to work before day 8
• Wiring it up without surprises
• What this costs you
• Frequently asked questions

The three models, and who each one fits

There are only three sensible shapes for a small product, and they trade the same two things against each other: how many people try it, and how much of your time each one costs.

A free plan earns its keep when free users create value for paying ones, through a network effect or a public page carrying your name. If free users are simply cheaper versions of paying users, it is a cost centre with a marketing story attached.

Why I took the card

One person builds and supports Lighthouse. That makes attention the thing in shortest supply, and a card is the cheapest filter available for it. Everyone in the trial has already looked at $19 or $29 a month and decided it is fine, which changes what their questions are worth.

The second reason is that it makes the numbers readable. With a free entry, a signup could be a buyer or a browser, and the two arrive through the same form. That ambiguity quietly ruins every other measurement you take, including which channel is working, which is the point of knowing which channel brought your signups.

I should be straight about the evidence: I have not run this long enough to have conversion numbers, so this is a reasoned bet rather than a result. If you want the version with numbers attached, wait for someone who has them.

How long the trial should run

Seven days, for most tools. The useful question is not how long someone needs to evaluate your product, it is how long until they can see it work once. If your product delivers something on day one, a thirty day trial does not give them thirty days of evaluation. It gives them twenty nine days to forget they signed up.

Match the length to how long a real result takes:

• Result in minutes, such as publishing a page or sending a campaign: seven days is generous.
• Result needs other people, such as collecting responses or inviting a team: fourteen, because the clock includes other people being slow.
• Result needs a cycle, such as a monthly report: either run thirty days or reconsider whether a trial is the right shape at all.

Lighthouse sits in the first group. Putting up a waitlist page takes a couple of minutes, so a longer trial would only delay the decision.

What has to work before day 8

A card-up-front trial moves the risk from getting people in to making the first session count. Three things decide it.

One result in the first session. Not a tour or a setup checklist: the actual thing your product does, done once. If someone reaches day 8 without having seen a result, the charge feels like a trick regardless of what the page said. The detail is in the first five minutes.

A reason to come back before the end. Something has to happen while they are away that is theirs: a signup arriving, a response landing, a number moving. A trial with no second visit converts on momentum alone, and there usually is not any. That is the same mechanic as why users do not come back.

An obvious way out. Cancelling has to be one click from the dashboard, and the page has to say so before they hand over a card. Hiding it does not keep anybody, it just converts a cancellation into a chargeback and a bad review.

Wiring it up without surprises

On Stripe this is one setting on the checkout session, a trial length in days. The work is in the events afterwards, and the mistake is treating them all as the same thing. The ones that matter:

One detail worth setting deliberately: Stripe can send a reminder before a trial ends, and it is off unless you turn it on. Lighthouse does not send one from the app. Whatever you choose, choose it on purpose, because a charge nobody expected is the most expensive support conversation you can have.

What this costs you

Three real costs, worth naming before you copy the decision.

• Nobody can look first. That is the whole point of the filter, and it is also its price. The landing page, the docs and the screenshots now have to answer every question the product used to answer by being tried.
• The phrase "free trial" gets harder. It is still free for seven days, and plenty of people read it as "no card needed". Say what you mean next to the button rather than in a footnote.
• You learn from fewer people. Filtering out browsers also filters out the ones who would have told you something useful on their way past.

If those costs sound heavier than the support load you are trying to protect, you probably want the no-card trial instead. That is a reasonable answer and most products start there.

Decide it by what you are short of. Short of information: do not ask for a card. Short of hours: ask. Then make the first session deliver something real, say plainly when the charge lands, and make cancelling easy enough that nobody has to email you to do it.

Frequently asked questions

Should my first product take a card up front?

Usually not. Early on you need to learn who the product is for more than you need to protect your time, and a no-card trial gets you more of those people. Switch when support starts eating the hours you would rather spend building.

Is seven days too short?

Only if your product cannot show a result in that window. Measure from signup to the first real outcome. If that takes two days, seven is plenty; if it takes two weeks, the trial is the wrong length or the onboarding is.

Do I need to email people before the trial ends?

It is kinder than not doing it, and it costs you some conversions that were never really conversions. Stripe can send that email for you, so the real decision is whether you want the people who forgot, and the answer is usually no.

What do I do when someone asks for a refund after day 8?

Decide the policy before it happens and put it on the page. Lighthouse does not refund partial months, and cancelling stops the next charge while access runs to the end of the period. Whatever you pick, the version that causes trouble is the one you improvise.

Can I run a free plan and a trial at the same time?

You can, and it usually confuses the pricing page. If the free plan is genuinely useful, the trial is redundant. If it is not, it is a worse version of a trial with no end date.

Lighthouse is a waitlist, survey, newsletter, and feedback toolkit for indie founders, at $19 or $29 a month with a seven day trial. From an indie dev, for indie devs and makers.