Trial length is not a preference. It follows from how long your product takes to prove itself — and longer is usually worse.
Free trial length is not a preference. It follows from one number: how long your product takes to demonstrate value. Get that number and the trial length answers itself.
Fourteen days fits most solo SaaS products. Thirty is usually a mistake disguised as generosity.
Longer trials feel more generous and convert worse. The reason is that trials do not fail from insufficient time — they fail from insufficient urgency.
Time-to-value is the gap between signup and the first moment the product visibly does its job. Measure it in minutes for the users who get there, and separately track what share get there at all.
Your trial has to comfortably exceed that number, with room for a person who signs up on a Tuesday, forgets, and returns the following weekend. It does not need to exceed it by three weeks.
If you have not measured this, that is the more urgent job — the onboarding framework covers naming and instrumenting the activation moment.
Three reasons, all behavioural.
Deadlines create action. Most trial activity clusters near the start and near the end. A thirty-day trial produces a longer dead middle, not more usage.
People forget. A month is long enough to sign up, get distracted by a busy fortnight, and have no memory of why the product mattered when the card is charged.
You wait longer to learn anything. With a thirty-day trial you find out whether a pricing change worked six weeks after making it. On fourteen days you know in three.
| Length | Fits | Watch for |
|---|---|---|
| 7 days | Single-session value, simple products | Too short if setup takes an evening |
| 14 days | Most solo SaaS | Needs a working reminder sequence |
| 30 days | Monthly-cycle products only | Long dead middle, slow learning |
The second decision, and it interacts with the first.
Requiring a card produces fewer trials and much higher conversion, because it filters for intent before anyone starts. Not requiring one produces more trials, more support load, and a conversion rate that looks alarming until you account for the volume.
For a solo founder the support load usually decides it. Two hundred no-card trials generate real inbound volume from people who were never going to buy, and that time comes out of the same week as everything else.
A reasonable default: require a card, keep the trial at fourteen days, and make cancellation genuinely one click. That combination is honest and it protects your attention.
Trial length is a smaller lever than what happens during it. A fourteen-day trial with no communication converts worse than a seven-day trial with three well-timed messages.
The minimum viable sequence has four triggers, and three of the four are behavioural rather than time-based.
Triggering on the absence of the activation event rather than on elapsed time is what separates a sequence that helps from one that reads as noise.
Before changing trial length, check your activation rate. If under twenty percent of trials reach first value, the trial length is not the problem and extending it will not help. Fix activation first.
Granting an extension when someone asks is nearly always correct. Someone who asks for more time is engaged, and the request itself is a strong buying signal.
Automatic extensions for everyone are a different thing and they do not work. A trial that quietly renews teaches people that the deadline is not real, which removes the mechanism that makes trials convert at all.
The trial length also has to match the price it protects — a higher price justifies a longer evaluation, which is one of the inputs in the pricing framework. GhostCoach runs a fourteen-day trial for this reason: the product's value shows up in the first session, and two weeks is enough to run several real decisions through it without the deadline becoming abstract.
If your trial length is right but conversion is still poor, the problem is downstream — see trial-to-paid conversion for the diagnosis, and freemium versus free trial if you are questioning the model itself.
It follows from time-to-value. Seven days if value is visible in one session, fourteen if it needs a few uses across a week, thirty only if value depends on a monthly work cycle such as reporting or billing runs. Fourteen days fits most solo SaaS products.
Usually not. Longer trials feel generous and convert worse, because trials fail from insufficient urgency rather than insufficient time. A thirty-day trial produces a longer dead middle, more forgetting, and slower learning about whether changes worked.
For solo founders, usually yes. Requiring a card produces fewer trials and much higher conversion by filtering for intent. No-card trials generate significant support volume from people who were never going to buy, and that time comes out of the same week as everything else.
Four triggers, three of them behavioural: one message on signup with a single action, a message when someone has not activated within 48 hours, a reminder three days before expiry, and one at expiry. Triggering on the absence of activation beats timing on elapsed days.
Yes. Someone who asks for more time is engaged, and the request is a strong buying signal. Automatic extensions for everyone are different and do not work — a trial that quietly renews teaches people the deadline is not real.
Give Marcus your activation rate and time-to-value. You get a specific trial length and the sequence to run inside it.
Try GhostCoach free →14-day free trial · cancel anytime · 30-day money-back on Lifetime