Signups are up, and revenue is not. People start the free trial, look around for a few minutes, and are never seen again. The trial is doing its job as a marketing offer and failing at the only job that counts: producing someone who would be annoyed if you took the product away.
A trial is a promise about time, not a discount.
Nobody signs up because thirty days are free. They sign up because they have a problem this week, and your page implied the product would take it off their desk by Friday. That is a claim about a short and specific window, and about whether they will feel anything happen inside it.
If you sell SaaS, the trial is the closest thing you have to a salesperson who works nights, and it is being asked to close without ever having heard the objection.
That framing changes what you build. A discount asks people to decide whether the price is fair. A promise about time asks you to decide what the first useful outcome is and what stands between a new account and that moment. Plenty of startups have never written that sentence down. Until you can say what the first win is and roughly how long it takes, the trial has no target, and neither does the product team.
The first session decides the trial.
Much of what goes wrong goes wrong immediately. An empty account with no data, a setup step that needs a colleague who is out for the week, a required integration nobody warned them about, a screen that assumes they already know your vocabulary.
Watch what a new account actually does in its first twenty minutes. Not a funnel chart of the whole trial, the first twenty minutes. That is where the trial is won or lost, and it is the cheapest thing on this list to fix, because the fixes are usually sample data, a shorter form, and three sentences of plain language where there is currently a term you invented.
The rest of the trial mostly plays out that first session.
A person in the trial beats an email sequence.
Automated trial emails do a job. They are not the job. At the scale a seed-stage startup is operating at, the founder can read the list of accounts that started this week and write to the ones that look like real buyers, individually, with a sentence that proves someone looked.
This is the part founders resist because it does not scale, and the resistance is misplaced. The point isn’t to convert the people in this week’s list. It is to learn, in their words, what they were trying to do and where they stopped, so that the ones after them convert without the email. Startups win on execution, and this is what that looks like on a Tuesday afternoon.
Ask for the card when the work is done.
Many trials never ask. The last email says the trial has ended and hopes the reader concludes.
The moment to ask is the moment the promise was kept: the first report they exported, the first invoice they sent, the first thing that would now have to be undone. Ask there, in the product, in one line, naming the thing they just did. Asking at the expiry date instead asks the customer to remember a feeling from ten days ago.
Whether the card comes up front at signup is a separate argument with reasonable people on both sides. Card up front gives you fewer, cleaner trials. No card gives you volume and noise. At seed stage, when the sample is small, and every conversation is research, we prefer volume because it correlates to more awareness of your brand.
An expired trial is still a conversation.
A trial that ends without a purchase is not a failure; it is a person who told you something specific and was never asked what. One simple email asking what they were hoping the product would do is answered more often than founders realize, and replies are cheap product research.
Keep those answers somewhere you can count them. Three people saying the same sentence about why they stopped is a roadmap item; one is an anecdote. That distinction is the whole of measuring marketing honestly at this stage.
Before we touch the marketing around a trial, we want to understand how new accounts start one. Trial length should accommodate onboarding, experimentation, and embedded operation—allowing prospects to get a feel for what you offer.
Tell us what you’re building.