Start with where visitors leave
Before you pick a format, find out where good-fit visitors actually drop out. Two leaks look alike from the outside, and they need different fixes.
- Before signup. The visitor reads your site and never creates an account. A recording doesn’t convince them, and an account feels like a commitment just to look.
- After signup. The visitor creates an account, meets an empty product and a pile of setup work, and never reaches the point where it’s useful.
Measure both ratios with what you already have: visitors to your product and pricing pages who sign up, and signups who reach a first moment of value. You choose what that moment is. Whichever ratio is worse is your leak. If you can’t say which one it is, no format will fix it by luck, so spend a week finding out first.
The four options
Almost every way of letting someone try B2B software is one of these four, or a mix.
| Option | What the visitor gets | What it costs you | Works best when | Breaks when |
|---|---|---|---|---|
| Free trial | A real account, empty, for days or weeks | A real tenant for every signup: running cost, emails sent, support, and abuse | The product is useful within minutes of signing up | Value needs setup or the visitor’s own input, or trials get farmed |
| Sample data on signup | A real account, already filled in | The same as a trial, plus a seed script and cleaning up the fake rows | The leak is after signup: the empty-product problem | The leak is the signup step itself |
| Captured demo (a tour) | A replay of captured screens or page HTML, with hotspots and steps | Little: someone builds it in an editor, with no engineering | Value is visible on screen and a recording tells the story | The visitor needs to see what the product does with their own input |
| Live sandbox | The real product, temporary and pre-filled, with no signup | A demo deployment and database, an integration, and limits on use | Value appears only after real work, and visitors leave before signing up | A tour would do, or the demo needs email, payments or integrations |
Free trial
- What the visitor gets
- A real account, empty, for days or weeks
- What it costs you
- A real tenant for every signup: running cost, emails sent, support, and abuse
- Works best when
- The product is useful within minutes of signing up
- Breaks when
- Value needs setup or the visitor’s own input, or trials get farmed
Sample data on signup
- What the visitor gets
- A real account, already filled in
- What it costs you
- The same as a trial, plus a seed script and cleaning up the fake rows
- Works best when
- The leak is after signup: the empty-product problem
- Breaks when
- The leak is the signup step itself
Captured demo (a tour)
- What the visitor gets
- A replay of captured screens or page HTML, with hotspots and steps
- What it costs you
- Little: someone builds it in an editor, with no engineering
- Works best when
- Value is visible on screen and a recording tells the story
- Breaks when
- The visitor needs to see what the product does with their own input
Live sandbox
- What the visitor gets
- The real product, temporary and pre-filled, with no signup
- What it costs you
- A demo deployment and database, an integration, and limits on use
- Works best when
- Value appears only after real work, and visitors leave before signing up
- Breaks when
- A tour would do, or the demo needs email, payments or integrations
Mixing is normal. A short tour on the homepage that explains the product, and a live sandbox for the visitors who want to try it themselves, answer different questions.
Choose by the shape of your product
Three questions get you most of the way.
- Can a visitor see the value without giving the product anything? If the value is visible on one screen, a tour is enough, and it’s the fastest thing to build.
- Does the value only show once the product has done something with their input? That might be a document, a dataset, a prompt or a configuration. A recording can’t do that. Look at a sandbox or a trial.
- Does the first session need setup you can’t skip? Connecting a third-party account or importing data is hard to offer anonymously. A sandbox works when you can seed what’s needed. Otherwise sample data on signup, or a plain trial, fits better.
What a free trial really costs
A trial’s cost is mostly paid by accounts that never convert, and it grows with every signup instead of every customer. Each one is a real tenant, so each one brings:
- Compute and storage for accounts that sit idle for days or weeks.
- Emails sent to people who never come back, which also hurts deliverability.
- Support questions from people who aren’t going to buy.
- Billing and integration code paths exercised by strangers.
- Abuse: throwaway addresses collecting free resources.
Teams then add gates, such as card-on-file, waitlists and verification, and the friction that the trial was meant to remove comes back. If you’d like to see the effect for your own product, the calculator on the pricing page takes your numbers. Its example values are placeholders, not benchmarks.
Where the tools fit
Capture-based tools build a demo from captured pages or media, in an editor that doesn’t need an engineer. That’s why they’re quick to ship. Navattic, Storylane, Supademo and Walnut work this way, and each also lists sandbox-style demos somewhere in its plans. TestBox puts buyers in your real product with synthetic data, built around sales teams sending sandboxes to buyers.
Trylane is the self-serve end of that range: any website visitor gets a private, temporary workspace in your real app with no signup, outside calls switched off, and a way to keep their work. It needs an engineer, and it fits TypeScript apps on PostgreSQL. Each comparison below says when the other tool is the better choice.
Test before you commit
You don’t need to build anything to learn which format fits.
- Measure the two ratios for a month, and write down what your first moment of value is.
- Ask the last five good-fit people who didn’t sign up what they were looking for and what stopped them. Their words will tell you more than a dashboard.
- Start with the cheapest format that matches the leak: a tour if visitors can’t see the value, sample data if signups find an empty product, a sandbox if visitors won’t sign up just to look.
- Decide what counts as success before you launch, such as more qualified signups from the same traffic, and judge it after weeks, not days.
No format suits every product. If yours runs on TypeScript and PostgreSQL and the leak is before signup, hands-on sessions are worth a look. If it isn’t, a tour or sample data will probably serve you better, and that’s a fine outcome.
Want visitors to try your product before they sign up?
Trylane is free for one product. It works with TypeScript apps on PostgreSQL, and your coding agent can do most of the setup.