Try it yourself demos that run on visitor data

A "try it yourself" widget lets a visitor run your product on their own data before they sign up, for example by pasting a website address, a domain or a file and seeing a real result. It works best when the product's value is easiest to see on the visitor's own material, and when a result can be produced in seconds without an account. It takes real engineering work, so it fits products where a feature list does not sell on its own.

How to do it

1. Pick the single result that shows your value. Choose one that a visitor can get from one input: a URL, a domain, a pasted text block or a sample file.
2. Put the input on the page people see first, usually the homepage or the product page. A short box with one button beats a form with five fields.
3. Return something real, fast. A canned screenshot looks like a fake. Aim to show a result within a few seconds, with a progress message if it takes longer.
4. Give part of the answer free. Keep deeper results, exports and history behind sign-up. That gives people a reason to create an account after seeing their own result. See free side product for the same idea as a standalone tool.
5. Ask for the email after the result, not before. Offer to save or send the full report.
6. Protect the widget. Rate-limit requests, block internal addresses if you fetch URLs, and cap compute per visitor. Public tools attract bots.
7. Track how many people use it, how many sign up afterward, and how those users compare with people who skipped it. Do not assume users of the widget are comparable to everyone else, since they are more motivated to begin with.

Examples

• Ahrefs Backlink Checker. Ahrefs offers a free version of its Backlink Checker on its site. A visitor enters a domain or URL and clicks "Check backlinks." The free page shows metrics such as referring domains, number of backlinks, Domain Rating and Ahrefs Rank. Ahrefs describes it as "a glimpse into the power of our premium tool" and lists fuller reports (filters, broken links, anchor text, export) as part of the premium tool.

When it backfires

• Heavy cost per run. If each run uses paid compute, a popular widget gets expensive. Add limits before you launch it.
• Abuse. Open URL fetchers can be used to probe other sites or your own network. Restrict what the server will request.
• Disappointing results. A weak or wrong demo result on the visitor's own data hurts more than no demo. Test it on real examples first, including messy ones.
• Privacy. If visitors upload files or paste content, say what you store and for how long.
• Giving away the product. If the free result answers the whole question, nobody has a reason to sign up. Hold back depth, not accuracy.
• Hard to measure. People who use the widget are self-selected. Compare against a similar group before claiming it lifts sign-ups.

What it looks like in practice

Say you sell a tool that checks a company's website for broken links. The homepage has one box: "Enter your website." A visitor types their domain and, after a few seconds, sees "23 broken links found on 400 pages", with the first five listed. To see the other 18, download the report, or schedule a weekly re-check, they create a free account. You limit the tool to 500 pages per scan, one scan per visitor per hour. This is an illustration of one setup, not a result.

Sources

• Ahrefs: Free Backlink Checker, read Oct 8, 2026