Catch leaving visitors with a useful exit offer

Most visitors leave your pricing or product pages without doing anything. An exit-intent prompt gives you one last chance to help them. It works when the offer solves the reason they are leaving. A generic “Wait! Don’t go!” box just annoys people.

Why it works

A visitor about to leave has made a decision based on what they saw. If one specific thing stopped them, such as needing a partner’s approval or wanting a cost estimate, a small and relevant offer removes that obstacle at the moment it matters. Swapping something useful for an email address also keeps the conversation going with people who were not ready to buy today.

Because the prompt fires only when someone is leaving, visitors who are still reading are left alone.

How to do it

1. Pick one or two high-value pages. Pricing, a key feature page or a long guide are good candidates. Avoid showing prompts on every page.
2. Find out why people leave. Look at session recordings, sales call notes and short on-page polls. Common reasons are unclear pricing, not being ready to buy or needing to convince a boss.
3. Match the offer to the reason. For “not ready”, offer a guide or a template by email. For “need approval”, offer a one-page summary to share. For “unclear price”, offer a quick cost estimate.
4. Keep the prompt simple. One headline, one sentence, one field or button and an obvious way to close. Show it once per visitor and never to people who have just converted.
5. Run a proper A/B test. Split visitors between a version with the prompt and one without. Use a sample size calculator to decide how many visitors you need before you look at the results.
6. Measure the real goal. Count trials, demo requests or sales from each group as well as prompt signups. A prompt can collect emails while lowering the main conversion.
7. Follow up well. Deliver what you promised right away, then send a short, relevant sequence.

Example

A project management tool for agencies sees many visitors leave the pricing page. Call notes suggest buyers often need to convince a partner. The team tests an exit prompt offering a one-page comparison of plans to forward to colleagues. They run it against a control group until they reach the planned sample size, then compare trial starts across both groups. The prompt group shows more trial starts over the following weeks, and the forwarded PDF brings a few new visitors from the same companies.

Metrics to track

• Main conversion rate in the prompt group and the control, counting trials, demo requests or sales.
• Prompt view-to-submit rate, which shows whether the offer feels worth it.
• Conversion of exit-prompt leads through the follow-up sequence.
• How quickly visitors close the prompt. Very fast closes suggest it appears at the wrong moment.

When it backfires

• The offer is generic. A newsletter signup in a popup gives nobody a reason to stay.
• It blocks the page. Prompts that are hard to close or fire too early hurt trust and can count against you in search on mobile.
• You stop the test early. Calling a winner after a few days often picks up noise instead of a real effect.
• Discounts become expected. Regular visitors learn to hover near the close button to get a lower price.

Questions people ask

Does exit intent work on mobile?

There is no mouse leaving the window on a phone, so tools use other signals such as fast scrolling up, time on page or the back button. These are less precise. Test mobile separately and keep the prompt small so it does not cover the page.

Should the offer be a discount?

Only if price is the real reason people leave. For most B2B products a checklist, a template, a short demo video or a quick answer to a common objection works better and does not train buyers to wait for deals.

What should the prompt ask for?

Usually just an email address, so you can send the promised guide or summary. Each extra field lowers completion, so ask for more only if you need it to deliver the offer.