How the recommendations are worked out
The tool follows a few simple rules. They come from the advice in our guide to popup timing and triggers, not from measured benchmarks, so treat every number as a starting point.
- Time delay. On front and category pages, the delay is 55% of the average time on page, rounded to whole seconds and kept between 5 and 60 seconds. With an average of 45 seconds: 45 × 0.55 = 24.75, so 25 seconds. Most visitors have then seen what the page is about and are still on it.
- Scroll depth. On blog posts the starting point is 50% of the page, on product pages 40%. When the average visit is under 60 seconds, the point moves 10 percentage points earlier; when it is over 180 seconds, 10 points later.
- Exit intent. Added as a second trigger on desktop, so people who are about to leave also see the offer. If visits average under 10 seconds, exit intent becomes the main trigger, because few people would stay long enough for a timer.
- Surveys. Asked as people leave (exit intent), or late in the visit: 80% of the average time on page, within the same 5 to 60 seconds.
- Games. A gamified pop-up also gets a small teaser button, so people can open the game themselves.
- Checkout. No pop-up at all.
When you combine triggers, the first condition that is met shows the pop-up.
Mobile: follow Google's guidance
Exit intent needs a mouse, so it does not work on phones. Google also advises against pop-ups that cover the main content, especially right after someone arrives from search. Small bars, slide-ins and pop-ups people open themselves by tapping a teaser are lower risk. The tool therefore never suggests a full-screen pop-up on landing for mobile. Our guide to mobile popup best practices explains what Google says in more detail.
Frequency and excluded pages
A good pop-up shown too often becomes a bad one. The tool suggests once per visitor, at least 7 days before it can show again (14 days for discounts, leads and B2B or content sites), hidden for people who have signed up, and no more than one pop-up per visit. It also lists pages to leave out: checkout, payment, login and the pages where people are already filling in a form.
Test it on your own website
Your visitors, pages and offer are different from anyone else's. Run the suggested setup for a couple of weeks, then test one change against it: a shorter delay, a different scroll depth or a teaser on mobile. Change one thing at a time and compare periods of the same length. Read how in A/B testing campaigns.
For the offer itself, see our email popup ideas and how exit-intent popups work. In Scratcher, you set the triggers, pages, devices and frequency for each pop-up without code. See how on our pop-ups page.
