Popups work, which is why they are everywhere, and they cost something, which is why most sites overuse them. The honest position is neither “never interrupt” nor “capture every visitor”: it is that an overlay is a loan against trust, and it needs to return more than it borrows. Here is how to tell the difference.
The test an overlay has to pass
Before adding one, answer three questions. Is the offer worth interrupting for, from the visitor’s point of view rather than yours? Is this the earliest moment it makes sense, or just the earliest moment it was technically possible? And is the exit obvious enough that someone who says no feels no friction at all?
A discount on the product someone is looking at passes. A newsletter signup thrown at a first time reader four seconds into an article does not, and the bounce it causes is real even when the capture rate looks flattering in isolation.
Timing and targeting do most of the work
The same popup can convert well or poison a site depending entirely on when it fires and who sees it.
Trigger on intent, not on a stopwatch. Scroll depth (the reader reached 60 percent, so they are engaged), exit intent on desktop, or an action that implies interest all beat a fixed delay. Never fire before someone has seen enough of the page to know what the site is, which realistically means never in the first ten to fifteen seconds.
Then respect the answer. Frequency capping is the single most neglected setting in this entire category: a dismissal should last weeks, not until the next page load, and anyone who already converted should never see the offer again. Suppress on the pages where the overlay competes with the actual goal, which means checkout, contact, and the pricing page you spent so long optimising.
The mobile rules are stricter, and Google enforces them
An interstitial that covers the content a visitor arrived for, on a phone, is a ranking liability as well as a usability failure. Google has treated intrusive interstitials as a negative signal for years, with narrow exceptions for legal notices like cookie or age consent.
In practice: on small screens use a banner or a bottom sheet that occupies a slice of the viewport rather than a full screen takeover, keep the close control large and in the expected corner, and never stack two overlays. And watch the layout shift, because an overlay injected late is one of the cheapest ways to wreck a Core Web Vitals score on an otherwise fast page.
Design the thing like it matters
Most popups look like they were made in a different studio to the site behind them, which is exactly why they read as spam.
Same type, same palette, same voice. One offer, one field, one button: every extra field is a toll, and a popup is the worst possible place to charge it. The copy should name the value in the visitor’s terms (“get the pricing guide”, not “join our list”), and the dismissal should be neutral rather than a guilt trip. Those “No thanks, I hate saving money” opt outs are a small, memorable act of rudeness attached permanently to your brand.
Accessibility is not optional either: focus moves into the dialog, escape closes it, focus returns where it came from, and the rest of the page is inert while it is open. That is the baseline for any modal, and it is the part most third party tools get wrong.
What usually works better
Before shipping an overlay, consider the versions that interrupt nobody.
Inline offers inside content convert well when they are relevant to the paragraph they sit in. A sticky bar takes a strip rather than the screen. A persistent, well written call to action at the end of a page catches people at the moment they finished reading, which is where a good page structure already puts them. And an offer that is genuinely useful rarely needs an overlay to be noticed.
We rarely ship popups on the sites we build, not out of principle but because the pages usually convert better without them, and it is easy to prove either way.
Measure the full cost
The trap is judging a popup on its own capture rate, which is always the flattering number. Judge it on what happens to the site: bounce rate on the pages it fires on, pages per session, and, most importantly, conversions on the primary goal rather than the email list.
If the newsletter grows while demo requests fall, the popup is not working, it is redirecting. That is a measurement question with a clear answer, and it is worth running as a real test rather than a debate.
Want an honest read on whether yours is helping? hello@beconfidency.agency.
If you would rather the whole site be built to convert without the tricks, that is exactly what our web design service is for.