Popup Timing, Triggers and Frequency Caps

Never fire on arrival. Trigger on behaviour that signals the visitor is settled and interested - time genuinely spent, depth scrolled, pages viewed, a return visit, or a specific intent action - and combine the trigger with a minimum delay so no fast action ambushes anyone. Cap frequency hard: once per visitor per campaign across a long window, with permanent suppression after a signup, a purchase or a clear decline, and a site-wide ceiling so separate campaigns cannot stack into a gauntlet. These settings are the difference between a popup programme and a popup habit, and they matter more to results than any headline or colour the popup carries.

A trigger is a hypothesis about openness, so choose ones with a meaning you can state. Scroll depth on a long guide means the reader is engaged with the content an embedded offer extends. A second product view means comparison has begun and help or a saving may be welcome. Time on site is the weakest common trigger - it counts the tab left open during lunch - so never use time alone, and always pair it with activity. Exit intent is its own case, covered in the exit-intent guide. Write each campaign's hypothesis in its log entry: we show this when a visitor does X, because X suggests Y. A trigger nobody can justify in that sentence is noise dressed as targeting.

Some contexts revoke openness whatever the trigger says: checkout and payment, where interruption risks the transaction; the middle of a form, where it breaks a task in progress; post-purchase pages, where the visitor has just done the thing; and any visit where another campaign has already fired. Maintain a site-wide suppression list and a one-interruption-per-visit rule at the programme level, not per campaign, because the visitor experiences the site, not your campaign structure. The catalogue of what stacked interruptions teach is in popup mistakes that teach avoidance; most of it is preventable at this settings level, for free.

Starting defaults, to be tightened by your own data: a given campaign shows at most once per visitor in 30 days; a declined campaign does not return for at least 90 days, and returns, if at all, only with a genuinely different offer; any visitor who converts is suppressed from that campaign permanently and from sibling capture campaigns while their subscription or order is fresh. Illustration of the alternative: an uncapped timed popup on a weekly-read site can meet the same loyal visitor fifty times a year - fifty lessons in closing your boxes fast, delivered to your best audience. Caps are cheap insurance bought with settings, and every popup tool of any seriousness supports them.

A visitor the site already knows - subscribed, a past buyer, a regular reader - has answered the capture question, and asking again announces that the business does not remember them. Suppress capture campaigns for subscribers entirely, and for buyers show only campaigns that respect the relationship: service and replenishment messages, delivered through the gentler formats in slide-ins, banners and embedded forms. Where the tooling cannot recognise returning visitors reliably, bias the triggers later and the caps tighter for everyone; being remembered imperfectly should cost the business restraint, not cost the visitor repetition.

Timing and trigger changes are real changes with real effects, so the testing discipline of the foundations bench applies: alter one setting, note the date and the hypothesis, and read exposed visitors' conversion and downstream behaviour over whole weeks, as the measurement guide sets out. Resist the common escalation - when a popup underperforms, firing it earlier and oftener - which reliably converts a weak campaign into an actively harmful one. The settings ladder runs the other way: later, rarer, more relevant is the direction of nearly every durable improvement on this bench.

Campaign settings rot quietly: a seasonal popup's end date passes unnoticed, a new tool resets caps, a copied campaign inherits arrival timing nobody chose. Put every live campaign's trigger, delay, cap and suppression list into a one-page register, reviewed with the parent bench, onsite campaigns, each quarter. The register is also where new campaigns earn their slot: a proposed campaign that cannot state its trigger hypothesis and its cap does not launch. Five lines of paperwork per campaign is the entire bureaucracy of a restrained popup programme, and it is what keeps restraint true after the person who promised it has forgotten.