Klaviyo pop up best practices: the settings that decide whether a form works

A pop-up is a targeting decision before it's a design

Let's face it: a pop-up has two jobs that pull against each other. Interrupt somebody who's busy, and be worth the interruption. Design decides how that feels. Targeting decides who has to feel it at all.

So the first thing we change on a form that isn't working is almost never the artwork. It's the audience, the moment, and what turns up in the inbox afterwards.

Targeting rules decide more than the offer does

Klaviyo will happily show one form to everybody, on every page, for as long as you leave it alone. That default is why a customer who subscribed two years ago gets asked again on the checkout page.

• Exclude people who are already on the list, so subscribers aren't asked twice
• Exclude the checkout and the account pages, where an interruption costs you an order, not an address
• Split the desktop and the mobile form, because Klaviyo treats them as separate forms anyway
• Point one form at the collection your paid traffic lands on, with copy that matches the advert
• Set a reappearance rule that respects a dismissal for longer than one page view

None of that is design work and all of it moves the number. It's also the half of the job an agency can do in an afternoon, so we do it before anybody opens the editor.

When the form fires, and why exit intent is oversold

The trigger is one setting and it changes the outcome more than the headline does. Fire too early and you interrupt a page that hasn't finished loading. Fire too late and your visitor has already gone.

Exit intent is the one we argue about most. On the accounts we run, phones are the majority of sessions, and a form triggered on exit alone simply doesn't appear for them.

Build the mobile form separately, and expect it to behave differently

Klaviyo gives you a mobile version and most brands treat it as the desktop one, squashed. Wrong instinct. On a phone the form owns the whole screen, the keyboard eats half of it, and a close button nobody can find buys you a bounce instead of a subscriber.

So we build the two separately and judge them separately. The phone version tends to be shorter, with the offer in the first line and one field above the fold, and it usually earns its keep.

Every field costs you something

The number of fields is the plainest lever in the whole form. Each one loses you people who'd otherwise have subscribed, so a field has to earn its place by changing what you send.

• An email address, which is the only field that isn't optional
• A first name, worth having if a flow uses it and dead weight if it doesn't
• A birthday, which earns its place only when there's a birthday email to send
• A phone number, on its own tick box, because consent for a text isn't consent for an email
• A preference question, if the answer actually routes somebody somewhere different

A two-step form is the compromise we usually land on. The address goes in on step one and gets captured the moment it's submitted, so asking for a number on step two costs you nothing when it's refused.

Half of a pop-up's job happens after the form closes

The commonest broken thing we find isn't the form. It's the gap between what the form promised and what arrives: a code that doesn't exist, an email slow enough to lose the visitor, or a welcome that never mentions the offer at all.

So test the path, not the form. Submit a real address, take the code, try to spend it, and time how long it took. If the offer arrives late or arrives wrong, the fault is in the welcome flow, and the form is getting blamed for it.