All features

A/B testing

Test copy, offers or whole designs against each other, and let the numbers show you the winner.

What it is

A/B testing, in plain words.

Most popup decisions are guesses: this headline or that one, 10% or free shipping, a bold design or a quiet one. An A/B test turns the guess into a measurement, on your own visitors.

A campaign can hold several variants. Switch two or more on, and traffic is split evenly between the active ones. Variants share the campaign's triggers and targeting, so the only difference visitors see is the popup itself.

Add a variant by duplicating one and tweaking it, by writing a new one in the designer, or by using a concept you didn't build yet. Not sure what to test? Ask the assistant: “Suggest an A/B test I could run”. It comes back with up to four ideas, each with a preview and a note on why it could work.

The campaign's Overview shows each variant's uplift and its chance to win, for the goal you choose, like submits or orders. When you're convinced, declare the winner: it stays on, and the others are switched off, not deleted.

When to use it

Where it pays off.

Copy that converts.

Test two headlines or two button texts and keep the one more visitors act on.

The right offer.

“Test 15% against the current 10%”: the assistant adds the variant. Remember to set up the real discount in your store too.

Design against design.

Put two finished concepts from the concept board in one campaign as A/B variants.

Judged on orders.

Set the campaign's goal to Order, so the winner is the variant that brings sales, not just sign-ups.

Key benefits

Why it earns its place.

01

Test more than a headline.

Copy, offer or a whole design.

02

Ideas on tap.

“Suggest an A/B test I could run” is one message away.

03

A clear read.

The Overview tab shows each variant's uplift and chance to win.

How it works

Set it up in three steps.

  1. 1
    Step 1

    Add a variant

    Under the variant table, click Add A/B test variant: duplicate a variant, write a new one, use an earlier concept, or add a control.

  2. 2
    Step 2

    Switch it on

    New variants start switched off. Turn them on, and traffic is split evenly across the active variants.

  3. 3
    Step 3

    Read the result

    Watch the uplift and the chance to win on the Overview tab, then declare the winner.

Settings & limits

Good to know.

Even split only.

Traffic is shared equally across the active variants. A custom split isn't available yet.

New variants start off.

A variant you add starts switched off, so it doesn't take traffic until you turn it on. Concepts added to a live campaign start right away.

You declare the winner.

The test doesn't end by itself. Declare the winner with the button, or ask the assistant to do it.

Uplift has a baseline.

With a control, variants are compared with no popup. Without one, they're compared with the weakest variant.

Questions, answered.

Step-by-step setup lives in the Help Center guide.

OptiMonk monk meditating in lotus pose

Try A/B testing
on your own store.

Free to start. No credit card.