How to Make an A/B Test on Wisepops?

Start with a question, select a success metric and decide how the experiment will end before publishing variants. A scheduled rollout and a statistically conclusive result are different outcomes.

Source: Wisepops experiment guide, checked September 6, 2026. The planning worksheet below is editorial guidance, not a report of a client experiment.

Set up the experiment

For popups, bars and embeds, open the campaign’s A/B test control, duplicate a variant or select an existing campaign, and set traffic allocation. Select the success metric and manual or automatic conclusion. Finish editing before changing Draft to Published.

The guide lists CTR, order rate, revenue per visitor and the campaign’s associated goal. It describes a 95% confidence conclusion and an optional maximum end date that can roll out the best-performing variant without statistical significance. Confirm which ending rule you selected.

Control groups, detailed results and feed/web-push tests have plan qualifications in the documentation. Confirm your account’s access before building a test around them; see Wisepops pricing.

Write a test brief

For example, a shorter signup form may increase submissions while removing a qualification field the sales team needs. Decide how you will assess that trade-off before looking at the results.

Check measurement before collecting data

Verify the intended goal on a controlled visit and inspect both variants on mobile and desktop. Check that the offer, destination and form behavior match the test brief. Keep test submissions identifiable so they do not confuse later review.

Record launch time and any interruption, traffic change or tracking failure. If you must change a variant during collection, document it and reconsider whether the collected data still answers one coherent question.

Interpret rates with their denominators

An illustrative result of 30 submissions from 1,000 eligible visitors is 3%; 40 from 2,000 is 2%. The larger submission count does not mean a higher rate. These teaching numbers do not establish statistical significance or a real campaign outcome.

A variant comparison answers a different question from showing a campaign versus showing nothing. To assess incremental impact, use an appropriate control design and consistent measurement. Click-attributed revenue alone does not prove additional revenue caused by the campaign.

Record the conclusion accurately

Keep the selected metric, dates, audience, uncertainty and ending reason with the result. If a deadline triggered rollout before a conclusive result, describe it as an operational choice rather than a proven winner. Check the deployed campaign after concluding the experiment.

For feed or web-push experiments, follow the channel-specific controls in the official guide rather than assuming the popup workflow applies unchanged. Continue with goal tracking to verify the measurement setup.