A/B TESTING · WCONVERT PRO

Two ways to say it.
See what resonates.

Turn “I think this could work better” into a campaign you can compare. Create a variation, see how it performs and choose what to keep.

Explore a sample test How testing works

Visual campaign editing Even assignment You choose when to stop

THE SATURDAY LETTERReal designs · Local demo

A SMALL EDIT, A CLEAR QUESTION

Does a more specific headline
earn a different response?

THE HEADLINE IN THE REAL EDITOR Updated WConvert Headline panel, with the Content tab and the text One good idea. Every Saturday.

Real WConvert designs and editor. Local demo drafts with the same photo, form and button.

FROM AN IDEA TO A DECISION

A simple workflow.
A clearer comparison.

Start with a question you can answer. Keep the conditions comparable, then look at both the counts and the rates.

Create a variation.

Start with a copy of your campaign. Edit one meaningful element and preview the draft before publishing it.

See the visual builder

Share the opportunity.

Published versions receive an equal chance of assignment. Keep audience rules and timing alike for a more useful comparison.

Explore display rules

Read. Decide. Repeat.

Review each version over the same dates. Keep observing or choose the version you want to become the campaign.

Explore campaign analytics

MAKE YOUR FIRST TEST A USEFUL ONE

Change something
your visitor cares about.

Start with one clear difference. These are practical questions you can explore with the campaign editor.

The promise

What makes signing up worth it?

AJoin our newsletter

BOne useful idea, every Saturday

Test a clearer benefit in the headline. Keep the offer, form and audience consistent.

The next action

Does the button make the next step clear?

ASubscribe

BSend me the guide

Compare button wording while keeping the same resource and submission action.

The visual

Which image makes the offer feel relevant?

AA product detail

BThe product in use

Change the image while preserving the copy, fields and display rules.

Compare the same kind of result. Keep the measured action consistent across versions, such as a form submission in both.

READ THE RESULTS

See the difference.
Keep the context.

Compare both versions over the same dates. Read submissions alongside appearances, then decide whether to keep observing or make a choice.

INSIDE WCONVERT · A/B COMPARISONSep 15–28, 2026

Real A/B comparison report. Variant B has 165 email submissions from 1,180 appearances, a 14% rate. Original A has 120 from 1,200, a 10% rate. The interface states that results do not prove a winner.

READ BOTH THE RATE AND THE COUNTS

A stronger observed response.
Your next question still matters.

B recorded 165 submissions from 1,180 appearances; A recorded 120 from 1,200. Review the dates, audience and sample before deciding what to keep.

Original A email submission rate: 10%, calculated from 120 submissions divided by 1,200 appearances. — A · ORIGINAL

Variant B email submission rate: 14%, calculated from 165 submissions divided by 1,180 appearances. — B · VARIATION

Observed results from demo data. WConvert does not calculate statistical significance or choose a winner automatically.

Try the sample-size and decision example

ONE CAMPAIGN. ONE QUESTION.

Which headline earns the signup?

Sample results

Compare two headlines using sample results. Switch the sample size, then preview what happens when you keep a version.

Even assignment across two published versionsActual appearance totals can differ.

AThe originalIn the example

Good spaces. Better weekends.

Times shown: 40
Submissions: 4
Submission rate: 10%

BThe variationIn the example

One good idea. Every Saturday.

Times shown: 40
Submissions: 6
Submission rate: 15%

Look at the sample size.

OCT 1–2, 2026 · SAME DATES FOR BOTH VERSIONS

Two extra submissions can look like a big lead.

B shows 15% versus A’s 10%, but that is only 6 submissions versus 4. This early snapshot is a reason to keep observing, not declare a reliable winner.

Illustrative numbers, not customer results or a sample-size recommendation. Rates are submissions ÷ appearances. WConvert does not calculate statistical significance or choose a winner automatically.

Try either “Preview keeping” action to see what ending a test means. This demo collects no information and changes no real campaign.

YOU MAKE THE CALL

A higher rate is a clue.
The decision stays yours.

A few responses can swing a percentage. WConvert keeps the underlying counts visible and leaves the decision with you.

When you choose a version, it becomes the campaign. The other versions stop being served, while their recorded results remain available.

Understand the reporting

EXAMPLE · AFTER YOU CHOOSE B

The Saturday letterYour chosen version

Kept

The original designPast results stay available

Retired

Keep the design you choose.
Keep the learning, too.

Illustration of campaign history, not a plugin screenshot.

INCLUDED IN EVERY PRO LICENSE

Make room for
your next better idea.

A/B testing joins the visual builder, advanced targeting and campaign analytics in WConvert Pro.

Explore WConvert Pro

PRO · ONE WEBSITE$49/ year

USD · Annual license

Also available for 5 or unlimited sites

Same Pro features in every site-count plan.

BEFORE YOUR FIRST TEST

Know what
you’re comparing.

Is A/B testing included in Free?

A/B testing is included in WConvert Pro. Free includes campaign creation and local analytics, so you can start collecting results before upgrading.

How do I create a variation?

Create a variant from an existing campaign. It starts as an unpublished copy, ready to edit. Preview your change and publish the variation when you are ready to include it in the test.

Is the traffic split exactly 50/50?

Assignment is even across published versions, with no custom traffic-ratio setting. Two versions have an equal chance of assignment. Actual appearance counts can differ because of chance, page eligibility, timing and visitor behavior. An even assignment does not guarantee identical report totals.

Will a visitor keep seeing the same version?

WConvert remembers the assigned version in that browser’s local campaign state while it remains available. This is not person-level or cross-device identification. Another browser, cleared storage or a changed set of versions can lead to a different assignment.

Does WConvert choose a winner automatically?

No. You review the results and choose when to end the test. WConvert does not provide a statistical significance score, confidence interval or automatic winner promotion. A higher observed rate alone does not establish a reliable winner.

What happens when I choose a version?

The chosen version becomes the campaign and the other versions retire from delivery. Their historical results remain inspectable in reporting. If the chosen version is unpublished, choosing it does not bypass publication checks or publish its draft automatically.

Can I compare clicks against signups?

Versions in the same test must use a compatible conversion action. A link click and a submitted form are different outcomes, so they should not be treated as competing rates in one test. Keep the purpose and measured action consistent.

How long should I run a test?

There is no universal duration or appearance count that makes a result reliable. Plan around your traffic, expected response and decision before starting. Avoid repeatedly stopping as soon as a small lead appears. WConvert shows the observed counts and rates; it does not certify the sample as sufficient.

Does the report show only the current test round?

Reports cover the selected dates and can include earlier or later uses of a design. They are not isolated, immutable test-round snapshots. Choose your date range carefully and review the campaign history when interpreting the comparison.

START WITH ONE QUESTION

What would you
try differently?

A clearer headline. A more relevant image. A button that makes the next step obvious. Your next test can start small.

Find your WConvert plan