Make variant B in the same builder, show it to half your visitors and see which version more people finish — with a confidence level, not a hunch. On the Pro and Business plans.
| Version | Views | CompletedDone | ConversionRate |
|---|---|---|---|
| A Full form, 5 questions | 640 | 71 | 11.1% |
| B Short form, 3 questions | 652 | 98 | 15.0% |
B converts 35.5% better than A, with 96.4% confidence. Use B · Keep A
Example numbers, not a customer’s test.
Click Create variant B and switch between A and B above the questions. Change questions, pages, logic, endings or the theme of B — A stays as it is until you say otherwise.
No second form, no redirect script. The link, the embed code, notifications and integrations are the form’s own, so they work for both versions without setup.
Each new visitor gets A or B at random. Someone who leaves halfway and comes back finishes the version they started.
No verdict until each version has 100 views and 10 completed responses. A leader is named only at 95% confidence; otherwise it says there is no clear difference yet.
Stop the test, then Use B — its content is published under the same link — or Keep A. Every response stays, with the version it was given on.
A Variant column and filter in Results, in the CSV export, in webhooks and the API, and abVariant in embed events and GA4 — compare the versions in your own analytics too.
Create B, start the test, read the results and pick a winner — with screenshots.
→ 08What to test first, how many views each version needs before you decide, and why checking daily crowns false winners.
→Showing half of your visitors one version of a form and half another, then comparing how many finish each. In ClarNow the second version, variant B, lives inside the same form, so both share one link, one embed code and one table of responses.
Pro and Business. If your plan drops below Pro while a test runs, the test pauses and everyone sees A; you can still stop it, read the results and pick a winner.
ClarNow gives no verdict until each version has at least 100 views and 10 completed responses, and says roughly how many more views it needs. A small difference takes far more than that to tell apart from chance — the results line shows the confidence so you know when to stop.
Most forms get a few hundred visits a month. Split three or four ways, each version gets too few for a clear answer in a reasonable time, so ClarNow tests A against one B.
Everything that is content: questions, choices, pages, logic, the welcome screen, endings, variables, the theme and the settings saved with each publish. The form type, the link, access, notifications, integrations and tracking are shared.
Yes, and the builder warns you: changes you publish during a test mix into its results. For a clean comparison, stop the test, make the change and start a new run.
As of 16 November 2026: no more than two versions, no split other than 50/50, no automatic winner, and B is edited in the builder only — the API and the MCP server work with A.
Free forever for 100 responses a month — the whole builder, no card.
Cookie preferences
Required for the site to work — security and your session. No tracking.