Available on | ❌ Free | ❌ Starter | ❌ Pro | ✅ Business |
What is A/B testing?
With A/B testing (also known as split testing), you show two versions of the same funnel to different visitors and compare how they perform.
The goal is simple: find out which version gets a higher completion rate (or more leads, submissions, etc.) so you can keep the winner and improve your results over time.
Typical things you might test:
Headlines and descriptions
Images or videos
Buttons (text, color, position)
Form length or question order
Any other change that might affect conversions
Note: Check out our plans and pricing for more information on A/B testing and other paid features!
Before you start: prepare your funnels
Pick the funnel you want to improve. This will be your control (original version).
Duplicate it in your dashboard. This duplicate will be your variation (modified version).
In the duplicate, make the changes you want to test (e.g. new headline, different image, shorter form).
Publish both funnels once you’re happy with them.
Tip: Duplicating and editing the copy saves time and ensures both versions are identical except for what you’re testing.
You can still edit both funnels even after publishing if you need to make corrections.
How to set up an A/B test
1. Create a new test
In your dashboard, click the “A/B Tests” tab in the top navigation.
Click “+ Create new A/B test”.
Give your test a descriptive name, e.g. “Pricing page - short vs long form”.
This is especially helpful if you plan to run multiple tests in parallel.
2. Configure test settings
On the test settings page you can adjust:
Name – change the test name at any time.
Domain – choose whether the test runs on a generic or custom domain connected to your account.
Useful if you want to hide the
involve.medomain from end users.
URL slug – customize the URL path for your test.
You’ll also see the option:
“Show the same funnel upon refresh during the same session”
If you enable this:
Visitors will keep seeing the same funnel variant they were first shown, even if they:
Refresh the page, or
Open the same URL again in a new tab
ℹ️ Bear in mind that this restriction is browser based and it is made possible via the use of cookies, which have a duration of 30 minutes.
For accurate test results, we recommend keeping this enabled so people don’t see different versions in the same session.
3. Select funnels A and B
Once your settings are saved:
Make sure both funnels you prepared earlier are published.
Choose them from the Funnel A and Funnel B dropdowns.
Click “➡️ Start A/B Test” to activate the test.
From now on, visitors who use the test URL (see next section) will be randomly shown either Funnel A or Funnel B.
How to share your A/B test (important!)
To actually use your A/B test, you must share the A/B test URL or embed code, not the original funnel URLs.
If you share:
The URL or embed code from the funnel’s own “Share & Embed” page →
Only that single funnel will be shown (no split testing).
Instead:
Open your A/B test and click “Share your A/B Test”.
Use the URL shown there if you just want to share a link.
Or use the embed code from this page if you want to embed the test on your website.
The “Share your A/B Test” page looks similar to the regular “Share & Embed” page, but it’s specific to the test and ensures your traffic is split between Funnel A and Funnel B as configured. For more details on embedding, see our dedicated embedding article.
How to read your A/B test results
Once your test is running, you’ll see performance metrics for each funnel:
Visits – how many times each funnel was loaded.
Starts – how many visitors actually started interacting with the funnel.
Leads – how many leads were generated.
Submissions – how many completed submissions you received.
Completion rate – the percentage of visitors who completed the funnel.
Avg. duration – the average time users spent on the funnel.
You can also click a funnel name from within the test to go straight to that funnel’s Analytics page for more detailed breakdowns.
The more data you have, the more reliable your conclusion will be. Avoid calling a winner too early based on very low traffic.
When you’re confident:
Decide which funnel is the winner based on your main goal (e.g. higher completion rate, better lead quality, or shorter completion time).
Keep the winner live and either:
Stop the A/B test, or
Set up a new test to continue iterating.
Best practices
Test one main change at a time
If you change too many things in one go, it’s hard to know what caused the improvement.Let the test run
Give it enough time and traffic before deciding on a winner.Use clear names
Name your tests and funnels so you can easily understand what was tested (e.g. “Hero CTA – ‘Start now’ vs ‘Get a quote’”).Keep everything else consistent
Target the same audience and channels while the test is running where possible.
Related resources
Testing Premium Features - how to try paid features if they’re not included in your current plan.
Analytics & funnel reports - learn more about how to interpret analytics for your funnels.