FERMÀT Experiments allow you to test different funnel variations against a control to understand what drives better conversion, revenue, or engagement. You can create multiple variants, control traffic allocation, and analyze results directly in the platform.
What’s in this guide:
What Is a FERMÀT Experiment
A FERMÀT Experiment lets you test strategic hypotheses by comparing a control funnel against one or more variants.
Common use cases include:
Testing different landing page layouts
Trying new messaging, imagery, or CTAs
Experimenting with pricing, bundles, or offers
Measuring performance vs a baseline (BAU)
Experiments are designed for larger, intentional changes, not just small optimizations.
Requirements Before You Start
Before creating an experiment, make sure:
You have a live funnel published
The funnel you choose will act as the control
You know what you are testing and why
If you haven’t built a funnel yet, start there first. Know more here.
Creating a New Experiment
To create a new experiment:
Open the funnel you want to use as the control
Click the Experiments tab in the top navigation
Click Create Experiment
This control funnel will remain unchanged and act as your baseline.
Naming and Describing Your Experiment
After clicking Create Experiment:
Enter an Experiment Name
Add a Description explaining:
What you are testing
Why you are testing it
What success looks like
Clear names and descriptions help differentiate experiments, especially when multiple are active.
Adding & Managing Variants
To add variants:
Click Add a Variant
Choose how to create the variant:
Clone the control funnel
Clone another existing variant
Repeat to add as many variants as needed
You can add multiple variants, but keep in mind:
More variants require more traffic
Experiments may take longer to reach significance
Editing the Control or Variants
To edit a funnel within the experiment:
Use the dropdown above the preview, or
Edit the variant as per the testing requirements
Each variant can be customized independently while remaining part of the same experiment.
Launching an Experiment
When everything is ready:
Review your control and variants
Confirm traffic allocation
Click Launch Experiment in the top right
Once launched, traffic will begin flowing to the variants based on the configured weights.
What Happens After Launch
After launch:
Your experiment runs continuously until stopped
Performance data is tracked automatically
Results can be reviewed in the Experiments Dashboard
Use the results to decide whether to:
Keep the control
Promote a winning variant
Run follow-up experiments