Step 1: Open Your Dynamic Page

Go to Dynamic Pages from the left sidebar
Click the page where you want to run an experiment

Scroll to the Variants section.

Step 2: Click “Experiment” on a Variant

In the Variants table:

Locate the variant you want to test
Click Experiment

The Start Experiment modal will open.

Experiment Setup Parameters

Below is a breakdown of each setting inside the experiment modal.

1. Experiment Name

Give your experiment a clear name.

Example:
• Main Experiment
• Headline Test – Urgency
• Mobile CTA Optimization

This helps you track and compare experiments later.

2. Primary Metric (Required)

This defines what the experiment is optimizing for.

Available options:

• Conversion Rate
• Conversions
• Button Clicks
• Form Submissions

Choose the metric that best reflects your goal.

Example:
• Choose Conversion Rate to maximize efficiency
• Choose Conversions to maximize total volume

3. Secondary Metrics (Optional)

You can track additional metrics alongside your primary goal.

This helps you monitor trade-offs or additional behavior signals.

4. Experiment Variations

You must add at least one variation to test against the original.

The interface shows:

• Original (Control)
• Variation A
• Traffic allocation percentage

You can:

• Add multiple variations
• Adjust traffic split
• Describe what the variation is testing

Traffic distribution must equal 100%.

5. Completion Criteria (Required)

This determines when the experiment ends.

There are four available options:

A) Statistical Significance (Recommended)

The experiment ends when a winner is determined with 95% confidence.

Additional parameters:

• Minimum Visitors per Variant
• Minimum Duration (days)

Best for:
• Performance optimization
• Reliable decision-making

B) Fixed Duration

End the experiment after a specified number of days.

You define:
• Duration (in days)

Best for:
• Time-based campaigns
• Promotional tests

C) Visitor Count

End the experiment after reaching a target number of visitors.

You define:
• Target Visitors

Best for:
• Traffic-based testing
• Controlling sample size

D) Manual

The experiment runs until you manually stop it.

Best for:
• Open-ended testing
• Continuous optimization

Step 6: Start the Experiment

After configuring:

Review traffic distribution
Confirm completion criteria
Click Start Experiment

Superfunnel will automatically split traffic according to your configuration.

How Experiment Traffic Works

• Traffic is distributed according to your defined percentages
• The original variant acts as the control
• Variations compete against the control
• Performance is tracked based on the selected primary metric

Best Practices for Experiments

• Test one major change at a time
• Use Statistical Significance when possible
• Ensure enough traffic per variant
• Avoid stopping experiments too early
• Clearly document what each variation is testing