Run an A/B Test Using Custom Parameters

  • 2 minute(s) read
Prev Next

Goal

Compare A/B test variants by passing variant information as custom parameters and analyzing results in Mapp Intelligence.

Prerequisites

  • Mapp Intelligence tracking is implemented

  • A/B test variants are assigned (for example, via custom logic or an A/B testing tool)

Procedure

Step 1 – Create Custom Parameters

Create custom parameters to store A/B test information.

Depending on your setup, different parameter types can be used (for example, page parameters, event parameters, or e-commerce parameters).

In this example, page parameters are used.

Navigate to:
Mapp Q3 > Custom Parameters > Page Parameter

Create at least two parameters:

  • One parameter for the test name (for example, ABTest)

  • One parameter for the variant (for example, ABTestVariant)

When defining parameters:

  • Choose a free parameter ID

  • Use clear and consistent naming

  • Activate the parameter

For detailed instructions on how to create and configure parameters, see: How to Set Up Custom Parameters

Step 2 – Pass Variant Information to Tracking

Include the test and variant information in your tracking.

For each relevant request, pass:

  • The test name

  • The assigned variant

Example:

  • ABTest = HomepageLayoutTest

  • ABTestVariant = layout_a

Ensure that:

  • The same values are used consistently

  • Each user remains in the same variant during the session

The variant must be available in tracking to allow analysis, but how broadly it is included (for example, only on specific pages or across multiple interactions) depends on the test setup.

Step 3 – Validate Tracking

Verify that custom parameters are correctly collected.

Check that:

  • Parameters are present in tracking requests

  • Values are consistent across the session

  • Variant distribution matches expectations

For detailed validation steps, see: Review of Data Collection

Step 4 – Analyze Variants in Mapp Intelligence

Use custom parameters to analyze and compare A/B test variants.

Analyze test and variant together

To analyze A/B tests, combine the test and variant parameters in a single analysis.

  • Navigate to: Navigation > Page Parameter (or the corresponding parameter type used in your setup)

  • Select the parameter used for the test (for example, ABTest)

  • Add the variant parameter (for example, ABTestVariant) as an additional dimension

This creates a pivot table that shows which variants belong to which test and how often they were tracked.

Use a suitable metric depending on your setup, for example:

  • Page Impressions to analyze how often a variant was tracked

  • Visits to analyze how many sessions included a variant

The choice of metric depends on how the variant is passed in tracking.

Performance comparison

Add business metrics to the same analysis to compare performance across test variants, for example:

  • Conversion Rate

  • Revenue

  • Orders

This allows you to evaluate how each variant of a specific test performs.

Note that this approach does not represent a strict Attribution Model. Results depend on how variant information is available within the tracked data.

Note

The interpretation of metrics such as conversion rate or revenue depends on how variant information is included in tracking (for example, within a session or across multiple interactions).

Segmented analysis

To analyze variants in more detail:

  • Filter analyses by variant values (for example, Variant A or Variant B)

  • Optionally filter by test name to focus on a specific A/B test

  • Apply these filters to different analyses, such as:    

    • Product analysis

    • Device analysis

    • Campaign analysis

In Mapp Intelligence, segments are reusable filters. You can either apply filters directly or create segments for repeated use.

Next Steps

  • Refine variant definitions based on results

  • Extend analysis using additional dimensions and segments

  • Explore more advanced tracking approaches if deeper attribution is required