Skip to main content

Setting Up Gate Experiments (A/B Testing)

Create and analyze A/B tests for your Formulayt gates to improve form conversion rates

Updated over 9 months ago

Why gate experiments matter

Form performance is a vital factor for web conversion optimization that's often overlooked. While marketers put significant effort into ads, landing pages, and follow-up emails, the form experience itself can make or break conversion rates.

Formulayt's gate experiments (A/B testing) enable you to test different form variations while collecting performance data. This allows you to:

  • Test specific hypotheses about user behavior

  • Identify which form elements increase conversions

  • Make data-driven decisions to improve results

Step-by-step guide to creating gate experiments

Follow these steps to create your first gate experiment:

  1. Navigate to Gates in the left-hand menu

  2. Click Create New > Experiment in the top right corner

  3. Enter a clear name and description for your experiment

    • Your description should indicate what specifically you're testing

  4. Add gates to test:

    • Include a variant description for each gate to track differences

    • Designate one variant as the default (shown when experiment is paused)

    • Set the weight for each variant to control display frequency

    • For experiments with more than two variants, enter the lowest weighting first

  5. Use the actions area to:

    • Edit variants

    • Edit gates

    • Preview gates

    • Remove gates from the experiment

  6. Update the experiment to receive embed codes for your website e.g. #exp-gate-3d173b95-da4b-4f43-bbe5-575491ec984d

  7. Click "Start Experiment" in the top right when ready to begin collecting data

    • Only start after completing all configuration

  8. After sufficient time has passed, view Experiment Analytics for behavior insights


Important notes

Editing Limitations: Active experiments cannot be edited. To make changes, you must reset the experiment, which will also reset all analytics data.

Did this answer your question?