Skip to main content

Adaptive Testing

Adaptive Testing (A/B Testing) enables you to send multiple message variations to a small portion of your audience, automatically identify the best-performing version, and deliver the winning message to the remaining recipients to maximize engagement.

K
Written by Ken Cantu

What is Adaptive Testing?

Adaptive Testing allows you to send multiple variations of a message to different segments of your audience, measure performance, and then send the winning variation to the remaining subscribers. This helps you optimize message content and overall campaign effectiveness.

Key Benefits

  • Data-driven optimization: Make decisions based on actual performance metrics rather than assumptions

  • Improved engagement: Discover which message variations resonate best with your audience

  • Easy setup: Intuitive interface that doesn't require training or external expertise

  • Flexible testing: Test different content variations to understand what drives results


Creating an Adaptive Test

Getting Started

  1. Enable Adaptive Testing: When composing a new message, toggle the "Adaptive Test" option to ON

  2. Create Your Variants: Build different message variations to test

Setting Up Your Content Test

Create Your Variants: You can create up to 4 different message variations

  • Each variant can have unique subject lines, message content, and images

  • Use the duplicate variant button to copy content and make minor adjustments

  • Use "Copy from Variant A" to quickly replicate content across variants

Configure Test Settings:

  • Test Size: Choose what percentage of your audience receives the test (the rest receives the winner)

  • Audience Split: Decide how to divide your test audience across variants (e.g., 25% each for 4 variants)

  • Winning Message: Choose whether the winner sends automatically or creates a draft for review

Winner Determination:

  • You can select when the winners are determined between 1 and 48 hours after the variants are sent. The winning variant is based on unique click-through rate (CTR)

  • The system automatically identifies the best-performing variant


Tracking and Analyzing Results

Performance Dashboard

Access comprehensive test results through the Performance Dashboard:

Test Results View: See all variants side-by-side with key metrics:

  • Recipient count per variant

  • Click-through rate (CTR)

  • Unique clicks

  • Donation metrics (when donation tracking is enabled)

Winner Highlight: The winning variant is clearly marked based on CTR performance

Detailed Insights: Review adaptive test insights and message goals

Individual Variant View: Click any variant to drill down into detailed performance metrics

Message List

The Adaptive Tests list view shows:

  • Test status at a glance

  • "Send Winner" action for tests waiting for you to schedule the winning message

  • "Winner variation scheduled" status for completed tests


Best Practices

Content Testing Tips

  1. Test One Element: Focus on testing one variable at a time (e.g., subject line OR call-to-action)

  2. Include Links: Add tracking links to measure click-through performance accurately

  3. Meaningful Differences: Make variations different enough to produce measurable results

  4. Donation Tracking: Use unique URL parameters for each variant (e.g., utm_campaign=appeal_A vs appeal_B) so your fundraising platform can differentiate donations by variant in its reporting. Tatango displays aggregate donation totals across all variants.

General Guidelines

  • Test Size: Use our recommended audience test size

  • Allow Sufficient Time: The 1-hour determination window ensures enough data for accurate results

  • Review Results: Check the Performance Dashboard to understand why the winner performed best

  • Apply Learnings: Use insights from tests to inform future campaigns


Troubleshooting

Common Issues

"Can't schedule message in past" validation error

  • Ensure all scheduled times are in the future

  • Double-check your timezone settings

  • Verify the winner message time is set correctly

INSERT LINK prompts not appearing

  • These appear automatically when variant content is blank

  • If you see [INSERT LINK HERE] in red text, you're viewing a cloned test; use "Copy from Variant A" instead


Last updated: April 2026

Did this answer your question?