A/B Testing Your Listings

Updated February 15, 2026

Set up split tests to find which title or tag variation performs better.

A/B Testing Your Listings

The A/B Testing Framework lets you run controlled experiments to compare listing variations.

Creating an Experiment

Go to Dashboard > Experiments.

Click New Experiment.

Select the listing you want to test.

Choose what to test: Title, Tags, or Description.

Enter the Control (your current version) and Variant (the new version).

Set a sample size target – the minimum number of views needed to declare a winner. The system calculates this automatically based on your desired confidence level (default 95%).

Click Start Experiment.

How It Works

Once an experiment is live, Etsy continues to serve your listing normally. Selletto Pro measures views, favourites, and sales for both versions by comparing performance before and after applying the variant.

Note: true A/B split-serving is not possible via the Etsy API. The experiment measures temporal lift — performance improvement after switching to the variant.

Reading Results

The results panel shows:

Conversion rate for control and variant

Relative uplift (%)

p-value and confidence interval from a chi-squared test

Status – Running, Significant, or Inconclusive

A result is declared Significant when p < 0.05 and the minimum sample size has been reached.

Applying the Winner

When an experiment reaches significance, click Apply Winner to permanently update your Etsy listing with the winning variant.

Was this article helpful?