Overview

Amazon has upgraded the Manage Your Experiments tool for brand‑registered sellers, adding the ability to A/B test both bullet points and product descriptions. The new “Run to Significance” option also lets the platform automatically stop a test once results reach a confidence level of 66 % or higher. Sellers who rely on optimized copy can now validate changes with real conversion data instead of guesswork.

Key Points

• Bullet‑point testing launched — Sellers can now compare different bullet‑point texts, orders, keyword placements, and formatting against the live version of a listing.
• Description‑section testing added — Full product descriptions can be split‑tested to identify which narrative or detail level drives more sales.
• Run to Significance introduced — Experiments automatically end when a variation achieves at least 66 % statistical confidence, removing the need to set arbitrary end dates.
• Fixed‑duration option retained — Sellers who prefer manual control can still specify exact start and end dates for any test.
• Brand Registry required — Only sellers enrolled in Amazon Brand Registry can access the expanded experiment capabilities.

How Manage Your Experiments Works

1. Select the content element — In Seller Central, choose whether to test bullet points, the product description, or another eligible field. Example: a seller targeting the “Eco‑Friendly Water Bottle” ASIN selects “Bullet‑point A/B test.”
2. Create variation A and B — Draft two distinct versions of the chosen element. Example: Version A leads with “Keeps drinks cold for 24 hrs,” while Version B starts with “Made from 100 % recycled plastic.”
3. Choose the duration setting — Pick “Run to Significance” to let Amazon monitor performance in real time, or set a fixed date range if preferred. Example: The seller selects “Run to Significance” so the test will stop automatically once one version reaches 66 % confidence.
4. — Activate the test; Amazon will split traffic evenly between the two variations and collect conversion data.