BigAtom Overlay enhances product images in Meta Catalog Ads using dynamic branded elements.
To measure the true impact of BigAtom Overlay, this must be run as a Meta A/B experiment — not by creating separate, unrelated campaigns. Both versions (with and without Overlay) should exist within the same campaign so that Meta fairly compares performance under identical conditions (audience, budget distribution, and learning).
BigAtom Overlay enhances product images using dynamic branded elements. This experiment helps you validate how those enhancements improve engagement and conversions.
Objective
Compare Overlay vs. No Overlay performance to measure improvements in CTR, CVR, ROAS, CPT, or any key KPIs you track.
Mandatory Pre-Requisites
To ensure results are accurate and not influenced by Meta’s automatic image edits:
Requirement | Why |
Disable Meta’s Media Enhancements (Image & Video enhancements) | Prevent Meta from applying overlays or edits where they shouldn’t |
|
|
Avoid frequent changes in product sets | Protect learning and maintain stable delivery |
If Media Enhancements are not disabled, Meta may choose overlay assets even in the “no overlay” group — which breaks the experiment.
Test Setup Overview
Component | Base (Without Overlay) | Test (With BigAtom Overlay) | Why It Matters |
Campaign | 1 Shared Catalog Sales Campaign | Same Campaign | Keeps conditions identical |
Ad Sets | Ad Set A | Ad Set B | Audience & budget split ensures fair comparison |
Audience | Same targeting across both | Same targeting across both | Removes targeting bias |
Creatives | Default catalog images | BigAtom overlay-enabled images | Isolates the impact of Overlay |
Budget | Even distribution between both ad sets | Even distribution between both ad sets | Prevents spending bias affecting results |
Optimization Event | Same for both ad sets | Same for both ad sets | Avoids optimization differences |
Test Duration & Learning
Recommendation | Why |
Minimum 7–14 days run time | Allows Meta to complete learning phase and stabilize results |
Avoid creative/product changes mid-experiment | Keeps test variables consistent |
Ensure both ad sets receive enough delivery | Prevents skewed distribution and inconclusive results |
Budget Requirement: Each ad set should have enough budget to get 50 conversions within 7 days based on your Account Average CPA | Meta recommends ~50 conversions/week per test group for learning
|
Want inspiration for what to test?
We’ve already curated a list of effective Overlay creative strategies.
👉 Refer to our Creative A/B Testing Playbook for Overlays