Test plan by ASIN
Eligible ASINs ranked by traffic and revenue, so tests run where a lift in conversion is worth the most and where there is enough traffic to reach a result.
Amazon · Listings · A/B testingAmazon A/B testing
Amazon A/B testing
Manage Your Experiments, run properly
CommerceKind is an Amazon agency that runs A/B tests for brands with Amazon’s Manage Your Experiments tool. Our in-house team picks which listings and which elements to test, writes the hypothesis, builds the alternative title, main image, bullets or A+ Content, sets up and monitors the experiment in Seller Central, reads the result on conversion and units sold, and rolls the winner out to similar products. Decisions come from shopper behaviour, not opinion.
Book a strategy call ★ Trustpilot 4.4/5 · $100M+ revenue for client brands · 60+ brands managedAt a glance
- Who it’s for
Brand Registry owners whose listings get traffic but convert below what they should, and brands that want proof before changing a best seller. - How we work
Test plan by ASIN first, then creative, set-up, monitoring and roll-out of winners. - Proof
$100M+ client revenue · 60+ brands managed in-house · $3M+ monthly ad spend
What our Amazon A/B testing service covers
Main image tests
Alternative main images planned and produced within Amazon’s image rules, such as angle, scale, pack shot or how the product fills the frame.
Title and bullet tests
Titles and bullet points rewritten around one clear idea per test, such as leading with the main benefit or the key search term.
A+ Content and description tests
A+ Content modules, comparison charts and descriptions tested against your current version.
Set-up and monitoring
Experiments created and watched in Manage Your Experiments, with no changes made to the listing while a test is running.
Reading results and roll-out
Results read on conversion, units and sales per visitor, and winning ideas applied to similar products across your catalog.
What we test first, and why
| Element | What it changes | When we test it |
|---|---|---|
| Main image | Click-through from search results | When click-through is low against similar products |
| Product title | Click-through and relevance | When the title buries the main benefit or key term |
| Bullet points | Conversion on the detail page | When shoppers visit but do not buy |
| A+ Content | Conversion and fewer returns | When questions and reviews show buyers are unsure |
| Description | Conversion where A+ is not used | On ASINs without A+ Content |
How an Amazon A/B testing programme works
01
Access
You grant Seller Central permissions through Amazon’s user-permission tools. No passwords, and no listing changes without your approval.
02
Plan
We check eligibility, rank ASINs by traffic and revenue and agree the first tests and what each one is meant to prove.
03
Build and run
We produce the alternative image or copy, set up the experiment and leave it alone until it has run its course.
04
Read and roll out
We report the result, apply the winner and carry the lesson to similar listings.
$100M+
revenue generated for client brands
60+
brands managed in-house
$3M+
monthly ad spend managed
4.4/5
on Trustpilot (13 reviews)
What Manage Your Experiments is
Manage Your Experiments is Amazon’s own A/B testing tool for brand owners enrolled in Amazon Brand Registry. It splits real shoppers between two versions of a listing element, such as the main image, title, bullet points, description or A+ Content, and reports which version sold more. Only ASINs with enough recent traffic are eligible, and Amazon shows which of your ASINs qualify inside Seller Central. Because the test uses live shoppers on Amazon itself, it measures what people actually buy, not what a survey panel says they like.
One idea per test
A test that changes five things at once can tell you that version B won, but not why. We write one clear hypothesis per test, such as “showing the product in use will lift click-through” or “leading the title with the main benefit will lift conversion”, so every result teaches something you can use on other products. We also check that seasonality, deals or stock-outs do not distort the result.
Let tests run to a result
Experiments need time and traffic. Amazon lets brands choose a duration of several weeks, and stopping early or editing the listing during a test can make the result meaningless. We plan tests around your calendar, so a test does not run through a big deal event, and we only call a winner when the data supports it.
Testing is part of listing work, not a substitute for it
A/B testing works best on listings that are already sound: correct keywords, compliant images and clear copy. Where a listing has deeper problems, we fix them first through our Amazon listing optimization and A+ Content services, and our main image work. In one account, fixing the listing before scaling ads turned a declining product around: see the listing-first case study.
Testing and advertising together
A better-converting listing makes every ad click worth more. We manage $3M+ in monthly ad spend across our clients, and we read test results alongside Amazon PPC data, so a lift in conversion shows up as lower cost per order. Results depend on product, traffic and category.
Why brands choose CommerceKind for Amazon A/B testing
CommerceKind has helped generate $100M+ in revenue for client brands, manages 60+ brands in-house and $3M+ in monthly ad spend, is an Official TikTok Shop Partner with a 10K+ creator network, and is rated 4.4/5 on Trustpilot from 13 reviews.
Brands that work with us
A few of the 60+ brands CommerceKind works with.
Verified Client Case Studies · Real Dashboards
See the numbers behind our client turnarounds.
Every case study below comes directly from unedited Seller Central, Amazon Ads, Search Query Performance, Sellerboard, TikTok Shop, and Walmart Connect dashboards.
+102% Ad Sales · −53% ACOSPrivate Label · Full Account Turnaround
Listing-first PPC turnaround: sales doubled, ACOS cut in half
Redesigned the main image for click-through, rewrote listing copy and A+ Content to convert, then rebuilt Sponsored Products and daily bid management.
Ad Sales$14,304 → $28,911+102%
Orders798 → 1,646+106%
ROAS3.99 → 8.48+112%
ACOS25.08% → 11.79%−53%
+117% Ad Sales · −29% ACOSPrivate Label · PPC Scale & Efficiency
Scaling PPC without ACOS spikes: 117% more revenue at lower ACOS
Rebuilt campaign hierarchy and segmentation with disciplined bid rules and A+ Content so 54% more ad spend produced 117% more ad revenue.
Ad Sales$19,680 → $42,712+117%
ACOS34.03% → 24.10%−29%
ROAS2.94 → 4.15+41%
Ad Spend$6,697 → $10,294+54%
+126% Ad Sales · −40% ACOSPrivate Label · 30-Day Turnaround
One-month PPC turnaround: sales up 126%, ACOS nearly halved
Tightened listing copy and imagery in week one, cut wasted spend across unprofitable targets, and set SKU-level ACOS goals with daily bid adjustments.
Ad Sales$15,170 → $34,257+126%
Orders553 → 1,233+123%
ROAS3.62 → 6.01+66%
ACOS27.64% → 16.65%−40%
+144% Ad Sales · +141% OrdersHealth & Wellness · Rapid Volume Scale
Health & wellness brand: 2.4× sales volume while holding ROAS steady
Mapped untapped high-intent keyword pockets and expanded placement coverage while upgrading creative so sales jumped from $38K to $93K at stable ROAS.
Ad Sales$38,207 → $93,266+144%
Orders1,545 → 3,719+141%
ROAS2.55 → 2.60Held
ACOS39.19% → 38.48%Held
+118% Click Share · +106% ClicksHealth & Wellness · SQP & CTR Growth
Keyword click share more than doubled in a declining search market
Diagnosed Search Query Performance funnel drop-off and redesigned the main image around the core buyer need—doubling click share even as market volume dipped.
Keyword Clicks34 → 70+106%
Click Share0.85% → 1.85%+118%
Impressions4,353 → 6,635+52%
Imp. Share2.53% → 4.08%+61%
~$97K → ~$160K MonthlyMulti-Category · Full Account Management
Multi-product store: broke a multi-quarter plateau to hit a record month
Sequenced listing conversion upgrades, A+ Content across hero SKUs, and structured PPC review cycles—compounding from ~$97K to ~$160K in monthly revenue.
Monthly Rev.~$97K → ~$160K+65%
Perf. Index~$128K → ~$190K+49%
Timeline7 monthsJun → Jan
OutcomeAll-Time HighRecord
+906% Purchases · 9.62% ShareBaby Supplements · Keyword Dominance
Baby supplement brand: 10× keyword purchases on #1 category term
Combined SQP search-funnel analysis, buyer-language listing SEO, main image CTR work, and exact-match ranking campaigns to capture 9.62% purchase share.
Weekly Orders16 → 161+906%
Purchase Share1.17% → 9.62%+723%
Keyword Clicks95 → 516+443%
Click Share2.71% → 6.81%+151%
+$87K Net Profit · −30.6% Ad CostLarge Catalog · Profitability Turnaround
Large-scale catalog seller: +$87K net profit turnaround in one quarter
Audited SKU-level contribution margins in Sellerboard, eliminated $34K in low-margin ad waste, and reversed YoY profit trend from −31.4% to +12.5%.
Net Profit$698K → $785K+$87K
YoY Trend−31.4% → +12.5%+43.9pp
Ad Spend$111K → $77K−30.6%
Ad Savings$34,024 saved1 Qtr
£256 → £297,128 GMV · 11,827 OrdersTikTok Shop · Creator-Led Growth
TikTok Shop (GBP): from £256 to £297,128 GMV driven by affiliate creators
Built an affiliate recruitment, seeding, Discord community, and brief engine from scratch—generating £197,522 in creator-attributed GMV and 11,827 orders.
Shop GMV£256 → £297,128Window
Orders11,827In Window
Creator GMV£243 → £197,522Creators
Hero SKU£71 → £145,065Top Item
+51.9% Shop GMV · +67% OrdersTikTok Shop US · GMV & Creator Scale
TikTok Shop (US): +51.9% GMV growth in two months by widening creator base
Broke through creator concentration plateau by expanding publishing affiliates from 503 to 556 and scaling GMV Max ads behind top-converting videos.
Shop GMV$77.3K → $117.4K+51.9%
OrdersFeb → Apr+67.0%
Creator GMV$48.7K → $70.3K+44.3%
Active Sellers120 → 133+13
$1.27K → $21.6K Sales · 2.72 ROASWalmart Marketplace · Channel Expansion
Walmart Marketplace: from $1,271 to $21,646 in monthly sales & Pro Seller
Rebuilt Walmart listing quality scores from Poor to Good, cleaned up catalog structure, and scaled Walmart Connect Sponsored Search to 2.72× ROAS.
Monthly GMV$1.27K → $16,855Oct → Mar
Apr Sales$13.3K → $21,646870 Units
Ad ROAS$1.80 → $2.72+51%
Listing Score57% → 67%Pro Seller
Original Client Reviews · Unedited
Every written review from our Trustpilot page.
Read all 12 original written client reviews verbatim—linked directly to Trustpilot so you can verify every single one.
Frequently asked questions
Is there an agency that runs Amazon A/B tests?
Yes. CommerceKind runs Amazon A/B tests for brands using Manage Your Experiments: a test plan by ASIN, alternative main images, titles, bullets and A+ Content, set-up and monitoring in Seller Central, results read on conversion and units, and winners rolled out across the catalog.
What is Amazon Manage Your Experiments?
It is Amazon’s built-in A/B testing tool for Brand Registry owners. It shows two versions of a listing element to different shoppers and reports which version sold more, so you can keep the better one.
What can you A/B test on Amazon?
With Manage Your Experiments, brand owners can test product titles, main images, bullet points, product descriptions and A+ Content on eligible ASINs. Price is not tested with this tool.
Do I need Brand Registry to A/B test on Amazon?
Yes. Manage Your Experiments is available to brand owners enrolled in Amazon Brand Registry, on ASINs that have enough recent traffic to be eligible.
How long should an Amazon A/B test run?
Long enough to reach a clear result. Amazon lets you choose a duration of several weeks, and we usually avoid ending a test early or running it through a major deal event, which can distort the result.
Can I edit my listing while a test is running?
You should avoid it. Changing the listing, price or images during an experiment can make the result unreliable, so we plan any other changes around the test.
What if my ASINs are not eligible?
We focus first on listing quality, traffic and advertising for those products, and use what we learn from tests on eligible ASINs to improve similar listings.
Do you also create the images and copy for tests?
Yes. Our in-house team writes the alternative titles, bullets and A+ Content and plans and edits the main images, within Amazon’s style and image rules.
Do you run A/B tests outside the US?
Yes. We work across Amazon marketplaces, including the US, the UK, Canada and Europe, wherever Manage Your Experiments is available for your brand.
Can my agency white-label Amazon A/B testing?
Yes. Agencies can offer listing testing and optimization under their own brand through our white-label Amazon listing and account management services, with an NDA and non-solicit first.
Not sure which
version sells more?
Book a free strategy call. We’ll look at your listings with you and show which tests to run first.
Book a strategy call