Main Image A/B Testing Masterclass
Discover a Simple Trick to Boost Sales by 15% Virtually Overnight
How to Make Alternate Main Images
Before You Start A/B Testing, You Will Need to Make Some Alternate Main Images.
First you will need to create some alternate main images. If you cannot do this yourself, get a creative graphic designer to do it.
Think about how to improve your main image so it catches more attention, and/or key information is easier to understand at a glance.
Example to follow:
Main Image A) Your current main image.
Main Image B) Just make the product bigger. Simply make sure bottle fills up the entire main image.
Main Image C) Show the capsules/gummies alongside the bottle.
Main Image D) Edit the image on the bottle and make the MG, COUNT and SERVING SIZE very LARGE. To do this you will need to get the editable label file and have a photoshop expert do it.
Main Image E) Make a main image with the box alongside/behind the bottle. This is a hack that allows you to show more key info in large text on the main image.
How to set up A/B tests
Step 1) Go to Manage Your Experiments
Go to Manage Your Experiments in Seller Central: https://sellercentral.amazon.com/experiments/dashboard.
Step 2) Create New Experiment
Click Create a New Experiment → Product Images.
Step 3) Upload Main Image B
Choose the ASIN you want to test → Upload Main Image B → Uncheck “Experiment with supporting images” → Schedule Experiment.
You want image B to be clearly different than the current one.
Click Shedule experiment
Step 4) Edit Details
Click View Details to change the details.
By default, it is set to 8 weeks.
Click Edit → Set Duration → 4 weeks (the minimum) → Submit Chances
Select "Start my experiement sooner if possible". By default it will set an automatic start date which is about 1 week in the future.
Select "Automatically publish the winning version"
It may take some time to get approval. Check back in a 24 hours to make sure it did not fail validation.
How to Analyze the Results
When an experiment is done, it shows you this data chart. The key metric to look at is Units Sold From Search. This means that you got more clicks in search – meaning a higher CTR.
For some reason, it does not show CTR (click through rate) specifically… It appears that Amazon is only measuring conversion data and not click data. I certainly hope they release CTR experiments.
The only way to AB test CTR specifically is to do it manually.
The rest of the data is more useful when testing gallery images and A+ content because it has to do with CVR (conversion rate).
- Units Per Unique Visitor - This is calculated by dividing the total number of units ordered by the number of unique visitors who participated in the experiment. It's important to note that only those visitors who were exposed to the experimental content are included; not every visitor to the product detail page is counted.
- Conversion Rate - This metric represents the ratio of customers who made a purchase to the total number of customers included in the experiment.
- Units Sold - The aggregate of units purchased by customers during the experiment.
- Units Sold From Search - This includes units purchased by customers who encountered the product through the search function on Amazon during the experiment.
- Sales - Represents the total sales from all product offerings to customers participating in the experiment.
- Sales From Search - Accounts for sales from all product offerings to customers who discovered the product via the search page during the experiment.
- Sample Size - Refers to the total number of unique shoppers who were logged in and saw the experimental content. Note that not all shoppers who visit may be counted as part of the experiment. Variations in sample size are considered normal, and adjustments are made during result calculation to account for these differences.
Step 1) Get a 3D rendering of the bottle (and packaging
Why get 3D renderings?
1
Editable
You can easily play with what’s on the label & packaging.
2
They Look Better
3D renderings often look better than real photos.
Step 3) Make key info larger
The purpose is so shoppers can understand the product just by glancing at the main image.
Keep in mind, they may not read the title!
Use complementary and/or opposing colors to make key info stand out and very easily ‘glanceable’
The below image demonstrates how on most bottles, the size and count is too small to read. The green arrows indicate bottles where you can see it at a glance. Overall, bigger is better!
Main Image Ideas for Supplements
Here are some examples of different styles of main images.
Showing the back of the bottle. Showing a highlight bubble “Advanced weight loss”. Showing the gummies next to the bottle.
Show the box. This is something to test.
Showing the box: this is a good strategy to test beacause you can put more info on there.
A very large box and a small bottle. It's an idea to consider.
Showing the powder inside the capsule. Also “30-DAY SUPPLY” is large and easy to read - good.
The box on the right is poorly lit - bad.
Show the flavor by putting the fruit, e.g. an apple.
Clear bottle so you can see the gummies inside.
Good example of showing the key numbers in large, glanceable form. 500MG, 240 CAPSULES, 240 SERVINGS.
Show the capsule and the fruit.
Showing the ingredient / herb flower.
Wide bottle = fills up the whole image.
45% Better CTR
For this healthy snack product, it's hard to tell by the main image how big the bag is actually. In reality, it's a pretty big bag – truly a value size. So we added "1.5 lbs" and "20 Servings - Value Pack".
We drastically improved CTR for a client selling healthy snacks after adding the size and servings in large text on the main image.
 |  | |
Before | After | |
9% Click Share ❌ | 17% Click Share ✅ |
 |  | |
Before ❌ | After ✅ | |
0.31% CTR | 0.45% CTR |
 |  | |
Before ❌ | After ✅ | |
Case Study: 61% Boost in Sales
This is my product. The main image had a white iPhone at launch. Just by changing the color, CVR almost DOUBLED! It increased from 0.9% to 0.17%!
I was very surprised that a simple color change from red to blue would have this drastic effect!
In turn the units sold nearly 2X’d – from 13 to 21! Absolutely incredible.
We had a white iPhone for the longest time. Then I had the idea if a more vibrant color iPhone would get more clicks & sales. So we tested it vs an orange one, and sure enough, it performed better.
But luckily, we didn’t stop there! The next text of orange vs blue was incredible! Had we just called it good after the first A/B test, we would’ve lost out on almost double the sales! The blue turned out to perform almost twice as well as orange. Gotta love it!
Test 1: Orange outperformed white
Test 2: Blue is nearly 2x better than orange! 🤯
Results:
Conversion Before: 0.09%
Conversion After Round One: 0.11%
Conversion After Round Two: 0.17%
Conclusion: Two rounds of A/B testing resulted in significantly improved KPIs. This took only approximately 5 weeks total.