Would you like to launch an experiment on Amazon but don’t know how to go about it?
Afraid it’s complicated and time-consuming?
Nothing could be further from the truth!
We show you how to launch an experiment on Amazon in 5 easy steps!
Read on, or watch our video.
Step 1 Select ‘brands’ tab and click ‘manage experiments’.
If you don’t see this tab in your Seller Central account, it means you don’t meet Amazon’s main requirement to be able to run experiments.
This feature is only available to brand owners registered with Amazon’s brand registry.
If you see it, select it and then click ‘manage experiments’.
Step 2 Choose the type of experiment
Next, click the “create new experiment” button and select which part of the listing you want to test.
Your choices are:
- main image
- A+ content
Step 3 Select ASINs
At this stage, you will encounter another Amazon restriction on experiments.
In order to subject any of your products to the experiment, you must have products with sufficiently high traffic in recent weeks.
This is because, for the results of the test to be reliable, they must be based on a sufficiently large sample group. If a product is rarely clicked on and bought by users, then the results could be distorted – e.g. a version that is not better at all but got one more sale could win, with a total of 3 sales combined for both versions.
Step 4 Complete the experiment data.
Once you have chosen a product for the test, you can define the basic elements of the test: title, hypothesis and duration.
The longer the test lasts, the more certain you are of the results – 8-10 weeks is recommended.
The start day of the test is always a few days away. This is because Amazon needs time to check that all elements of the test meet the requirements.
Step 5 Add the B version and submit your experiment
Add a title, image or A+ content for the B version.
You can select the A+ content for the test from the list of available versions, or you can create a new version for the test.
Then click ‘submit experiment’.
This will submit your test for validation.
Come back in a few days and check if your test has been accepted and run.
If not, make changes and run it again.
Important – Variants in the experiment
You can also add tests for products with variants.
However, you may find that not all variants are qualified for the experiment because, for example, they generate too low traffic.
If you don’t want to test some of the product variants, just don’t add a B version of the title, image or A+ content for them. However, you must always add a B version for the parent ASIN.
Want to know more about Amazon experiments?
Follow our blog and channel and stay up to date.