Creating a New Split Test – Step-by-Step guide

Here’s how to create a Split Test in Trendle:

Step 1:

Go to Split Test page and click “Create New Test”

Step 2:

You then select which element you want to run a test on.

You can choose: Price, Sale Price, Main Image, Bullet Points, Description.

You can only test one element per test. If you were to test many elements at once, your tests would not be conclusive and you would not be able to say which element was responsible for the change in performance.

Please note: You must have full listing control in order to test these fields. This is especially important for the Title. To know if you have listing control, change the title manually in seller central and wait to see if this gets applied to the Amazon site. If not, and you have to contact Amazon to get this applied then you do not have full listing control and some of your tests are likely to fail (especially on title). Price should not be affected.

Once you’ve selected the element to test, you can choose which country this test should run in.

You then search for the SKU you will perform the test on. Once you’ve entered the SKU or ASIN or Product Title, click on “Search”. The application will then show you the matching results based on your inventory catalogue.

Select the SKU you wish to run the test on and click “Next”.

Please note: If you’ve chosen to test the Main Image or the Bullet Points, the application will need up to a minute to fetch the latest data from Amazon. Please be patient as we retrieve this live information. We do this to ensure you have the latest listing details.

For Price and Sale Price, there is no wait as the application updates these values daily.

Step 3:

  • Element: This is the variable or element being tested (for example, “Price”, “Sale Price”, “Image” etc)
  • Test Name: You can give your test any name to help you refer to it later
  • Original Value: This is the current value live on Amazon. This will be the “A” of the A/B Test. For example, $50
  • Test Value: Enter here the alternative element to use for the test. This will be the “B” of the A/B Test. For example, $53
  • Test Start Date: This will always be the next day by default. It’s not possible to start the test on the day you created it as this would mean using an incomplete day for the test which could make the test results less reliable.
  • Test End Date: This is the last day that the test will run (including that day). By default this is set to 1 month.
  • Rotation: This is how often the application will change between A/B. We recommend changing every 1 day, but you can choose any value up to 30 days.
  • Save & Schedule Campaign: Click this button to finalise your campaign. Once this button is clicked, you can no longer edit that campaign. To make changes, you would need to end or delete this campaign and to create a new campaign
  • Save as Draft: You can save a campaign as draft if you are not ready yet to commit it to going live yet. You can later go back to this campaign and edit it further.

 

Step 4:

In the overview table you can see all your past, current and future planned tests.

By clicking on the test name you will be guided to view the details of that test including any data such as Sessions, Conversions, Sales and Revenue.

Step 5:

  • In the first section you are shown the elements being tested

If the Test is in draft mode, then you can still edit the variables.

If the Test is running you can choose to End the Test early.

If the Test has ended, you can choose which of “A” or “B” you would like to implement as the permanent solution.

  • In the following sections, you can see the data associated to the test. Read this article for more information on how to analyse the results

Currently you can see the Sessions, Conversions, Sales and Revenue per day for the SKU being tested. This allows you to conclude which variant A or B has the desired impact.

Note: In order for us to display Sessions and Conversions, we require limited access to your Seller Central account. For more information, please read this article.

Was this article helpful?

Related Articles

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.