Updated 2024 Amazon Experiments Data
Each year we work with over 150 brands to elevate their SEO and Visual content on Amazon. Since 2023, we’ve completed over 200 split tests on Amazon, 82.1% resulted in successful results which we define as an increase in sales.
Our best single split test generated a 1008% increase in orders. Our best performing client overall saw an increase in sales of 1752% from a complete content refresh with us before investing in advertising.
We used a combination of Manage Your Experiments and manual testing explained further below. We also ask all clients to keep all other factors as they were including PPC, no turning spend up or down during these periods of testing.
How is data collected manually if brands don't qualify for MYE?
If a brand does not qualify for Manage Your Experiments (Amazon Experiments), here's how we pull the data manually:
1) We take the previous 90 days of sales data for a baseline average.
2) We upload the new content and make no other changes
3) We pull the data 15 to 30 days later depending on traffic to compare it to the baseline
Note: This is the increase in order seen from one individual test not across all testing done for each brand. See individual break below for each test type.
What else can we learn from the split testing data?
Sessions data will vary based on the stage of the brand. Early stage and new-to-launch brands can see larger gains as they may have had a false start and have more room for growth than established brands already selling well on Amazon.
More than 50% of clients can expect to see $5-120K increases in revenue while that same 50% that achieves a best case scenario could see $15-205K increase in revenue. Best case scenario is achieved when brands have a strong traffic strategy after optimizing content.
Which type of split test should you start with?
We often get asked how often content should be refreshed on Amazon. The truth is that once you have a strong foundation of high quality content, you should move on to testing and you should always be testing.
But then, which test should you start with?
We’ve created a breakdown of the results we are seeing with each type of split test we run including titles, main images and A+ Content.