Updated 2025 Amazon Experiments Data

When a brand doesn't qualify for Manage Your Experiments, we handle the data collection ourselves. We grab 90 days of sales data as our baseline, upload the new content without changing anything else, then pull fresh data 15-30 days later (depending on how much traffic they get) to see how it compares. Keep in mind, these results are from individual tests, not combined across all the testing we do for each brand.

Your mileage may vary depending on where your brand is at. Newer brands or those that had a rough start on Amazon tend to see bigger jumps since they have more room to grow compared to brands that are already doing well.

More than half of our clients see revenue increases between $5-120K, while that same 50% in a best-case scenario could hit $15-205K increases. The best results happen when brands have solid traffic strategies backing up their optimized content.

Most people think main images are the way to go but after running 250+ split tests per year, we've learned that A+ Content actually has the highest success rate at 80.5%, followed closely by main images at 79.45%.

Even still, here's where you should really start:

Start with SEO copy (titles & bullets) - they're the easiest to test and can give you up to 990% increases in orders. Just create alternate titles using keywords that are already converting well for you.

Then tackle main images - while A+ Content has the highest success rate, main images are much quicker to edit and test. You can see up to 547% increases in orders by optimizing your main image and product image stack.

Save A+ Content for last - it takes the most time and effort to get right, but when done well, it delivers the best returns with up to 1,080% increases in orders.

The key is to keep testing throughout the year, not just run one test and call it done. And remember, our results come from working with eCommerce brands (mostly under 50 SKUs doing less than $10M annually on Amazon) whose listings are typically under-optimized, which is why we can squeeze so much value from testing.

We recommend doing a full content refresh once a year, ideally when you have any branding updates and sometime between Q1-Q3. Once that refresh is done, you should be split testing different elements like titles, bullets, main images, product images, and A+ content throughout the year - right up until peak holiday season hits.

Here's the key: don't mess with your listings during peak holiday sales unless you already know your current content isn't working and you've got nothing to lose. The last thing you want is to accidentally tank your sales during your biggest revenue period of the year.