Your browser is out of date. The site might not be displayed correctly. Please update your browser.

October 11, 2021

How to Maximize Amazon Profits by Split Testing Your Listings

Your Amazon success is essentially made up of two layers:

  • Your product listing should be discovered by surfers;
  • It has to be strong enough to convert them to bring you sales.

Your product listing is the most important element that affects both layers. With no one-size-fits-all approach, the only way to maximize the efficiency of your listing is to test, test, and test.

In this article, we’ll go over:

  • Why you should split test your Amazon listings
  • How to go strategic about split testing
  • Tools for running a split test on your Amazon listing

Sellzone partnered with Kaspien to run a webinar that covered how to go smart about split testing your product listings to see the maximum impact. Jenn Johnston, a digital marketing specialist at Kaspien, and Dan Saunders, an e-commerce channel manager at Stanley Black & Decker, joined the conversation to share their expertise and insights on the matter.

Why You Should Split Test Your Amazon Listings

With more than 1.3 million sellers, the Amazon marketplace is oversaturated with various offers and products. And even if you manage to find that winning product, lack of visibility can hinder your potential.

This is where split testing comes into play and can potentially save the day.

Once you see what split tests can do, you’ll never ask why you need to run them:

  • Maximize visitors and engagement

As mentioned, the key part of Amazon’s success is still to be discovered. And while we do have some insight into Amazon’s ranking algo, it’s not as straightforward as steps one, two, three. A successful test, however, can help you maximize the number of visitors that come to your listing and drive higher engagement and conversions.

  • Eliminate any uncertainty

If you’re only relying on hunches and intuition, thinking that this or that image should help you with conversions or even at the attraction stage, you’re in the wrong. Split testing is just about the only way to crack the Amazon algo and show you what really works and doesn’t work in the marketplace.

  • Increase profitability

Higher profitability is the end result of all the perks listed above. If you’ve split tested the heck out of your listing, it should be set for maximum results, with its high rankings and top conversion rate. And this means you’ll drive more people to your listing and successfully turn them into buyers – all leading to higher profitability.

  • Rank higher

Listing conversion rate is a ranking factor. When you have more conversions (aka sales), your listing will get to rank higher. In turn, that increased traffic can help you get more conversions, further improving your product's overall ranking. And split tests can help you improve both rankings and conversions.

How to Go Strategic About Split Testing Your Amazon Listings

Having a plan, strategy, and tactics in place is the best way to go about your tests:

  • Pick the right listings (if you own/manage multiple product pages); and
  • Choose the right elements to test.

Selecting the right listing for the test

While you can and maybe should test multiple listings at the same time, you still have to make your picks. And there are two key ways to prioritize which product listings should get the first go:

  • Split test your top-performing listings

Dan shared that on Amazon, 20% of listings drive 80% of revenue. So a good strategy could be to pick the listing that’s already bringing you the biggest results and further optimize it for even more success. Spending the time to perfect your high-performing listings helps to give you incremental advantages over your competitors.

  • Start your split test with underperforming listings

If you believe that your product page has the opportunity to bring in more sales, but it hasn’t reached its maximum potential, you can start with underperforming listings. Sometimes split testing the lower performers can have a more dramatic impact on conversion, helping you to identify what is causing your listing to underperform and hone in on the message that will resonate best with your customers.

Pro tip: If you want to go all data-driven on your listing choice, you can take Dan’s advice and use Sellzone’s Listing Quality Check tool to see which listings have more issues than others. Take a look at the total score of your listing quality, and if it’s at less than 85%, it has a high potential for suppression. So this would be a good one to start with.

Choosing the right elements to test

Both Jenn and Dan agreed that if you have the time and resources, it’s best to test each and every element on the page:

  • Title
  • Bullets
  • Images
  • Videos
  • Standard description
  • A+ content

Most sellers focus on title, bullets, and images, but our experts’ experience shows that one small tweak to other elements can also make a big difference – so when it comes to split testing, the sky’s the limit.

And even with the more ‘traditional’ testing elements—say, images—you can find an original testing angle. For instance, Dan shared that some sellers now experiment with image metadata, adding keywords to the file name, which apparently also makes an impact.

But, and this is important to follow, always test one element at a time. Otherwise, you won’t understand what caused the impact.

Things to consider before you start testing

Now that you know the ins and outs of a split-testing strategy on Amazon, there are a few things to think about before you go on this experimental journey.

  • Make sure you have the perfect timing for split testing

While you certainly can run split tests whenever, there are some periods when you should stay away from experiments. Dan—someone who has run tens of thousands of split tests during his career—emphasized that “peak” seasons and times of big sales (think Black Friday or Cyber Monday) are a big red flag for testing. First of all, you might get unrealistic results (sales do go up during those periods anyway). And secondly, if your hypothesis is wrong, you might lose revenue.

  • Formulate a hypothesis

It’s best to go about split testing with some kind of hypothesis in mind. For instance, you might think that a more friendly look could give your listing a boost. Then, you have to test all elements of the page, tweaking them according to your “friendly look” hypothesis, using a more conversational language, brighter imagery, etc.

Pro tip: as you are formulating your hypothesis, remember that there are also external factors to consider that could affect your testing. It could be as simple as particularly harsh weather impacting the sales of your outdoor items or an influencer promoting your products on Instagram.

  • Factor in test duration

With split testing on Amazon, the longer, the better logic applies. Many Amazon experts recommend keeping the minimum test duration at 4 weeks. So if you’re running the split test manually, this means you get your listing locked up for four weeks, as you can’t test simultaneously.

But this is where special split-testing tools can really help you keep that duration at 4 weeks, enabling you to split test your listing in parallel.

Free unlimited split testing

Tools for Running a Split Test on Your Amazon Listing

Jenn and Dan both shared some insights on the top free solutions for split testing on Amazon. Jenn covered Manage My Experiments – a split-testing tool provided by Amazon itself. Dan shared his experience with Sellzone’s Split Testing tool.

Why use a special tool to run your split tests?

As you already know, the best part of using a special split testing tool is that you don’t have to lock up your listing for weeks while your test is running. But there are other perks as well:

  • You get to see real-time data.
  • You can automate the whole split-testing process.
  • You’ll quickly see any negative impact and be able to troubleshoot promptly.

But not all split-testing tools are created equal.

Amazon Experiments vs. Sellzone’s Split Testing tool

Jenn and Dan shared their pros and cons for each tool, comparing the two solutions:

Amazon Experiments

Amazon Experiments offers some basic split-testing capabilities but comes with a lot of limitations:

  • You have to prove your ASIN’s eligibility.

Amazon doesn’t let you test just any listing. And its decision-making process is not very transparent. Jenn mentioned that listings that are allowed to participate have to have enough sales and traffic. But Amazon never clearly states how much traffic and sales you must have exactly to qualify.

  • Amazon Experiments is only available for brand-registered sellers, while the Split Testing tool is available to any seller for free.
  • You can only test three elements: product title, main image, and A+ content. The rest is up for manual testing.
  • The testing period is 4-10 weeks, with the earliest cancellation time being at 4 weeks. So if you see some dramatic drops, you won’t be able to stop your test.

Once your split test is over, Amazon Experiments will showcase metrics that should help you decide which listing performed better:

  • Sales
  • Conversion rate
  • Units sold
  • Sample size

Despite some of the cons, Amazon Experiments comes with one unique feature which makes it really valuable – probability. Using Amazon’s internal sales data, the tool shows the probability on which listing version could perform better, with an estimate of potential incremental sales and units sold.

Split Testing tool by Sellzone

Capabilities-wise, the Split Testing tool comes with a lot more freedom. You can:

  • Enter any listing and set it up for a split test.
  • Set up your test for any time period (the setting is 7 days, but change to at least 4 weeks).
  • Split test any element of the page – but do make sure to test one thing at a time.
  • Get live data and stop your test at any time – this helps you take timely action and interrupt the test.
  • Keep track of all your tests in a single dashboard.
  • See metrics for sales and units sold and decide which version of your listing is a definite winner.
Free unlimited split testing

Over to You

Thanks to our experts, now you know all about the ins and outs of split testing your product listing. And that’s already a good start to your Amazon success.

Read further:

  • If you want a full guide on how to create efficient A/B tests for your product listings on Amazon, make sure to head over to this post.
  • Read this post to learn about the anatomy of a best-selling Amazon listing – to know which parts of your listing to tweak first and how.Make sure to go through this extensive guide that will walk you through the key steps to seeing top results on Amazon.
Send feedback