What is Amazon AB and Split Testing?
Split testing, or AB testing, on Amazon is the method of listing optimization through experiments with product content and pricing. There are four core variables of the listing that affect conversion rate: Image, Title, Description, and Price
This page covers the basics of Amazon split testing:
- How to split test your Amazon Listing
- What elements of the listing you can tweak
How to split test your Amazon Listing
You might have good sales, but there are always ways to improve. You can’t know for sure, unless you have the test data to prove it. Your Amazon listings are most probably underperforming right now, and you are losing money.
Split testing is a method used by Amazon sellers for data-driven optimization:
- Take one of your product pages, make a note of how many views it has and its conversion rate.
- Tweak your listing in some way and wait for a period of time before collecting a new set of data.
- Compare the results with the initial numbers, and see if you’ve benefited from the change.
- Rinse and repeat.
Split testing is completely in line with Amazon's rules and regulations, just remember to follow its official Style Guide.
Running multiple tests for multiple listings manually would be inefficient and insanely time-consuming, but you can automate that process with AB testing software for Amazon.
Our AB testing tool is free and there are no limits, so you can run as many tests as you like.
How long does it take to get test results
You will be getting live testing data with our tool, but to get enough data for a definitive conclusion, we recommend you wait for at least two weeks per experiment.
Remember that there are various forces (e.g. holiday season) that can disrupt the statistical confirmation, so you might want to add a couple of weeks for the experiment to be valid.
When is the testing complete
Strictly speaking never. Split testing is all about gradually improving upon improvements.
Of course, there’s a scenario where you hit your peak conversion rate, but it’s not going to stay that way forever. The market is a boisterous scene, and you need to keep working to stay on top.
What result numbers are good/bad
It’s really on a case by case basis, but even a 1% bump in your conversions might be worth hundreds and thousands of dollars that you could be making.
What elements of the listing you can tweak
You can perk up your product pages by changing and split testing the imagery, wording, and pricing.
The buying decision process is inherently visually guided, and the importance of products’ images cannot be overstated. Your main product image is the first impression you will make on a person browsing the merchandise, so try to stand out while showing the best qualities of your products.
Invest in professional photos, try different angles and positions, try packed and unpacked presentation, show products in use. But before you start snapping away new pictures for your listing, it might be a good idea to freshen up on Amazon's official image guidelines.
Once you’ve grabbed your customer’s attention with a great image, the title of your listing comes into play. It should be relevant and concise, but still informative and inspiring further examination.
Description is the next element for you to tweak here. Descriptions provide more space to elaborate on a product’s strengths, so use them wisely – start by telling how your product solves your customer’s problem. Then describe other benefits, your type of product and its functions.
Just remember, while chasing for the most compelling phrasing, that you should actually name your product in the title, and give it an honest description. Also remember to add keywords to your Amazon listings, so that your products will be visible in search.
Pricing probably should have been first on the list, but it would have been too obvious. Besides, if you’ve crafted a perfect listing with all of the previous elements, your price is right.
Unfortunately, it’s not as simple as “raise price - get more money.” You shouldn’t be afraid to test out higher prices, but if you’ll shift fewer units for a higher price, your eventual revenue might go down, while actually lowering your price may significantly increase your profits. All of this is a definite “maybe”, until you proof test it. There’s a lot of intricacy and some can be predicted, but still only to a certain point.
Leaving economics aside, there’s also a lot of psychology involved. Many, many subtle factors affect the price processing of your customers. What would you go for—$22.00, $21.97, or $21.99?
Don’t jump to a conclusion too quickly. Remember—it’s not all about math. Studies show that even the syllabic length of the numbers in your price have an influence on decision making. The longer it takes to say the number, the more mental space it occupies, and the larger the number seems.
Factoring in all of the variables and making the right choice on the first try is close to impossible. Testing out theories and collecting data is critical for the successful seller.