top of page

Split Testing on Amazon


Split testing on Amazon is similar to traditional A/B testing marketers usually use to test different variations of websites to find what performs the best and what users respond to strongly. In a nutshell, split testing basically means comparing two versions of a website (or Amazon product listing in this case) to determine which one people like the best. This is done by displaying the two variations to similar visitors at the same, and whichever one has the best conversions is the winner!

Why Split Test on Amazon?

It sounds like it isn’t that complicated; most people would think they can just make changes to their product listings, let it run for a period of time, and then compare it to the previous version. While this is certainly doable, it won’t provide the best, most accurate data. Amazon is a complicated system with many everchanging components. Between things like seasonality, demand, competition and market conditions, it would be difficult to determine if your product is selling more due to market changes or your listing updates. This is why it’s best to run a split test of two variations simultaneously so both variations are shown to customers under the same market conditions.

How to Split Test on Amazon

Split testing on Amazon works by helping you send traffic to two variations of your listing in order to gather statistically significant data regarding how customers react to each variation. Amazon has become very competitive, so split testing helps you figure out the best way to stand out from your competitors. By split testing on Amazon, sellers can craft the perfect product listing in an measurable, verifiable manner that results in attracting more visitors, higher conversion rates, and ultimately higher profits. While you can test variations of all aspects of your product listings, the most recommended things to try split testing on Amazon are price and main product image.

Split Testing: Price

Split Testing - Search Results

To start price split testing on Amazon, you should raise or lower your product’s price by 10%. Raising your product’s price can potentially position it as a higher-quality product. The downside to increasing your product’s price is that it may lead to less sessions and conversions because those customers who are price-sensitive will shy away from your listing but your profit margin for each sale will likely increase.

On the other hand, lowering your product’s price by 10% could attract price-sensitive customers who are focused on limiting spending. It can also make it more appealing than competitor offerings. The reduced price may have an impact on your profit margins, but the idea is that your product will sell more and not affect your profits.

By simply adjusting your price by a few dollars in either direction can have a dramatic impact on your overall sales and profit. Sellers have seen profits significantly increase by raising or lowering their prices and the best way to determine what price changes to make is to do split testing on Amazon.

Split Testing: Main Product Image

You only get one chance to make a first impression, right? So don’t you want to make the best first impression possible, especially on a cluttered platform like Amazon? Of course, the answer is yes. Customers make the decision to click on your product listing in a split second, and your main product image is a big part of that decision. Therefore, you want your main product image to be as compelling as possible.

Take the following search screenshot as an example. If you were searching for a dinosaur costume, you’d find that most of the main product images are the same. It’s hard to stand out based on your main product image if everyone is using the same or similar images.

Even if you are selling the same thing as someone else, try to find new and creative ways to photograph the product. Test this new main product image against your existing one to see if it helps your listing stand out in search results.

These are just a couple ways you can do split testing on Amazon. You can also test things like your headline, bullet points in the product description, or the full product description. Just remember to only test one thing at a time so you can definitively say what is and is not performing well.


The Manual Method

Source: Splitly

Automated Split Testing on Amazon

If you don’t want to hassle with running split testing on Amazon manually, there are tools that can help you manage it.


PickFu

PickFu is instant polling software that works entirely outside of the Amazon marketplace. Rather than splitting your traffic to various versions of your Amazon listing, you show all the versions at once and survey people as to which version they like better. This method is particularly effective when it comes to testing your main product image, as your survey respondents will not only vote on which image they would more likely click on, they provide a written answer about what appealed to them. These comments are enormously insightful as you optimize your product listings, and can also work to test product names and descriptions. On PickFu, you can use demographic and behavioral targeting to make your survey audience look more like your product’s customer base. For example, you can survey a group of Amazon Prime members, women within a certain age group, dog owners, people whose incomes are over $100K, etc.

One advantage of polling outside of Amazon rather than splitting traffic within Amazon is you aren’t risking an adverse effect to sales. Another advantage is that your Amazon listing doesn’t have to be live, so you can test images, descriptions, and product names before launching, and even test product variations before going into production or buying inventory. PickFu’s results are also lightning fast, as most polls complete in about 15 minutes.

Listing Dojo

If you are looking for a free tool for split testing on Amazon, check out Listing-Dojo.com. You can easily sign up for a free account and authorize your Amazon Marketplace Web Service (MWS) account. By granting this access, Listing Dojo will pull your product catalog, make updates to listings, and pull your sales reports. You can even gain access to see clicks, sessions, units ordered, and total orders data by gaining access to your business reports.

Then you’ll just need to select your test elements (price, image, title, description). Listing Dojo does not recommend testing content as it involves re-indexing your listing and that can have a negative impact on your sales. If you are a newer seller, it can be an option but for the more established sellers who already know what content works best in their product listings, it’s best not to split test your content.

After you’ve decided what element to test, then create a campaign by selecting the number of days (between 1-31) you want the test to run. It’s recommended that you run each variable for 7 days. Then enter your independent variables.

Whenever one variable finishes testing, you’ll be sent a metrics report. You will be able to see detailed information about how that variable performed, including the baseline, which is your control. You can see the data as a side-by-side comparison and also a detailed performance of each tested price.

Splitly

Splitly is a paid split testing tool that requires no manual work and claims to have the most sophisticated algorithms in the industry. You start by uploading your variants: price, features, images, descriptions or back-end keywords. Then you run the experiment and let Splitly optimize your listings by rotating through variants. Splitly will pull reports directly from your Amazon Seller account to determine the statistically significant winner. Splitly offers a wide range of plans, so you should be able to find something that fits your needs and budget.

Keith Villano with Splitly had the following to say about how split testing on Splitly can help Amazon Sellers:

“With Splitly, you can test your price, images, tiles, bullets, and descriptions to find out what resonates and converts best among your customer base. We’ve helped over 1,000 sellers net an increase of $86,072,958 simply by testing their listings with Splitly.  Unique to Splitly is Profit Peak, it’s a dynamic repricer that you can set & forget. Using Profit Peak, you’ll be able to find your optimum price where you can both profit and sell the most for each unit sold thanks to the built-in machine learning algorithm which tracks how your product’s price affects your product’s demand so that you can always be sure that your products are priced accordingly by the day of the week, by the time of the year. It’s about time sellers take control of their listings once again and start uncovering money left on the table with Splitly.”

Sellozo is Here to Help

Split testing on Amazon can be tricky and intimidating. If you have questions regarding running split tests for your Amazon Product Listings, our knowledgeable Product Specialists are happy to help! You can schedule a time to chat with one of them by clicking here.


Try Sellozo!


Sellozo offers a simple, powerful platform that makes it easy to manage and optimize your Amazon Sponsored Product Ads and – thanks to machine-learning – programmatically alters bids to help your ads beat out the competition.

.fusion-button.button-4 .fusion-button-text, .fusion-button.button-4 i {color:#ffffff;}.fusion-button.button-4 {border-width:0px;border-color:#ffffff;}.fusion-button.button-4 .fusion-button-icon-divider{border-color:#ffffff;}.fusion-button.button-4:hover .fusion-button-text, .fusion-button.button-4:hover i,.fusion-button.button-4:focus .fusion-button-text, .fusion-button.button-4:focus i,.fusion-button.button-4:active .fusion-button-text, .fusion-button.button-4:active{color:#ffffff;}.fusion-button.button-4:hover, .fusion-button.button-4:focus, .fusion-button.button-4:active{border-width:0px;border-color:#ffffff;}.fusion-button.button-4:hover .fusion-button-icon-divider, .fusion-button.button-4:hover .fusion-button-icon-divider, .fusion-button.button-4:active .fusion-button-icon-divider{border-color:#ffffff;}.fusion-button.button-4{background: #55c2f4; background-image: -webkit-gradient( linear, left bottom, left top, from( #03a9f4 ), to( #55c2f4 ) ); background-image: -webkit-linear-gradient( bottom, #03a9f4, #55c2f4 ); background-image: -moz-linear-gradient( bottom, #03a9f4, #55c2f4 ); background-image: -o-linear-gradient( bottom, #03a9f4, #55c2f4 ); background-image: linear-gradient( to top, #03a9f4, #55c2f4 );}.fusion-button.button-4:hover,.button-4:focus,.fusion-button.button-4:active{background: #03a9f4; background-image: -webkit-gradient( linear, left bottom, left top, from( #027fd8 ), to( #03a9f4 ) ); background-image: -webkit-linear-gradient( bottom, #027fd8, #03a9f4 ); background-image: -moz-linear-gradient( bottom, #027fd8, #03a9f4 ); background-image: -o-linear-gradient( bottom, #027fd8, #03a9f4 ); background-image: linear-gradient( to top, #027fd8, #03a9f4 );}.fusion-button.button-4 .fusion-button-text {text-transform:uppercase;}.fusion-button.button-4{width:auto;}Get Started


Meet the Author: [wbcr_php_snippet id=”18211″]


[wbcr_php_snippet id=”18212″]


[wbcr_php_snippet id=”18210″]

Comments


bottom of page