Ecommerce Marketing / Enterprise Ecommerce / How to Sell Online

A/B Testing for E-commerce Stores: What it is and why you should use it, plus 10 tests to try

Smriti Chawla / 5 min read
Table of Contents

    Tell me honestly — is your e-commerce website optional? Or does it really mean some serious business for you? I ask because if your site is really something you value, then you absolutely need to understand the power of A/B testing.

    You invest a lot of time, money and effort into making your online store successful. And when you’re receiving tons of traffic, it’s natural for you to hesitate in making even a small change on the website. Okay, you “feel” like it should help increase your conversion rate. But the doubt remains — what if it backfires?

    You probably have a lot of ideas in the back of your head about changes in your heading, discount offers, pricing, layout, and so on. But maybe the fear of plummeting conversion rates has always stopped you from trying them out.

    That’s where A/B testing can really move the needle. It allows you to measure the success of proposed changes while limiting potential damage.

    What is A/B testing, exactly?

    Imagine how great it would be if you could just poll your live visitors and ask if they would like a certain change on the website or not. Unfortunately, that’s not how it works.

    A/B testing lets you do something very similar, except people will not even be aware that they are voting for something. It allows you to make two versions of your landing page. The first will be the Control page, which is the original page on your site. The second is the Variation page, which will be nearly identical to the Control except for the one change that you want to test. These two versions are then tested with live visitors on your site to check the conversion rate and find the winning page.

    Here’s a visual of how A/B testing works:


    Your testing goal can be anything from selling more of a certain product to site engagement, form completion and more. The goal completion of every visitor is tracked for both versions of the page and the winning version is declared once statistical confidence of the data has been achieved.

    How does A/B testing help e-commerce sites in damage control?

    Sometimes when you’re apprehensive about changing your website, trying out a variation that seems to have, say, a 50-50 chance of success requires courage. After all, if the variation doesn’t get the desired response, it can cause lost leads or conversions. And if yours is a website that receives heavy traffic, that loss can accumulate and become overwhelming within a day or two.

    A/B testing tools help you keep that potential damage in check. For instance, you could show the Variation page to only 20% of your visitors, while the remaining 80% of the visitors still see the Control. You test the two versions of the web page with your live visitors and see how they respond to the change. You can always change the percentage who see the variation according to your specific requirements.

    You can even track your revenue while this test is going on, and see for yourself which page (Control or Variation) is performing better for your unique set of visitors. As the statistical confidence of the test is achieved (based on the number of visitors on the site), the winning page is declared.

    In simple terms, you’re no longer hanging in limbo, contemplating if you should go ahead with making a permanent change to your site. You simply A/B test for a few days and get a definite answer based on facts and proven data. After all, it’s about what works for you and not what you think should work.

    A/B testing has gone mainstream

    The primary reasons why A/B testing has become increasingly popular for e-commerce sites of all sizes and a must for e-commerce giants like Amazon and Ebay:

    1. Major potential for improvement — There are so many elements to an e-commerce site that it presents a multitude of opportunities for testing: from the checkout process and product pages to site layout, or even a complete design overhaul.

    2. Conversion rates directly impact revenue — For e-commerce sites, the impact of A/B testing can be directly measured in terms of revenue. It’s not about lead generation, form filling, brochure download or anything else; it’s a direct sale. Successful tests can help you reduce your cart abandonment rates and average order value, or analyze your user behavior to find out what interests them on your site and what doesn’t. This case study details how increased their sales by 15% just by testing three simple changes to their checkout page.

    Tests to try for optimized conversion

    If you’re considering A/B testing your site, here are some of the most commonly tested elements for e-commerce sites:

    Pricing and Shipping Strategies

    Many people would visit your site just to take a look around, check out the latest products, etc. But how do you reel in those window shoppers and make them click that “buy now” button?

    We all know pricing greatly influences consumer decision making. So how do you make it work in your favor and not against? Should you start ending your prices with .99 or other variations as seen in tons of other stores (you can see the pattern of ending prices with 95 in the image below, taken from How is your latest discount offer working for you? Should your prices be displayed on the left-hand or right-hand side of the image? Are your visitors abandoning carts after seeing your shipping charges?

    You can find answers to all of this and more through A/B testing your website. Check out how SmileyCookie increased their sales by 41% by offering next-day shipping to see exactly what I’m talking about.

    Product Pages and Product Filters

    There are so many things you can test on your product pages. For example, raised their conversions by 9% by increasing the size of their product images. Of course, there are best practices that you can implement and test right away, like placing “buy now” buttons above the fold, using bright and contrasting button colors (as you can see on Sony’s official website), including trust badges, and so on.

    Some of you may wonder why you should A/B test if these are proven practices. That’s the whole deal with A/B testing — you can never be sure about the results until you test them on live visitors. For example, here’s a case study about how removing a product filter increased site engagement by 27% for But in other tests, including product filters has been found to be a good practice that increases revenue.

    For some sites, displaying best-selling products may work better for cross selling. For others, showing the latest products might be more effective. Similarly, you can test which search result pattern works best for you: the grid or the list. You can even test if cross-selling increases revenue (by providing more choices and encouraging customers to buy more), or decreases revenue (by offering too many choices and confusing your customers about what they should buy).

    Once you get into the groove of A/B testing, the entire site is your playground. Figure out a solution to reduce your cart abandonment rates, try out different product page layouts, experiment with product descriptions, headlines, button placement/colors and more — and track your revenue for each of them.

    You might not always find that your changes work, but many times they will. And moving forward, the insights into shopper behavior will help you make your site a lot more effective and eliminate the friction that causes lost sales.

    The best part is that you do not have to be a techie to get started with A/B testing. With tools like Visual Website Optimizer, it takes just a few minutes to create a test and you don’t need any knowledge of HTML or coding. Try out A/B testing and see how it works for you!

    Most Popular Reads


    Leave a Comment

    Less Development. More Marketing.

    Let us future-proof your backend. You focus on building your brand.