Using A/B Testing to Improve Sales in E-commerce
There’s no better way to improve sales than to understand how your customers behave. As a business owner, having the peace of mind to predict what is going to perform better based on behaviour pattern is a superpower.
By testing hypothesis about customer behavior you’re capable of influencing your sales in a larger scale. Sometimes, this is actually fairly easy and fast, in case you have a small or medium sized e-commerce store.
What is A/B testing?
A/B testing is a method where you compare different versions of something to see which one performs better. This can be used on a website, focused on UX, advertising, email or any sort of contact you have directly with your clients. It is, in essence, an experiment, where you change one single variant and show it to different users. Later, you analyse the data to see which one performed better and stick with the one with the bigger conversion rate.
This runs unnoticed by the client, since they do not see the other version of what they are looking at, but provide you dozens of insights about them. Sometimes, just by modifying a single word on your website, you are going to experience big changes.
One of the best pros of running A/B tests is that some of them can cost you absolutely nothing or very little. Unless you’re implementing a new layout or some fancy new strategy or design, you can try small changes yourself if you have a good way of measuring them.
A/B testing is also extremely useful for when you’re planning on implementing something new in your ecommerce, blog or website, but wants to test the waters before going all in. It’s a way to understand how people are going to react to the changes before modifying it completely.
A/B testing can be used on basically anything, from design to product descriptions.
Make sure to check out this real case we had with two merchants using EBANX operating within the same industry and see for yourself the power of a/b testing.
“The average conversion rate for boletos processed through Direct API on the same period was 59.5% while the conversion rate for boletos processed through checkout integration it was a scant 24.55%, an average difference of 35% for the entire period.”
How an ecommerce can benefit from A/B testing?
As an owner of an eCommerce store, you’re aware of all dozens small and big triggers to make customers shop. Look at this as your playground! You can move bits and pieces to try and have a better conversion rate, from layout to email marketing campaigns to highlighted products.
Even though A/B testing can be done quite easily, the results can be tremendous.
Especially in your homepage, you can strategically move key-products to be displayed so you get a better grasp of what your customers are looking for. This is especially useful to those who do not have a big selection of products but want to retain their user as long as they can on their store.
You can create identical landing pages for different products, advertise them on the same platform using the same strategies and analyse which one performs better. Or just change a small call to action. For instance, take this example below if you want to boost your email list. Just by adding your sign-up forms in a different place around your page can provide big changes in your results. That’s how you’ll now which one to implement in the long run.
The rules of A/B testing
Even though some people swear by the change-one-thing-at-a-time rule, this is not always the case. Sometimes it’s better to test a whole new website than just keep doing small changes over time – also, this would take you a lifetime and wouldn’t reflect on the final result.
In any case, to try smaller experiments you should stick to one or two things at a time. That way you are going to be able to measure exactly what brought you the better or worse results and revert it. If you decide to change a CTA, don’t change the background color. Try this later, when you already have the results of the new content.
Pick one thing to test and stick to it. Don’t forget that you can always run another test as soon as you have the results from the previous one.
Before you begin your a/b tests, however, here’s a list you should remember:
- Make sure you have a schedule. Establish the day you are going to stop the experiment and analyse the results.
- Have an audience: even if it’s a small, it needs to be consistent. It’s difficult to make A/B tests when the data is not quite there yet.
- Define what’s the goal: small changes can bring big results in very specific fields. Even though your conversion rates are going up, the changes might affect negatively your time on site. Make sure you stick to the goal but understand what’s happening to your website in general.
- Keep things simple. The more variables you have, the harder it is to keep track of which one is causing which effect.
- Keep testing. Make sure you analyse results and keep improving your store little by little.
Three A/B tests to try on your eCommerce store
A/B Test #1 | The Promotion
When creating a campaign offering a discount code or any sort of promotion on your eCommerce, create two identical banners offering the same discount but phrasing it differently, like the following example.
Same product, same design, same discount. Even though the discount is exactly the same, one might attract more traffic or sales than the other.
A/B Test #2 | The Promotion
Take different photos of your products, place them in the same layout and see which one converts better.
If you have a t-shirt store, try a model wearing the t-shirt and also only the t-shirt. If you sell stationery products, focus not only on taking photos of the products but also creating a scenario where it might be displayed.
It’s up to you and what you sell, but those small changes might result in a completely different conversion rate.
A/B Test #3 | The Background
This is a fairly easy A/B test. Try modifying the background color of one of your pages to see how people react to it. Analyse how the page performs, customer retention and sales.
Keep in mind that this last test be done on basically anything you want: you don’t have to change the background of the whole page, but just the color of a button, or the color tone of your content. Work with what you have in a way that does not take long to be implemented.
A/B tests can be performed on literally everything you want to be improved. Find a way to make those changes fast and keep an eye out for the results. If they perform better than the previous one, keep it that way and improve other sections of your website using a new A/B test.
You can do the analysis yourself by just checking conversion rates, time on site and every data you have in place. If you need some help setting this up and analysing your results, we suggest in another post the tools Optimizely and Klaviyo.
Check out the whole post: 9 best ecommerce tools for small business
A/B tests are an effective strategy to understand what will attract the eyes of your customers.
Taking the time to perform various A/B tests can provide you data that you can use for years. If one word performs better than the others, you will start focusing more on this vocabulary to improve your whole website. If a button attracts more clicks when it’s blue and not pink, it’s time to make some changes and try the same strategy in another pages as well.
If you’re a small or medium business, take advantage of this by being more flexible with easy-to-implement changes. The faster you move, the better you will know your customers and the better your ROI will have.