For Business

Find your marketing sweet spot with A/B testing

The dos and don’ts of split testing.


When it comes to growing any ecommerce business, the formula for finding success invariably includes iterative, ongoing A/B testing. Once you have the basic building blocks of your ecommerce growth strategy down, one of the more direct paths to achieving effective optimization and boosting conversions is experimentation.

A/B testing helps you collect objective data that you can use to make informed decisions about growing your business. Specifically, iterative testing provides insights to find optimal messaging, image selection, layouts, fonts, colors, checkout flow, calls to action (CTAs), and more—all of which can impact the ROI of your marketing programs. And with insights from A/B tests, you can identify new ways to drive shoppers to your site and turn them into your customers.

For ecommerce businesses big and small, A/B testing can also help determine what doesn’t work.

“Amazon itself does an extensive amount of A/B testing,” says Adrian Bell, Amazon’s Principal Product Manager overseeing testing for Buy with Prime. “Everything gets tested, and each year, Amazon saves billions of dollars in potential losses by not releasing features that didn’t work as anticipated.”

The scope of A/B testing can grow exponentially depending on the scale and complexity of the business and what information you’re seeking to uncover. But most small and midsize businesses simply don’t have the time nor the resources to test every incremental design tweak or marketing decision.

So, what marketing, advertising, and user experience (UX) elements should you prioritize for testing? Check out these tips for using A/B testing to find your marketing sweet spot.

Get picture perfect

Whether you use BigCommerce to host your ecommerce site, MailChimp or HubSpot for your email marketing, or Meta and TikTok for your social media campaigns, most major outlets now offer integrated A/B testing functionality. (Amazon sellers can run split tests on their brand’s listings content from Seller Central.)

The value you get from A/B testing on any platform depends on how you choose to take advantage of the available testing tools.

Testing your product imagery across channels, particularly on your site and social media channels, is a simple way to gain meaningful insights into what resonates with your customers.

“In any A/B test, to get a measure of the impact, you have to have an easily perceptible change between the control and the treatment groups, and imagery is one of the better ways to do that,” says Adrian. “For many brands, it often comes down to a test run between lifestyle imagery versus pure product imagery to see which version drives higher shopper conversion.”

While testing copy tends to be a bit more complicated, costly, and time consuming, the basic process of testing imagery requires fewer iterations, particularly if you already have a set of lifestyle and product images to compare.


“We look at the imagery of the social ads and then the click-through rates and the return on ad spend to understand what’s working and what’s not,” Adrian says. “With imagery, a lot of it is contextual. Some products lend themselves to lifestyle photos, and some perform better as stand-alone product images—and that’s why you test.”

“A/B testing product imagery has been big for us,” explains Logan Dunn, head of ecommerce at Wyze. “For example, we’ve found that putting a phone next to an Internet of Things device leads to much stronger conversion because people want to see the mobile app, which is also a part of the product.”

Tweak your CTAs

Regardless of the medium, platform, or marketing channel, continually A/B testing the effectiveness of your CTAs can offer outsized rewards.

“In terms of priority, the CTA is always number one,” Logan says. “While we have found that colors are somewhat effective for testing, placement and messaging are much more important.”

Whether you’re offering free shipping, a deal on a particular product, or any other kind of promotion, the language you use to entice people to make a purchase, along with the CTA’s placement and design, can be crucial. “Testing and iterating on the wording around taking the necessary action, whether it’s clicking a button or clicking the image to get to the next stage, is something that should always be tested,” Adrian explains.


That applies to everything from CTAs on your product pages to the emails and SMS messages you send out to the social media ads you deploy.

“Determine the specific action that you want shoppers to take and then create the test around that,” Logan adds. “If you want add-to-cart clicks, make sure you’re tracking the times that button gets pressed.”

Harness the power of badging

When it comes to social ads, testing the logos and badges you include on imagery can deliver real results. There are many ways to use A/B testing to boost conversions, and adding the Buy with Prime badge to your social advertising is one tool at your disposal. From optimizing your site layout to adding collection pages and taking advantage of marketing integrations, testing can provide insights to help you make more informed marketing decisions.

Social Ads for Buy with Prime syncs your Buy with Prime catalog with your Meta catalog to automatically generate relevant social ads featuring the Buy with Prime badge. Testing gives you the opportunity to determine how adding the Buy with Prime badge to your social ads impacts engagement and traffic to your site.

“For social ads that use the Buy with Prime badge set over a standard product image there was a big increase in return on ad spend and click-through rates when tested against to the same product images without the Buy with Prime badge,” says Adrian.

Wyze had already seen a 25% spike in the conversion rate since adding Buy with Prime to its site. The company used iterative testing to determine the impact the Buy with Prime badge had on engagement with social ads. The results offered promising signals for customer acquisition.

“We saw that Buy with Prime-branded assets help a lot with prospecting,” Logan says. “Someone may not know your brand, but there’s a high chance they know the Prime brand—and that builds immediate confidence.”

Just start testing

It’s easy to get overwhelmed by what’s possible with A/B testing, but don’t let that stop you from digging in.

“Our biggest issue is that you can test everything,” says Logan of his experience with testing at Wyze, “so prioritizing the based on expected return is important.”

Once you’ve used A/B testing to optimize the basics of your UX, layout, and checkout flow—meaning everything from button sizes and placement to fonts—start testing some of the marketing fundamentals.

As you start to formulate your testing strategy, here are some questions to consider:

  • What subject or opening line will get more people to open an email or SMS?
  • After they open the email or SMS, what CTA or color will get them to click through to your site?
  • What product images lead to more conversions on pages on your site and across your marketing channels?
  • Can you change the language in a headline or product description to get more shoppers to make a purchase?

Eventually, you can explore more complex A/B testing tactics, like personalization and pricing variations. But as you start your testing journey, it always pays to experiment with simple marketing tests that are easy to implement to understand how to move the needle. And remember, you don’t have to be a big spender on ad platforms to get results.

“I think testing your own site is the best return on your time and testing resources,” Logan says. “That’s where you pay the least for the most effect.”

Whether it’s a social media channel, email newsletter, or a new ecommerce solution, if the platform you’re using offers it, get to testing.

Logan adds, “We’ve found that while we have an inherent desire to do things 100% right out of the gate, we always strive to tap into the best A/B testing tools to determine if we were right at the beginning. Do a scrappy test and see what happens. If it doesn’t work, refine it and do something else. It’s easy to feel satisfied after a single test, so we try to fight that urge and keep testing. Always have some sort of a test running.”

As a Buy with Prime merchant, it’s easy to follow Logan’s advice with an A/B testing solution from Buy with Prime. Uncover critical insights that can help you make key business decisions by setting up A/B testing for Buy with Prime on your ecommerce site. Reach out to your Buy with Prime sales rep to get started.

Learn more about Buy with Prime.

Kelby Johnson