a person sitting at a table

At 9:00 AM EST on January 1, 2022—1,916,978,813 websites were online. That’s two billion ecommerce companies, online stores and other commercial sites vying for people’s attention. How can your business stand out with so much presence on the internet? Website A/B testing can help.

Many businesses use website A/B testing tools for conversion rate optimization. Not running these tests could mean leaving money on the table.

The number of websites online as of January 1, 2022, at 9:00 AM EST.

image source

Key takeaways:

  • A/B testing is an experiment where website visitors see one of two (or more) versions of a webpage or a page component. Then, a business analyzes which version drives more of the desired results.

  • Website A/B testing maximizes potential business revenue.

  • The steps in A/B testing are: determine the goal, make a hypothesis, create the variable, run the A/B test and measure results.

  • While it’s acceptable to perform tests on multiple variations (multivariate testing), experimenting with too many elements at once makes pinpointing the successful variation challenging.

  • Elements to test are the call-to-action, blog and webpage intros, shipping cost placement and menu items.

What is website A/B testing?

A/B testing is an experiment where website visitors see one of two (or more) versions of an element. The element could be a webpage, a page component (like an image or text box) or an email. Then, a business analyzes which version drives more of the desired results, such as increased engagement, signups or high-quality leads. Other terms for A/B testing are split testing, bucket testing and multivariate testing.

The original element is the control, and the proposed change is the variation. What changes can we test? Anything you think would appeal to customers? Here are some possible variations:

  • An image

  • location of the image

  • size of the image

  • color of a webpage

  • color or size of the first letter on a webpage

  • wording for a call-to-action

Let’s use a simple example. An ecommerce business uses gray “Buy Now” CTA buttons. Could a simple button color change increase sales?

The company performs an A/B test on CTA button color. Fifty percent of website visitors see the original gray “Buy Now” buttons, while 50% see the red “Buy Now” buttons. The company notices that the red buttons yield 4.5% more sales than the gray ones during the experiment. At the end of the testing, the website switches to red “Buy Now” buttons.

A/B testing software takes multiple factors into account when it comes to effectively running A/B tests: sample size (the number of visitors), statistical significance, measuring heatmaps, ensuring there aren’t too many different variations of landing pages that confuse the process, accurately recording test results to improve the customer experience, etc.

Website A/B testing where the variation is a color change on a button

image source

Why website A/B testing is vital

How do businesses benefit from A/B testing? Here are some real-world statistics to cement the advantages of website experimentation:

  • Workzone saw a 34% increase in form submissions by A/B testing their customer testimonials area

  • A/B testing increased China Expat Health’s lead conversion rate by 79%

  • Discovery Communications A/B tested components of their video player, resulting in a 6% increase in video engagement

  • ComScore performed A/B testing on logos and testimonials on a product landing page, which led to a 69% increase in leads

  • Secret Escapes doubled conversion rates by testing variations of their mobile onboarding pages

Split testing can improve user engagement, reduce bounce rates and increase conversion rates. A/B testing improves click-through rates (CTR), maximizes your ability to convert, streamlines the conversion rate optimization (CRO) process, enhances user experience and leads to increases in revenue.

Video: A/B Testing Intro: Why, What, Where, & How to A/B Test

Video Source: https://www.youtube.com/watch?v=rLCTkUnBfCw

Steps of A/B testing

Each element represents an aspect of your business where we can maximize success. For all A/B tests, make sure to:

  1. Determine the business aspect to improve

  2. Make a hypothesis regarding the experiment

  3. Create the variable

  4. Run the A/B test

  5. Measure the results

While it’s acceptable to perform tests on multiple variations (multivariate testing), experimenting with too many elements at once makes pinpointing the successful variation challenging.

Four elements to A/B test

  1. The call-to-action (CTA)

The CTA is usually an imperative sentence that urges the visitor to perform the desired action, such as subscribing to the site or purchasing a product. The call-to-action correlates with your conversion rate. Some variations for A/B testing a CTA are:

  • wording

  • font type, color or size

  • CTA location

  • anchor text

  • adding a button

  1. Blog and webpage intros

A catchy title can draw people to your site and increase website traffic. Some variations to test for blog and webpage titles are:

  • asking a question: “Should You Perform Website A/B Testing?"

  • incorporating a number: “Four Website Elements You Should Be Testing This Year”

  • using hyperbole: “The Ultimate Guide to Website A/B Testing”

  • using the first-person: “How Website A/B Testing Helped My Business”)

Another characteristic of blog and website are images. You can also try testing two different post images to see which drives more website traffic.

  1. Shipping cost placement

Today, seven out of ten shoppers abandon their shopping carts. The placement of the shipping cost may lead to a decrease in shopping cart abandonment. Could the customer provide their location at the beginning of their product search?

For example, you could ask for the zip code or use the current location. Then, factor in the shipping cost when displaying the product price. Could the request for the shipping address appear earlier in the checkout process? 

  1. Menu items

Your homepage layout and number of menu items can affect your website bounce rate. Too many menu items may overwhelm potential customers, and they may leave your site. Some suggested variations for A/B testing are:

  • side menu bar

  • top menu bar

  • seven or fewer parent menu items

  • sub-menu items

  • no sub-menu items

  • menu bar color

  • the font color, type and size

List of website elements and corresponding business metric

Get started with Optimizely website A/B testing

When you want to improve your business revenue, turn to Optimizely. Our software solutions and a team of experts can assist with sophisticated website A/B testing. You can run tests to track user behavior that will provide insights into website optimization including webpage title, image and CTA variations. Immediately integrate results into your production environment in real-time. You’ll optimize the conversion rate and receive a greater return on investment for your digital marketing campaign.

Contact Optimizely today to discover how A/B testing can increase your business revenue.