You’ve spent a lot of time and effort finding the right design agency, but the question remains:
How do you know your website design is effective? Wouldn’t it be nice to move past hunches and assumptions, and to look at things more methodically? That is exactly what A/B testing can do for you. Embracing it allows you to use data to track the impact of individual design elements, as well as systematically improve the overall design. Effective website design is part art and part science.
You have the art component covered by hiring a designer who understands principles of composition, typography, and aesthetics. A/B testing offers a scientific approach to conduct experiments, track your results, and improve conversions.
Keep reading to find out what A/B testing is… and why it’s a key element in an effective design.
What is A/B Testing?
A/B testing is a process in which you test two variations of your website design simultaneously to determine which is more effective. This process helps you validate design elements and determine which ones affect your conversion rate.
Each A/B test begins when the website owner chooses one design element. Half of the visitors see the original version of the website (also called the “control”) while the other half see a version (called the “treatment” or “challenger”) with the element slightly altered.
The metric of success is ultimately up to you. Typically, it is a marker related to conversions – like sales, sign-ups to an email list, or bounce rate (the percentage of visitors who click away from the site after viewing only one page).
When your test runs long enough, you eventually generate statistically significant results – those that cannot be explained by chance. The version that performs best becomes the new “control.” Then, the process repeats itself with another design element.
Why Should You A/B Test?
Running A/B tests helps you challenge your assumptions about what makes your website design work.
The vast majority of businesses want more than just offering visitors a pleasing user experience. They need a design which ultimately results in action (service enquiries, sales, etc.). A/B testing leads you to the best combination of design elements to reach those conversion goals.
A/B testing also makes your website design even more effective over time. Good designs become great designs when they are optimised to foster the highest engagement and response. It’s a cost-effective way to do business because you get more “mileage” (in the form of email sign-ups, sales, etc.) from the same amount of traffic. This makes marketing initiatives like SEO, content marketing, and paid traffic more profitable than before.
A/B testing also reveals new insights about your target audience. You’ll learn which design elements resonate the most (if, for instance, they prefer longer or shorter landing pages). This increased understanding fosters more design ideas, as well as ideas to better serve your customers.
Finally, A/B testing is a definitive process to test theories and resolve disputes within the company about how to best handle the design. Everyone gets the chance to offer his or her input, and the results generated from tests provide clear guidance how to move forward.
Which Elements Can You Test?
One of the most challenging aspects of getting started with A/B testing is figuring out where to start. The typical business website is loaded with different design elements, all of which can be validated through testing.
This process need not be intimidating, however. The idea is to tweak just one element at a time to analyse its overall impact on conversions. It’s a much simpler process than multivariate testing, where you test multiple elements at the same time to see which play a key role in meeting your website goals.
Here are just a few design elements which influence visitors’ behaviour:
- Call to action – the specific action you are trying to get visitors to take (text and buttons)
- Copy – written content designed to drive visitor response
- Forms – contact forms, customer support forms, sign-ups for email lists, etc.
- Layout/positioning of design elements
- Length of page
- Navigation structure
- Social proof – testimonials, case studies, etc.
- Value proposition – a concise statement of what your business does, whom it serves, and how it does it better than competitors
The list might seem overwhelming at first glance, but you can chip away at it one test at a time. Majestic Wines, a UK wine retailer, systematically worked its way through the elements above to reduce clutter, emphasise their value proposition, and ultimately add a short video on their homepage.
Conversions increased by 201 percent (via online form submissions).
How to Run Your First A/B Test
Before you can run an A/B test, it’s time to choose which design element to focus on first. It’s easiest for first-time A/B testers to focus on “low-hanging fruit,” design elements that the vast majority of website visitors 1) see, and 2) engage with. Tweaking these elements typically has the most significant impact on conversions.
First, identify which web pages receive the most traffic. This will usually be your homepage, along with a few key landing pages related to specific products and/or services. From there, choose an element that almost every visitor engages with – things like headlines and other “ above the fold ” content.
Once you’ve settled on an element, research which A/B testing tool would suit you best. There is a large variety of these tools available, and their prices range just as much as their functionality. Many tools allow you to start at a lower price tier and upgrade (if you want) at a later time.
Once you start a test, keep it running until you are statistically confident in your results. The time the test will take depends on how much traffic your website receives. Ideally, you want statistical significance to be over 95 percent to ensure sure your results are valid and not based on chance. A good testing tool will tell you when you’ve gathered enough data to reach that conclusion.
Don’t run tests longer than necessary. Once you have a clear winner, move on to another high-priority design element. If you’re short on testing ideas, the scientific method offers a helpful framework:
- Ask a question (e.g., “Why are so few people joining our email list?”)
- Form a hypothesis (“We are doing a poor job communicating what visitors get from signing
- Test your hypothesis (Tweak the sign-up form copy to make it clearer that subscribers will
receive a valuable free eBook)
- Analyse your data and draw a conclusion (See which design variation converts better)