Why A/B Testing Makes for Better Websites (and What to Be Careful Of)

The importance of a modern, functional website to online businesses in 2020 is greater than ever before.

From Gen Z’s to Gen Xers, a majority of users do their shopping online, a trend which only shows signs of growing. E-commerce’s share of total retail sales has doubled in just the last 4 years, and it’s poised to double again in the next decade.

Your website needs to be fast, secure, easy to navigate (especially on smartphones), and efficient at directing visitors toward making a purchase or conversion.

Now, you may think you have a great website, and even if it is bringing you a decent amount of business, how do you know if it’s reaching its full potential?

The answer is in analytics and reporting. The right tools can tell you where your visitors are coming from, who they are, which pages they’re visiting the most, how long they’re staying on each page, what images or buttons they’re clicking on, and when, where and why they’re leaving your website.

Looking at all of this data can tell you two main things: 1) What your customers’ priorities are; and 2) What their experience is like on your website.

Ultimately, we use analytics on a website to understand and optimize customer flow. 

For example, when we brought our analytics experts to client Carefree Dental’s website, they found that the problem was that customers weren’t ready to sign up for a membership — they were more interested in learning more about the service first. 

A relatively small tweak to the website to focus on learning more instead of pushing a sale right off the bat led to much higher website engagement numbers. The result was a 120% increase in sales.

So how did we do this? With a little something called A/B testing.

What is Website A/B Testing?

A/B testing is a practice of optimization used in many fields, from scientific research and software development to product design, moviemaking, and yes, website development.

Essentially, it’s an approach to web design and development in which you create two versions of the same page and let them loose on the world. By comparing the performance of each, you can determine which version is best for your goals.

So in the above case with Carefree Dental, the goal was to drive more sales. The interesting thing is that A/B testing of a homepage that focused on sales performed more poorly than a version that focused on learning about how the service works. 

Turns out that when people understood the product better and knew how it would help them, they were much more likely to make a purchase.

What’s great about A/B testing on a website is that you can really compare any number of variables. You can try different ad copy, different images, videos, banners, various layouts, and more. 

With online A/B testing software, you can create a lot of subtle variations of your website and test them all against each other. Ideally, each variation is the same as the next except for one difference. The software analyzes the data and shows you what’s working and what’s bombing.

But here’s the thing: A/B testing isn’t helpful if you’re not optimizing for the right thing. 

So before you head off and go A/B crazy, let’s look a little at the pros (and potential pitfalls) of A/B testing so that you can make the most of this powerful approach to website marketing.

The Perks of Website A/B Testing

1. Finds the Best Content for Your Goals

This is really the point of A/B testing. Instead of arguing over creative decisions in your next marketing meeting, you can simply try out different ideas and let the numbers do the talking. 

For example, you can test different marketing slogans to see which one has the highest click-through rate. Or you can compare image and video content to find out which is resonating the most with your audience. You can settle on the most effective call-to-action for any given landing page.

Furthermore, you can A/B test product descriptions. Is a short and sweet one the winner, or are your customers looking for more detail? Same goes for page content — does a general overview lead to more or less conversions than an in-depth explanation?  And which blog headlines are best?

You can even use A/B testing on testimonials and reviews to see which ones lead to more sales.

2. Improves SEO

This is a less obvious perk, but it really matters. When you make changes to your website, it can change the site’s load time. Faster load times lead to higher SEO ranking.

With A/B testing, you can try design variations for which one loads the fastest while still hitting your conversion goals. Getting this balance right will keep your website high in search engine results so that you have traffic coming to engage with your awesomely optimized site.

3. Improves Customer Experience

That previous pro lead to this one. No one likes a slow website. By A/B testing for speed, you not only improve SEO, you also improve your customer’s experience.

Additionally, by testing website variations to achieve the highest possible engagement and conversion rates, you’re ensuring that your site fulfills the priorities of your visitors.

They aren’t encountered with confusing imagery, poor layout, or meandering copy that has little to nothing to do with what they’re looking for. The result is a happy visitor — aka a lead who is more likely to bring you another sale.

4. Increases Marketing ROI

You’re no doubt spending money on digital marketing to bring traffic to your website. Search engine marketing, social media ads, and original content all cost money. If they do what they’re supposed to do and bring people to your website, it’s a waste if your website sucks. 

It’s like if a friend called you up and said “Hey, there’s a great rock band playing in town tonight, wanna check it out?” So you put on your studded choke collar and head down to the club, only to discover that the “rock band” is 4 guys hitting literal rocks against each other. 

Not exactly what you were expecting. Needless to say, you won’t be visiting the merch table.

With A/B testing, you can make sure your website lives up to its promise and helps lead to sales, not drive them away. A more effective website means a higher ROI on your marketing spend.

5. Lowers Bounce Rate

A visitor who doesn’t immediately see what they’re looking for on your website is gonna head for the door like asap. In website lingo, that’s called bouncing. 

You know when people say “aight, I’m gonna bounce?” That slang started with web developers. (No, it didn’t. We just made that up. It was a joke.)

Some marketers put far too much importance on bounce rate, but it can be a helpful metric if you know what you’re doing. After all, if people are leaving your site from the first page they visit, it’s unlikely that they’re able to convert.

Your bounce rate goes up when your site is slow and hard to navigate. It goes down when it has a clean, accessible design that puts the content your audience is looking for front and center.

With A/B testing, you can see which page variations have lower bounce rates and land on the perfect design to keep people on your site for longer.

6. Saves Money on Site Design

Say your website isn’t performing so hot. Before you go spending a bunch of cash on a complete overhaul, why not try A/B testing? Maybe all it needs is a new call to action, or a page reorganization to drastically improve conversion rates. You don’t know until you try.

By focusing on small factors that lead to big site improvements, you can stop wasting money on unnecessary web design and only pay for the changes you need.

The Potential Problems with  Website A/B Testing

After that pretty persuasive list of reasons to do A/B testing on your site, it’s hard to imagine any real flaws with the approach. But they do exist if you don’t set up your tests properly, and if you make the wrong decisions based on the results of A/B testing.

One risk of A/B testing is that if you don’t act on the results correctly, you could exclude a portion of your audience and ultimately drive them away.

For example, let’s say you run an A/B test on a CTA on your homepage. 75% of your visitors prefer version A, while 25% prefer version B. If you just go ahead and say, “Great, let’s go with version A,” you’re neglecting 25% of your traffic!

The good news is that A/B testing tools have a much higher threshold than 75% — they typically won’t point towards A or B unless one is appealing to 95% or more of your visitors. The takeaway? Trust the data, and only go with changes that have a very high success rate.

The problem is, if you’re a small business without a ton of traffic, it can be hard to accumulate enough data to do meaningful A/B testing.

In this case, an option for A/B-style optimization is contextual testing. Using the power of Google Analytics and UTM parameters, you can find similar attributes among visitors who prefer version B over version A. These might be the kind of device they’re viewing your site on, where they’re located, the time of day, or where they came from (an email vs. social media vs. SEM vs. external link). 

We know, in general, that version A performs better, so that will always be the default version of your site that is displayed to new visitors. But when a visitor who matches the profile of a version B fan comes to your site, they’ll see that version instead of version A. 

This is powerful stuff, but you can’t do it without data, data, and more data.

Ultimately, testing can be a very helpful arrow in your quiver, but it doesn’t have to be the only one. No matter what, you can always take steps to optimize your site for better performance and conversions. 

In the end, it all comes down to making sure you’re collecting data. Data will help you spot issues or trends — the kind of stuff that might inspire an A/B test or (if you have a smaller site) new changes. And the data is where you’ll be able to see the results and see that your change has had an impact.

Here at BKMedia Group, our analytics team relies on Google Analytics and other tracking tools to get the information they need to help our clients reach their sales goals.

Read more about the 5 Google Analytics reports that’ll make your website rock »