A/B testing (aka split testing) is so much cooler than its buzzword-y name sounds. It lets you gain full control over each element of the content you’re putting out there and helps you understand exactly how each of these pieces affect your bottom line.
Plus, you start to feel pret-ty smart once you realize you’re learning new things about the way your audience’s brains tick — stuff they probably don’t even consciously know about themselves.
What Is A/B Testing?
At its core, A/B testing is just pitting two different things against each other to see which one comes out on top.
Have you ever gotten into an argument with a friend about which route is fastest to get from your house to theirs? (Los Angeles readers, I know you’ve heard this one before!) How’d you settle that bet? You tested it! You both left the same place at the exact same time, went your separate routes, and found out once and for all whose way is the best, right?
A/B testing on the web is basically the same thing. Any time there are two different routes you could take — two versions of a sales page you can’t decide between, maybe? — you can settle the argument with a quick and simple test.
How? Check out this short and sweet video explanation from Hutchinson Web Design:
Why Should I A/B Test?
Here at LKR, we don’t like to just sit and wonder. We like ACTION! When we’re working on some copy for a new product or figuring out where the purchase button should go, we don’t spend much time hemming and hawing over which option miiiiiight lead to more sales. We like to KNOW. And the only way to really gain that knowledge is to test, and to test often.
Not convinced that the results will be worth the effort? These off-the-charts results make for some solid arguments about why you should be A/B testing:
- Increase your click through rates: How changing a single word increased click through rate by 161%
- Use testing to grow your list: How to Quadruple Your Opt In Rate
- Increase conversions in your email campaigns: Use A/B Split Testing to Improve Your Email Marketing Campaigns
What Do I Test?
You can test anything that you’ve ever been curious about, ranging from small items like the “from” field on your promotional emails to big attention-grabbing things like the color scheme of your entire sales page.
My team and I regularly test different elements of our blog, sales pages, and landing pages. Occasionally, we can’t resist testing out a tiny tweak here and there just to see what all of you guys like better, but most of the time, we like seeing CRAZY results by testing the things that are more likely to have a bigger impact.
If you’re an A/B testing newb, start off with these simple, but heavy-hitting items:
- Headlines: We’ve found that creating an intriguing headline is THE hardest part of writing anything. So, any time we have a writing task, we make sure we write a TON of headline options. If we’re super-torn between a couple of good headlines, we always test to see what our audience will respond best to.
- Call to action buttons: Simple tests for which words work best on your purchase buttons can actually make big improvements in your conversion rate. In our recent Famous in Five challenge, we swapped out the button text on our sign-up page and saw some pretty interesting results!
- Page layouts: This example from 37signals shows how you significantly alter how people respond to your site by changing around your page layout. Here, they talked about testing huge, radically different designs on their landing pages which ended up getting HUGE results.
And when you’re done with those, move on to these intriguing test subjects next:
- We loved these suggestions from KissMetrics. Their 3 overlooked elements to test are usually some of the first tests we’ll run on a brand new product’s sales page.
- MailChimp’s guide to A/B testing reminded us that it’s not just our site we can tweak. There’s plenty to optimize and test with email campaigns, too! We’ve tried experimenting with different days/times of the week, subject lines, and swapping out the text links back to our site.
How Do I Set Up An A/B Test?
For all of our split tests, we LOVE Visual Website Optimizer. It’s easy to use and they make analyzing your results (the most important part of your test) a dead-simple task. Plus, the clickmaps that show you where customers are interacting with your pages are just so cool.
Other tools available that may fit your needs include:
- Optimizely: You can easily create a test using their point-and-click editor, and implement your results right away. No coding or fancy technical knowledge needed.
- Unbounce: Design landing pages to test using a drag-and-drop editor, plus they have a huge library of templates you can pick and choose from.
- Convert: Use a quick wizard to modify elements of your page design for an easy A/B test that integrates nicely with Google Analytics.
Here’s an easy to follow beginner’s guide to setting up an A/B test that takes you through how to get your test up and running, and how to understand the results.
If you’ve decided that you want to dip your toe into the A/B testing pool with your email marketing campaign, here’s another helpful guide that outlines the basics.
A/B Testing Case Studies
You want more proof that this stuff works? Well, you’re looking at it! Here are a few case studies of how I was able to see real results for my business with A/B testing:
- We increased signups for The Dash by 24% with just a small tweaks to our homepage.
- And most recently, my team and I split-tested the homepage of our Famous in Five Challenge. We changed 5 words and boosted sign-ups by 8.39%.
How To Become An A/B Testing Wizard
So, you’ve run a few basic tests, and now you’re ready to REALLY step up your game. Sure, the concept is simple and the process (thanks to awesome tools like VWO) can’t be easier… but there are still some basic mistakes I’ve made that you can learn from.
- DON’T end your test too early. Even after you’ve had a couple of successful conversions that are heavily skewed toward one version, don’t pull the plug! You always want to make sure that the test has been exposed to enough people to make the answer undeniable. Just because the first 10 people bought your product while looking at Version B of your sales page doesn’t mean that Version A’s the long-term loser.
- DON’T forget about your results. Yeah, it’s super important to be testing all different variables all the time. But don’t get too caught up in trying new things that you ignore what’s already worked for you in the past. Keep a record of every A/B test that you’ve run and make sure to reference it often so that you don’t waste time testing for something you already know.
- DON’T overthink it! Just because you can A/B test literally every. single. piece. of your website, doesn’t mean that you should. If you put two completely different opt-in forms up against each other, each with their own unique copy and color scheme, don’t over-complicate the test by trying to dissect whether it was the COLOR or the COPY that did it. You know which form works better now, so don’t sweat the details.
- On a similar vein, DON’T vary too many things with your test at once! At least not if you really must know EXACTLY which variable is responsible for boost (or drop… yikes!) in results. If this kind of micro-level understanding isn’t a huge concern for you (as it often isn’t for me) and you just want to know which of two options is more effective – then DO go on ahead and test two completely different options against each other. The key is to just be aware and to be realistic about the information you’ll get back and the conclusions you can legitimately draw from it.
- And lastly, DON’T let your tests dictate your vision. Just because an element of your page wins out in a split test doesn’t mean you’re stuck with it forever. Knowledge is power, but don’t let it boss you around.
These are the exact tools and resources we used when we first started out with A/B testing. And boy are we glad we did.
The next time you’re on the fence about a site change or copy tweak that a team member SWEARS will be the ticket to your sales shooting through the roof, don’t wonder if they’re right — test it out to find out for sure.