Why guess when you can test, right?
Working with WordPress may have already had a contribution to developing a clear sense of what’s working and what’s not. And that’s great.
But, instead of relying solely on your gut feeling, you could simply go for the WordPress A/B testing solution. I mean, think of this way: if used the right way, A/B (or split) testing can take the guesswork and uncertainty out of the equation and empowers you to make the best possible decisions for the growth of your websites. Sounds like something you would definitely want to have, doesn’t it?
Ok then, let’s dig into it and find out how exactly we do this A/B testing for WordPress.
How to do A/B testing on WordPress – Best Practices
Just like the name implies, an A/B test has to do with putting two options (A and B) against each other to see which one of the two options is the best option.
If the testing involves websites, then this generally has to do with checking two versions of the same page with a key difference. This key difference can be:
- a variation of a call to action (CTA)
- a different headline
- an alternate color scheme.
And these are all valid types of A/B tests.
Now, as your next steps you’d have to:
Step.1 Select your two versions
Step 2. State a hypothesis (I’ll get back to this)
Step 3. Choose a goal
For example, if you want to test two calls to action, at the end whichever one accumulated the most clicks is clearly better from a conversion standpoint. Right? So, that would be our hypothesis, and obtaining the most clicks would basically be our goal.
Now, let me just give you a definition of what’s a hypothesis in A/B testing and understand a little bit more about what’s the point of it and how it works.
What is a hypothesis in A/B testing?
A hypothesis is nothing more than a prediction you create prior to running an experiment.
A hypothesis is a clear statement, not an open-ended question.
Therefore a hypothesis will clearly state the following:
- what is being changed
- what you believe the outcome will be
- why you think that’s the case
The bottom line is that running the experiment will either prove or disprove your hypothesis. And, before you start an A/B test, you always perform a hypothesis.
Why is a hypothesis so important?
Because it helps you to find an answer to the following question: “What am I hoping to learn (discover) from this experiment?”
A test hypothesis consists of two elements:
A proposed solution
A hypothesis is not a fact. Therefore it shouldn’t be argued as right or wrong until it is eventually tested and proven.
Also, when you are thinking about the solution you want to implement, you need to think about the needs of the customer. Then make sure to think about the solution that can address the problem the customer is facing.
If your customers are confused, for instance, you need to explain the issue they are confused about in a better way or provide them with more information that would bring more clarity.
An anticipated result, the solution will facilitate
Make sure to reduce your customers’ confusion by adding the element that would create a clearer understanding of what they have to do on the website. By adding this helpful element you’ll provide your customers with the solution they need – a simplified action process on the website.
Which means you’ll have the result you anticipate you’ll have. And, remember, your proposed solution and your KPI need to be directly correlated.
Additionally, besides these two main elements, a hypothesis consists of, there something you shouldn’t overlook: the problem.
After completing your analysis and research, make sure to identify the problem that you will address in your testing.
Here’s an example of a good hypothesis sample:
Image via MarketingExperiments
The bottom line: your hypothesis should follow a structure of: “If I change this, it will have this effect“. And, it should always be informed by an analysis of the problems and rooted in the solution you consider to be appropriate in that specific case.
A/B testing best practices
Here’s what you need to look out for in your A/B testing:
- Ensure that your variations are different enough to avoid confusion
- Decide on a clear hypothesis and goal for your test before starting
- Try to select your pool of testers randomly
- Allow your test to run until you accumulate a statistically significant result
- Make sure you don’t show both variations to the same user
The truth is that any WordPress user is frequently faced with certain decisions that could be A/B tested – therefore internalizing these steps is extremely important. And if you come up the right hypotheses, a well run split test will tell you exactly how you need to continue.
Note: If you decide to simply make assumptions about your target audience, you’ll sooner or later get it wrong. Why? Because human behavior is difficult to understand without hard data – which means you need A/B testing that will help you produce the data on which you can rely.
Here’s a great example of A/B testing: Humana
Humana, an insurance carrier, created an A/B test with amazing results. The company had a banner with a simple headline and CTA as well as an image.
Through A/B testing, the company came to the conclusion that they needed to change a few details in order to make things to make the banner more effective.
Their goal was to increase its click-through rate on the above-described banner.
Image via The Daily Egg
If you look at the first banner, it had a lot of text. Although it was a pretty good banner, with good copy, it wasn’t giving Humana the results they wanted.
In the second version of the banner, they reduced the copy significantly. Also, they changed the CTA from “Shop Medicare Plans” to “Get Started Now.” They made a couple of other changes as well, such as the image and color scheme, they rounded out the differences between the control and eventual winner.
The result? 433% increase in CTR.
How? By simply cleaning up the copy and changing the picture. And, after changing the CTA text as well, the company enjoyed a further 192% boost.
Check out more cool A/B testing examples, here.
What tools can you use for A/B testing
Image via Optimize
Google Optimize is a really great re-developed, very easy to use and very powerful tool from Google – and it is integrated with Google Analytics. And the best part is that it is free for up to 5 simultaneous experiments.
Image via Optimizely
Optimizely is one of the industry’s leading CRO platforms and it provides way more than A/B testing. This means you shouldn’t outgrow the platform as your conversion optimization needs progress.
Optimizely offers a range of products, however, it’s actually Optimizely Web which provides the A/B testing and other features for optimizing web pages. Additionally, you can expand on this with Optimizely X Personalisation that allows you to deliver personalized messages to different target audiences.
Tools & plugins for A/B testing on WordPress
You can implement your own A/B tests in WordPress using two simple plugins. And the great thing about these tools is that they also offer free and premium versions.
Nelio A/B Testing
Nelio A/B Testing may actually be the most powerful and versatile conversion optimization service for WordPress.
Here are some of its characteristics: Nelio A/B Testing –
- helps you define, manage, and keep track of your A/B-testing experiments
- is compatible with WooCommerce
- enables you to run almost any type of test you want, including entire pages, headlines, theme variations, widgets, and more.
Image via Nelio AB Testing
The plugin’s only downside is its view-based pricing system – the free version will enable you to display your tests to up to 1,000 viewers (which you can increase by referring friends). Even the premium version offers impractically small view limits for sites that receive lots of traffic, thus forcing them into purchasing higher tiers.
However, as a conclusion, you should definitely consider Nelio AB Testing – obviously if you don’t mind paying for a premium service because its free plan is really limited. But, on the bright side, it will help you to get familiarize with the inner workings of the plugin which can be useful when trying to make a decision.
Simple Page Tester
Image via WordPress
Simple Page Tester can easily handle complex A/B tests as well.
In fact, its free version:
- allows you to A/B test entire posts and pages
- allows you to make any changes you desire using the WordPress editor
- doesn’t limit you in any way when it comes to the number of views you can get – the plugin allows you to run as many tests as you want at once
- enables website owners to run A/B Tests in WordPress without having to edit the code as required by so many of the split testing services out there
As a conclusion, Simple Page Tester is a very easy to use, free solution.
Image via WordPress
Title Experiments is a very basic WordPress plugin, however, it’s really great at testing your titles for free and discover which ones get a higher click-through rate.
There is also a premium version available that allows you to test multiple features, images, detailed statistics, and even automatic experiment freezing. So, it sounds like a great tool you can use in your A/B testing.
WordPress Calls to Action
Image via WordPress
What’s great about the WordPress Calls to Action plugin is that it allows you to test your:
- social media buttons
- file downloads
- and set up call-to-action templates.
And, this free plugin will also help you in tracking your conversion rates and run multivariate tests.
What can you A/B test on WordPress
If you are interested to test individual elements on a web page, A/B testing is the best method as it allows you to check each specific page element. It also allows you to determine which element serves your readers the best and converts more traffic to sales.
The great thing about A/B testing is that (while it can sometimes be a long process), it gives you the most accurate and detailed results for the elements you want to test.
Here’s what you can split test on your WordPress site:
- Footers and sidebars
- Color schemes
- Page titles
- Graphic elements
- Page layout
- Calls to action (CTAs)
- Themes or landing pages
- Opt-in forms
Measurement and Analysis
The final part of the A/B test is the measurement and analysis. This is often a quite neglected part of the test and therefore it turns into a major source of later analysis issues. So, it’s important that any hypothesis has a good measurement strategy in order to actually be useful.
Here’s what I mean: A/B testing isn’t just about comparing CTR A with CTR B. What you also need to focus on is to analyze the source of the traffic as well. Make sure you know whether the traffic came. Is it a desktop? Is it mobile?
When Google is doing tests, they are taking things way further, analyzing OS’es, gender, age, location, browsers etc. Ideally, all the properties of the environment should stay the same. But because ideal doesn’t happen in real life, we should do our best to be as close as possible.
As I’ve said, the traffic sources are extremely significant as well when you are running a campaign for instance. Are you running a Facebook campaign or a Google Ads one?
Clearly, you won’t do this testing for every campaign. You would only pick an ad and set up two different landing pages and then rotate evenly. And then depending on the quality of the campaign, you could actually attract more or less qualified users on the website, and they would act accordingly.
The point is if you overlook these details in the A/B testing process, you could end up with the wrong conclusion. In the end, the proper analysis and measurements will bring the most value and will help you develop more hypotheses.
A/B testing is one of the most effective ways to gather information about your copywriting and design choices. The reason A/B testing is valuable is that different audiences have different behavior that you’ll eventually need to discover.
However, A/B tests can also be very complex. And if you’re not careful, you can end up making the wrong assumptions about what people like and what keeps them engaged.
So, it’s extremely important to follow the right methods in A/B testing in order to achieve the accurate results you want.
Here are some steps of the A/B Testing process you need to focus on:
- Decide on what you want to test
- Set your goals
- Analyze the data
- Select the page that you want to test
- Set the elements to A/B test
- Create a variant
- Pick an A/B testing tool
- Design your test
- Gather your data
- Analyze the A/B testing statistics
An A/B test will always help you discover a new way to make your marketing efforts more effective – but testing is a never-ending task. Why? Because there’s always room for more optimization. So, go ahead plan your next A/B test and improve your conversion rates as well as boost your bottom line.
Remember, the possibilities are endless when it comes to A/B testing. So test, refine, then test again.