Stop Guessing, Start Testing Your Content For A Successful WordPress Site
Stop Guessing, Start Testing Your Content For A Successful WordPress Site
There are two approaches to making sure that your content is helping your WordPress site achieve its goals.
The first is to guess what works and keep your fingers crossed. The second is to test versions of titles, content, calls-to-action and see what what gets your audience to behave the way you want them to.
If you are serious about your WordPress site’s success then you need to stop guessing and start testing your way to better engagement on your WordPress site.
Online testing is usually referred to as A/B, split or multivariate testing. The definitions of each term are so blurred that they are often interchangeable but what it all boils down to is randomly (usually) displaying different versions of the same content to a visitor, tracking behavior to assess if one version is more effective than another and then, and most importantly, working out why.
How Should You Test?
The reason for testing is to find out what significantly influences visitor behavior so that you can repeat this across your site. It’s worth stopping and thinking about that because it needs to be at the core of all your testing. This is not simply an exercise in discovering whether one piece of content is more effective than another, it’s about working out why so that you can incorporate this into all your content.
For that reason, you should stick to the either-or approach of A/B testing. This will make your testing far more effective as it helps resist the urge to try endless combinations, makes it much easier to interpret results, allows results to be incorporated into subsequent tests (when running multiple A/B tests) and makes it easier to pinpoint the causes for any success.
To be able to do this you’ll need to:
- Ensure that you run your tests for as long as it takes to be satisfied that you can call a winner
- Have a clearly definable goal to be able to track success
- Not have too many differences in the content so that contributing factors to success can be easily identified
The 3rd is critical. If you run an A/B test with two completely different versions of content (different images, different copy, different call-to-action) then you won’t be able to say with any certainty why one version was more successful than other, you’ll just know that one package performed better.
The more you can target the test the better. So, running an A/B test on just the text on a button is always going to be more useful than running a test using 2 completely different versions of a landing page.
That said, if you do want to test a landing page then you can start with radically different versions if you approach it as a series of A/B tests:
- Start with the 2 very different versions and keep them both running until you have enough data that you can confidently call a winner.
- Replace the loser with a tweaked version of the winner
- Keep replacing the loser with a tweaked winner until you don’t see any significant difference in the results
Useful Plugins For Your Testing Toolkit
If you work your way through the myriad of A/B testing plugins on the WordPress repository, you’ll find that many integrate your WordPress site with a 3rd-party service. That level of integration, and therefore how much of the testing can be controlled from the WordPress Admin interface, varies considerably and is very much a personal preference.
For that reason, I’ve stuck with standalone plugins that work entirely within your WordPress installation:
Generally the standalones offer a simpler set of functionality and analysis, you can always go for a service, very few of which are free, once you are comfortable with how to test.
Title Experiments Free
Titles are important. If your titles are not engaging, if they don’t pique your visitor’s attention, if they don’t get clicked or tapped then it doesn’t matter how good your copy is or how great that final call-to-action is because your visitors are not going to see it.
Title Experiments Free is, not surprisingly, focussed on allowing you to test multiple versions of your post titles to see which generates those valuable clickthrus, all via the post edit screen.
The included analytics are basic but entirely adequate for working out what’s working with your audience and, in fact, the plugin will start skewing the display of the post title towards the most successful variant, including in your RSS feeds.
To find out more about how to use Title Experiments Free, read this recent post about testing titles with the Title Experiments Free plugin.
WordPress Calls To Action
The WordPress Calls To Action plugin is a solid plugin that not only provides templates to kickstart your call-to-action design but also includes a form builder.
1.6 million WordPress Superheroes read and trust our blog. Join them and get daily posts delivered to your inbox - free!
The plugin has some idiosyncrasies that can trip you up – for example, variants are always created with a blank template rather than the template of the original and so you’ll need to use the Choose Another Template option to align them – but once you get around these, building multivariate calls-to-action is very easy.
The call-to-action components are inserted using shortcodes which means that anywhere you can use a shortcode you can use WordPress Calls To Action making this plugin a great choice for widgets, popups, in-page promotions and, in conjunction with the built-in form builder, data collection.
The provided analytics cover all the basics allowing you to quickly and easily make decisions about the individual and combined performance of your variants.
A great tool for any WordPress user who wants to get into the world of A/B testing.
Simple Page Tester
This plugin takes an entirely different approach to Call To Action by providing split testing (actually A/B testing) at the page level.
Using existing content to build a test, the plugin places a metabox on the post edit screen to create a test using the content as the master page. Then it’s simply a matter of specifying an alternative post or page as the variant (the plugin allows you to select an existing post, duplicate the master – which is a nice touch, or create a new post).
When the a visitor navigates to the master URL, the plugin will either display the master or redirect the user to the variant. As the variant is delivered in its own right visitor behavior can be tracked via a standard site analytics service such as Google Analytics (where the page could form part of the goal funnel).
Simple Page Tester is best suited when you want to test an entire page with landing pages immediately springing to mind, especially in conjunction with WordPress’ page templates feature, expanding the testing possibility beyond just content to the entire look and feel of a page. But there’s plenty of other opportunities as well including, for example, making use of the static home page feature to test different homepage layouts.
The plugin keeps tabs on how many times each page is displayed and allows you to control the view ratio between the master and the variant, although why you’d want to test with anything other than a 50/50 ratio, I’m not sure. However, to get a real insight into the success of the pages you’ll need to use the plugin in conjunction with an analytics service unless you upgrade to the premium version.
Content Developers Need A Testing Culture, Too
No WordPress developer or implementor would ever launch or perform a major upgrade on a site without first testing that its going to function as expected. Testing code is an absolute given.
Testing content, though, is not on the radar of most site owners. Not until, that is, the site is deemed to be underperforming and someone is charged with improving performance. Usually such a task is approached as a traffic problem rather than a content problem.
Yet, considering that in the majority of cases, a WordPress site’s content is created using little more than intuition to determine the audience’s needs, it’s hardly surprising that the content will often be ineffective in influencing visitors.
Just as code developers (should) have a culture of code and test, so should content developers. It’s ludicrous to think that we can correctly define our audience needs and wants or even that those needs and wants don’t change over time.
The best way to ensure that content is helping the site achieve its goals is to experiment and test content just as developers test and experiment with code.
The key is to have clearly defined and trackable goals. Then develop a plan. Perhaps it just covers landing pages, or traditional call-to-action components like popups and widget based adverts. Perhaps you just want to more of your home page visitors to click on a post title.
Whatever it is, you need to take action and you need to test your content. In fact, there’s an excellent argument for continually testing content, if only to make sure that your current approach is maximizing its influence.
Do you test your content? If so, what do you test and how? Let us know in the comments below.