Today, we’re going to break down the “Top 7” split tests you should perform on your website first. Hopefully, this guide will give you a concrete, actionable strategy for getting started with split testing on your site.
Test #1 – Headings & Sub-headings
There’s an often-cited statistic that claims that the average website visitor spends no more than seven seconds on your website before deciding whether to stay and look through your content or hit the back button and look for a different option (although recent research shows that this “window of opportunity” may be even smaller).
Because of these short attention spans, one of the first split tests you should carry out involves the specific wording of any headings and sub-headings you use on your site, as changes made to these elements can make a big difference in terms of getting visitors over this seven-second hump. Consider testing your current headline against one of the many headline formulas out there to see how effective and engaging you can make it.
Test #2 – Images
Although most graphics are added as an afterthought based on what “looks best”, the truth is that images can have a dramatic impact on the effectiveness of various areas on your site.
To see how much of a difference images can make, consider the following case study, carried out by Jason Thompson of EmptyMind.org. Jason was interested in finding out whether a generic icon or a personal photo would lead more people to click through to his “Contact Me” page, and the results of his A/B split test were astonishing. Swapping out his plain old telephone icon for a closely cropped personal image led to a 48% increase in his clickthrough rate (CTR):
To test how effective the images on your website are, consider split testing images in any of the following places:
- In your graphic header
- Embedded in your page background
- In your “About Me” featured section
- In any buttons or icons that lead to personal connection opportunities
Test #3 – Calls to Action
In addition to the subject and placement of your images, consider split testing your calls to action. A call to action occurs any time you ask people to complete some behavior, whether that’s clicking through to read your “About Me” page or purchasing a specific product from within your store.
Website readers are surprisingly responsive to different wording in these areas, so you might be surprised to find that changing your opt-in subscription button from reading “Click Now to Subscribe” to “Claim Your Free Reward Today” could result in a significant improvement in the number of opt-ins you receive. Remember, any time you’re able to increase the percentage of visitors who take action on your promotions or offers, you’ll likely see a corresponding increase in your overall profits!
Test #4 – Offers
Similarly, the way you word and structure your offers can play a big role in how effective these items are – but you won’t know for sure until you split test them!
As an example, consider a case study of mobile phone retailer, Mobal, carried out by the staff at Visual Website Optimizer. The company wanted to determine whether offering three or four phones on their international cell phone sales page would lead to more sales. After carrying out the appropriate A/B split testing, the company was able to achieve a 27% increase in sales by including four phones in their product offerings:
Test #5 – Your Opt-in Form
If you aren’t selling products but are instead trying to encourage as many of your website visitors as possible to sign up for your email newsletter, be aware that there are plenty of different elements surrounding your opt-in form that benefit from split testing.
Consider experimenting with any of the following options in order to maximize your opt-in rate:
- The location of your opt-in form on your website
- The style of opt-in form you use (pop-up versus lightbox versus embedded)
- The pages on which you place your opt-in form
- The colors and text you use in your form
- The number of fields required to subscribe to your list
Most email subscription programs operating today include split-testing features built-in to their programs, so consult with the customer service department of your email provider to see what opportunities are available to you.
Test #6 – Linked Text
Any time you use the text within your website to link out to another page (whether within your own website or on another page entirely), you’re inviting your readers to click through and explore this new content. So why not make these invitations as appealing as possible by split testing the specific anchor text you use within your links?
As an example, suppose you want to filter visitors from your blog posts to your current “Specials” page to increase their likelihood of making purchases. Obviously, you’ll want the text within your “Click here to view our specials” link to be as effective as possible, making it a good idea to test different variations of your linked text against each other to see which option offers the best results.
Test #7 – Colors
Colors play a powerful role in creating the “mood” of a website, but this effect goes beyond controlling the way visitors feel – it can actually influence how likely they are to buy from you!
As you likely incorporate different colors into a variety of different places on your website, the potential number of split tests you could run on your site’s colors alone is quite high. Consider any of the following as possible staring places for this type of split test:
- Your background color
- The text in your header and headings
- Different colored structural elements throughout your site
- Your opt-in boxes or checkout buttons
- Call out text colors throughout your website
Remember to focus on a single element at a time when you’re first getting started with split testing. As you build more experience and become more advanced with your split testing skills, you can begin to experiment with multivariate split testing, which will allow you to compare multiple elements at the same time while still understanding the statistical significance of your results.