Wouldn’t it be great if you could be sure all your headlines, website copy, images – just about every aspect of your business website – are working for you and not against you? Well you actually can. By split testing different variables on your site, you can make sure your site is the best it can be, and converts the best it can.

What is Split Testing?

Split testing (also known as A/B testing) involves conducting controlled experiments in order to improve website metrics and conversions. It is a great way to find out what marketing strategies work best for your business and your audience.

What Should You Test?

You can split test just about any aspect of your website, like your content, and even your advertising.

However, just because you can test everything, it doesn’t necessarily mean you should. Focus your efforts on the things that will have the biggest impact on your business. Great places to start are:

• Headlines
• Calls to action
• Images to attract attention
• Sales copy
• Paid ad copy
• Email subject lines or text

Where to Start

If you have trouble discovering where start, start by looking at your own stats.

Log in to your Google Analytics dashboard (or similar website analytics dashboard) and take a look at your exit pages. Why? These are the pages where readers tend to leave the most. So it will pay off to do some tweaking to keep readers on your site.

Take a look at these pages and pick one to start. It’s best to do your testing in a controlled environment, so test one thing at a time. Pick a page and look at the different elements on the page. Could you alter the headline to see if it attracts more people to reader further? Maybe your content needs to be more enticing? Or what about images? If you have images on the page, try adding more attention-grabbing ones, changing the location of the images, or adding more (or taking away some!) You can even play around with color on your buy or call to action buttons.

There are many different aspects you can change, and it’s amazing how tweaks that seem so small can make a huge difference on email subscriptions, sales conversions, and website bounce rates.

But! It’s important to make sure your tweaks give you the results you desire. That’s why you test. Change something, measure results, and compare to previous numbers.

How Long to Split Test

Split testing isn’t going to be an overnight thing. Depending on how much traffic your site has, you should run your test anywhere from a week to a month. Avoid testing for too long, as you may end up with unreliable results.

Split testing is a great way to find out exactly what is, and what’s not working on your site. By making small changes and measuring your results, you can ensure your pages offer the most value and convert as well as possible.