How to Measure Things That Matter When Split Testing

I love split-testing, as it’s a surefire way of putting an end to ‘guesswork’. When it comes to increasing the rate of conversion of a landing page, most of the online marketers rely on guesswork which is definitely not the best way to go. If you’re ready to devote the time and energy that it requires, split-testing the different elements of your landing page and tweaking them appropriately can immediately boost the rate of conversion, increase your online sales and elevate your business to new heights.

Most of the internet marketers continue to optimize their web pages or landing pages to enhance the rate of conversion. Rather than optimizing your pages over and over again, you should get started with split testing.

Getting Started

You might have heard about split testing or A/B testing before. To get started with it and do it the right way, you need to be aware of the basics. During a split test, you’ll be able to compare two (or more) versions of a web page so that you know how each of them actually performs.

It’s true that split testing consumes time. If you are not quite well versed with what the different elements of a web page and completely understand the analytical data available, it’s tough to conduct this test on your own. If you think you don’t have enough expertise to do it on your own, it’s always advisable to use the services of a professional.

First and foremost, you should create a list of all the essential elements that comprise your landing page. Initially, the list that you prepare may be too long. But you won’t test all of them. Instead, you’ll need to filter out only those vital elements that you think are highly impacting. At this stage, you really need to act like a critique and rank the elements on the grounds of urgency or priority.

Setting Up the Test

Once you’re ready to measure things that you should, it’s time to create different versions of the page that you want to test. If you’ve chosen to test color, content, images, headings, call to action and banners for example, you should create different versions of the page by tweaking these elements differently.

Next, you’ll need to install a script on your web server so that the different versions of the landing page are displayed to visitors randomly (or simultaneously).

If you’re working in WordPress then you’ll definitely want to check out our tool, Simple Page Tester.

Tracking the Results

Split testing will help you increase the rate of conversion only when you’re able to measure the results accurately. To get the best results from the test, you should focus on three things which are as follows.

  1. Effectiveness of the Data
  2. Size of the Sample
  3. Timespan

Unless you work with really strong data, a significant sample size and appropriate time-span, you may get skewed data (which can turn things the wrong way). Ones you’ve tested the first few elements and improved them as well, you can move on to test the other elements on the page.

Bonus Tip

Before you start to split test, you should ensure that you have one call-to-action on the landing page and that the elements that you select for the same are pertinent to that specific action.