Split Testing: Why It Is So Important and How To Do It

There are many wonderful things about owning an online business.

The overhead is relatively low, marketing and promotion are typically less expensive and risky than in the “real world” and you can start without a great deal of capital.

Split testBut that’s not all.

One of the greatest things about owning an online business is that you can get your hands on quantifiable data that can help you to analyze exactly what you need to do in order to increase your bottom line.

There are many ways in which you can do this, but today I want to talk about just one: split testing.

This is something that I have been aware of for a long time but avoided up until the past few weeks, but it has already made a huge (and positive) difference to sales of my information product. In this post I want to introduce you to the concept of split testing, share the results I have attained by doing it, and then finish by giving you a step-by-step guide to setting up your first split test.

What is Split Testing?

If you have never heard of split testing before you would be forgiven for assuming that it is some sort of advanced marketing technique. But whilst its application can be rather complicated (if you want it to be), the principle is in fact very simple. Here’s Wikipedia’s definition:

Split testing is an experimental approach to web design (especially user experience design), which aims to identify changes to web pages that increase or maximize an outcome of interest.

Okay. so that may not have made it seem any less complicated, so let me reveal to you a real case study of how I used split testing to boost my sales in order to make its benefits more clear.

My First Split Testing Campaign

I have an information product that I sell on my blog. It has its own sales page. The number of sales I generate is influenced by how well I market the product on that sales page. The words that I use can be more or less persuasive and lead to a greater or lesser chance of the visitors purchasing.

With regards to my sales page, split testing can essentially be broken down to the following steps:

  1. One of two different versions of the same page is served up to each visitor that comes along.
  2. The split testing software makes a note of each sale that is made and attributes it to whichever page was served up to the visitor.
  3. Once the split testing experiment has been completed, the page with the most sales can be deemed to be more successful.


Split testing enables me to make successive tweaks to my sales page in order to boost my sales. Here’s the results from my first experiment:

Split testing

As you can see, the “Original” sales page generated 20 sales from 443 visits at a conversion rate of 4.51%. Meanwhile, “Variation 1″ generated 0 sales from 64 visits. The original sales page (which was in fact a new sales page that I implemented) is the runaway clear winner.

In practical terms, implementing that new sales page resulted in me earning far more money from the sales of my information product than I would have otherwise, and I can now tweak the sales page and carry out additional experiments to increase my conversion rate even further. Who wouldn’t want to split test?

How to Split Test

With that last thought in mind, the only reason I could imagine you not wanting to split test would be its perceived complexity. That’s actually what held me off for so long. But fortunately, the process is not that complex at all. In fact, I am going to walk you through it now.

Step 1: Create Your Two Test Pages

First of all, consider what page you want to split test. Then simply make a copy of it, tweak it as you see fit, and save that as another page in WordPress. You now have two versions of the same page, ready to be split tested.

Step 2: Create Your Goal

I assume that you have a Google Analytics account for your site if not, get one now (you can follow these instructions).

Once you’re up and running you need to create a “Goal” relating to your page to be split tested. For instance, the “Goal” for my sales page was for the visitor to hit a “Thanks for Purchasing” page on my site, as them hitting that page confirmed that they had purchased the product. Your goal could be something very similar.

Click here to learn how to set up goals in Google Analytics (don’t worry, it’s a pretty straightforward process).

Step 3: Set Up Your Experiment

By now you should have your two pages and your goal set up. Now it is just a case of telling Google to split visits to your main test page between it and the variant and track the goal completions per page. Here’s how to do just that:

  1. Log into your Google Analytics account and click on Content > Experiments in the sidebar.
  2. Click on “Create Experiment” on the following page.
  3. Enter the URL for the main page that you want to test and click “Start Experimenting”.
  4. On the following page you can name your experiment, add the variation page and also name both pages as you see fit (these are the names that will show up in the experiment results).
  5. On the next page you need to select the goal that you set up previously as the “experiment objective” (i.e. what Google should track in order to determine the effectiveness of each page in doing its job).
  6. Finally, you need to add your experiment code. This may sound complicated but it doesn’t have to be with the Google Content Experiments WordPress plugin. Just install and activate it, then add the Google experiments code to the meta box on the Page Edit screen (for your original page only) in WordPress.
  7. Once you’re done, click on “Save & Next” in Google Experiments and your code will be validated.
  8. Finally, you have a chance to check over the experiment settings and can then hit “Run Experiment” to start!


Your experiment can be accessed via that same Content > Experiments menu option that you used to initially set up the experiment. You can check in at any time to see how the experiment is progressing (although it will take some time for you to see any worthwhile results), and Google will pick a winner once it has received enough data.

It’s Time to Start Split Testing!

If you own an online business where you make sales through your website, there really is no excuse not to start split testing.

Setting up experiments will only take you a few minutes once you’re into the swing of things, and the benefits in terms of boosting your income can be huge. So what are you waiting for? It’s time to start testing!

This post was written by Tom Ewer, a regular contributor for MyWifeQuitHerJob.com!

Ready To Get Serious About Starting An Online Business?


If you are really considering starting your own online business, then you have to check out my free mini course on How To Create A Niche Online Store In 5 Easy Steps.

In this 6 day mini course, I reveal the steps that my wife and I took to earn 100 thousand dollars in the span of just a year. Best of all, it's absolutely free!

Give Me Access To The Free Course!
Enter Your Email Address:

Similar Posts

Have you read these?

2 thoughts on “Split Testing: Why It Is So Important and How To Do It”

  1. Great tutorial! I have found that many web design companies & SEO firms don’t really use A/B testing that much, but it is so critical to the success of any website.

    Based on my experience, I have found that you should let it run for at least 1-2 weeks. Also, you need to have enough of a sample size to make sure that the tests are statistically valid – and make sure that the test achieves 95% statistical significance (or significantly close to that.

    On websites that have thousands of visitors, some simple A/B tests can make a huge difference in your bottom line.

    Split testing is such a powerful way to maximize a website’s profitability – I don’t understand why more companies don’t do it.

  2. Split testing is the key to success. Gotta figure out what’s working and what’s not so you can do more of what is and scratch anything that isn’t.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>