San Antonio, TX

Uplift ROI delivers pay on performance conversion rate optimization. Get in touch today.


San Antonio, TX

Mon - Sat 8.00 - 5:00


Split Testing

What is Split Testing?

Split Testing simply means that you “split” the traffic of your website between two site versions. Then, you compare the conversion rate between the two versions to determine which site variation has a higher conversion rate. Split Testing is also referred to as “AB” testing, splitting traffic between variation A and variation B.

Split testing is a random process generating statistically significant results. A true split test is the only way to tell whether changes to your website are delivering a higher conversion rate, as simple before and after comparisons are not accurate.

Split Testing Example

Here’s a simple graphic that shows the basics of a split test (taken from our friends at VWO):

split testing

In this example, traffic is “split” between the two site variations. The first version delivers a 22% conversion rate, whereas the second site measured a 52% conversion rate. Here, the winner is clear! The split test shows that the blue version of the site delivers a higher conversion rate when traffic is randomized.

Why Split Test?

Split Testing is just a method of conversion rate optimization, or the process of maximizing your website’s conversion rate. Testing scientifically allows you to determine what elements of your website are truly contributing to sales vs. what elements you need to change.

Without performing experiments on your site, you potentially miss out on revenue.

Why optimize your conversions?

Because SEO and PPC have gotten so competitive, achieving a higher conversion rate can be your competitive advantage. If you can convert double the number of customers compared to your competition, you can afford to spend more on your PPC campaigns, thus bumping them out of a competitive spot.

Another way to think about this is that you are widening your marketing funnel:


With more customers getting in the door, you don’t need to spend as much on marketing to acquire them, thus decreasing your marketing expenses.

Split URL vs Standard Split Testing

It’s important to differentiate between a split URL test vs a “split test” that most people talk about.

In a split URL test, you are literally sending traffic to two different site variations. For example, some traffic would go to www.site.com/A and www.site.com/B.

However, in a split test, all traffic goes to www.site.com. By using testing software that employs javascript, you can show or hide elements to 50% of traffic, and not display those elements to the other 50% of traffic. This is traditionally what we refer to as a “split test.”

When You Should Split Test

Split tests are helpful in all of the following situations:

  • Product Changes: Do you have a new product that you are rolling out? Why not try a simple split test on your website, offering it to some users and not offering to others? Testing in this way will give you data to make decisions rather than testing everything on your own.
  • Design Changes: Do you have a “gut” feel that you need to change the design of the site? Try testing it! Again, performing a split URL test will give you data to make the decision rather than trusting your gut.
  • Fixing “Leaks” in the Funnel: If you know the biggest areas of drop off, you can fix the “leaks” in your funnel by performing tests. For example, if you see that visitors are getting stuck at a sign up page, you can try performing a split test to improve sign up rate.

Try not to overtest

The flipside of split testing is “overtesting,” or trying to improve conversion rate too fast. Don’t do this! Split testing takes time to do properly, often months of effort. It is a long-term investment and not simply a quick win for more revenue.

If you test too many things at once, or do not give tests enough time, you won’t get the type of meaningful results that you need, forcing you to retest things that you’ve already tested.

Need help getting started?

Our team has completed literally hundreds of AB tests for clients, averaging a 28% improvement in conversion rate. Get in touch today and get a free evaluation for your website.