A/B Testing: What is It and What Should I Test?

Most online marketers spend a lot of their time generating traffic to their sites and rightly so, after all it is traffic to your site that drives sales right? Whilst no one can argue that without traffic you won’t have sales, the number of sales you make can be calculated from a very simple formula: Visitors X Conversion Rate = Sales. A conversion rate is the rate at visitors convert to sales (or any other desired action such as newsletter sign-ups, clicking on AdSense banners, buying an affiliate product or filling in a lead generation form). The easiest way to work out your conversion rate is the number of sales (or other measurable actions such as newsletter sign-ups), per 100 visits. So if you make 2 sales (conversions), per 100 visits your conversion rate is 2%.

Working from the above formula the 2 main ways to increase sales are by increasing traffic or increasing conversion rate. So how can you increase conversion rate? Many website owners or online marketing professionals will make changes to their site based on ‘gut feel’ of what’s right and wrong. However this is far from a scientific approach and given how important conversion rate is you really should use an approach that allows you to easily measure the results and run fair tests between different changes you make to your website.

The most simple of method is called A/B testing. A/B testing allows you to test different versions of your website pages to find the best converting option. This is done by sending half of your traffic to page ‘A’ and half to page ‘B’ and measuring the resulting conversions from both options. A/B testing can be run free of charge using Google website optimizer; there are some good tutorials to get you started including this guide to getting started in 5 minutes.

One of the major issues many people have when they are new to A/B testing is figuring out what they should test. The aim of this article is to short cut that learning process to reveal the most effective things to test to increase your website conversion rate rather than to give an in-depth view of the technology and how it works. If you do want to read up more on how A/B testing works, you can read Maxymisers’ A/B testing guide which gives a good overview.

When running an A/B test, your first inclination may be to test the design of the site. While the site’s appearance has a strong impact on the viewer and the impression they have on your product or service, oftentimes one forgets that there are other elements just as important and in some cases, more so. After all, what works for one company may not work for another in terms of selling products. In this article I will present key elements to test in determining which factors are the most effective in relation to your site and its offerings.

There are a large amount of elements on your page to test that can potentially double or triple your conversion rates. In some cases they may even have a larger impact. Here’s a brief list of the web copy elements on your site you might test:

Photo Courtesy of: Stephanie Hamilton


A headline draws the visitor in and gets them engaged with your content. In creating compelling headlines, test different wording in conjunction with different text sizes for the most optimum results. The headline with the most clicks in the duration of the test wins.

Call to Action Text

The call to action button is perhaps the most important element that a website should test. After all, it’s what leads to conversion and sales. Reasons your call to action may not being working include its text. It may not be motivating or visitors may not realize it’s a ‘Buy Now’ button. For instance, if you’re an ecommerce website simply changing the text on your button to convey limited time offers can have drastic results.

Call to Action Position

If nobody notices your call to action it’ll be hard to convert. Placing your call to action button in an easy to scan location promotes usability and conversions. Test a few different positions (above body copy, beside body copy, below body copy, within body copy, etc.).

Call to Action Style

For the most accurate results, test a variety of different styles individually. How you style the call to action includes whether it’s a text link or a button. Or styles may include size and colors used.

Copy Length/Style

Besides the design of the page, the copy plays a pivotal role in influencing and converting visitors into customers. The length and style should be considered as it plays a role in how many visitors read it. Whether you wish to test different copy altogether or various formats such as lists, remember to test your copy frequently for maximum results.


In terms of e-commerce websites, images play a large role in whether a visitor converts or not. Test out which images work best for the product you’re selling, how many of them to use, and how large they should be for the most impact. In other scenarios, human photos were shown to have significant increases in conversion rates when A/B tested. Subconsciously, we are attracted to images with people.

Different Offers

Lastly, you’ll want to test which offers are working the best for you. A visitor may respond differently to an offer according to the context of the page the offer is promoted.


Once you’ve determined what you test, figure out how long you should run that test. After you figure out how long to test, look at the tools available to help you carry out the testing, we’d suggest starting with. By determining the elements to test and how to execute them, you’ll have a better way of ensuring successful results of A/B testing.

Click to rate this post!
[Total: 0 Average: 0]

Madalin Tudose

A web developer with a crush on SEO. Having my skin in the game of website development and digital marketing for more than 10 years already, you might consider me an expert. At least this is what people call me. Honestly, I HATE that term. I prefer to describe myself as a person who takes action and risks. I test every hypothesis, document every step of the process, and implement what works.