Using A/B Testing to Get the Most Out of Your Website

Posted on June 5, 2013 | Written by | Print Post

A/B Testing can be defined as the statistical comparison of one webpage’s element against another to see which results in better performance, and when done properly, it can serve as an effective method for increasing revenue, understanding customer demands and identifying sections of your website that are critical to achieving higher conversion rates.

A Typical A/B Test consists of a Control Page, which usually features an existing element, and a Challenger Page, which features a new element that you’d like to test the effectiveness of. With the help of Google Analytics, both pages can be run simultaneously and relevant data can be gathered from any users that visit the site. The results of the test can be used to determine which page results in higher conversion rates as well as why that particular page is performing the way that it is.

While A/B Testing has proved that it’s effective in determining if a new addition (or tweak) to a webpage will actually improve the performance of the entire website, there are a number of things that must be taken into consideration in order to achieve optimal results.

Identify What you Want to Achieve – Setting goals and identifying ways to track progress are critical to the A/B Testing process. A goal can be as simple as increasing website sales but the more specific the objective, the easier it is to measure. Being very specific also helps you to segment and accurately identify target groups that the changes to your website are affecting.

When developing your goals it’s helpful to obtain feedback from customers in order to understand why something is happening. In many cases, surveys can be utilized to pinpoint both the problematic sections of your site that need to be addressed and the strong sections that can serve as templates for success.

Determine How to Measure Progress – Once a goal has been identified, it’s important to identify quantifiable metrics or Key Performance Indicators (KPIs) so that you can track progress. A KPI will give you some indication of how well your challenger page is performing with respect to your goal. If your objective is to sell more products, then your KPI can be as simple as the number of sales over a designated period of time.

Identify Which Pages You Need To Test First – While it’s certainly possible to go through and test every part of your website, it’s much more efficient to identify specific sections that have the highest potential for helping you increase performance, and address them first. Even though you may think that your recent drop off in sales is being caused by a poorly designed “Products” page, it may be because most visitors are landing on your “Category” page and they don’t like what they see. Use analytics to see where the majority of traffic is going and then use A/B Testing to figure out how to take advantage of it.

Develop a Hypothesis, Test and Retest – Once you’ve identified which part of your website you’re going to focus on, you can begin to test. If you’ve seen a recent drop in sales, you need to figure out why. Examine your page and make changes that you think could potentially boost conversion rates; then test to obtain results. If the results aren’t what you were expecting, make changes and test again. Changes can include alterations to the page design, layout or written content. The more tests you conduct, the more telling your data will be and the more efficient your website will become.

It’s extremely important to note that when trying to interpret the statistical results of a test, sample size and confidence interval must be taken into account. Data gathered from two thousand visitors is more telling than two hundred visitors. Similarly, if you see that you’re getting a specific result 95 percent of the time, it is much more telling than if you’re getting a result 60 percent of the time. The higher the percentage, the less likely that chance is to blame for your results and the more likely that your change (whatever it may be) is directly responsible for any trends that you observe.

A/B Testing is a very simple and very controlled way of measuring the performance of your website, and through methodical planning and careful observation it can serve as a useful process for gathering critical information about potential customers, growing your online presence and continuously streamlining business operations.

wrote this on June 5, 2013

About Dusty Dean
Dusty is the CEO/Chief Web Business Analyst at BitCadet.

Read all of 's posts