Why A/B Testing & CRO are Important

Let's get right into it.

What is A/B Testing?

Simply put, A/B testing is a method that compares two versions of a webpage to decide which performs better among users. This is basically a scientific experiment where there are control and a variation of the webpage and the users will be randomly picked to view one or the other. Then the pages are analyzed to see which is more apt to users completing the conversion goal. As consumers, we may have been apart of A/B testing without even knowing it!

What's the Difference Between Site A & B?

When performing an A/B test anything on the webpage can be modified to create the second version. Maybe its the change of a title or the size of a button or maybe it's a full-fledged redesign of the page. Any of these modifications can be part of the variation.

How Does This All Work?

There's an easy framework many people use to get the process started. It's as follows.

1. Collect Data: As we learned in my last blog about Google Analytics, data matters. Analytics can provide a much-needed insight into how a company can optimize their webpage or app. The key is to look for pages that have low conversion rates.
2. Identify Goals: What is your conversion goal? This is what helps determine if your variation is more or less successful than the original page.
3. Generate Hypothesis: You thought as a marketer you didn't have to deal with scientific experiments? WRONG. Time to put on a lab coat and go back to high school chemistry because we're making a hypothesis! This is where you decide why you think the variation will be better than the original.
4. Create Variations: With an A/B testing software you can make the variations needed to test out your website or app. Again this can be simply changing the color of a button to a full revamp of the page. A software that's great for this is Optimizely.
5. Run Experiment: Time for some excitement! This is where visitors will be randomly assigned to the control or the variation of the page. These interactions are how you decide which site (control or variation) most optimal for conversions.
6. Analyze the Results: The software you use will collect the data and from here you can analyze the results on how each page performed. This is where you decide if there is a statistically significant difference between the pages. (I know how to do this, thanks DSCI 205 & 305!) If your variation came out victorious you can apply these changes to your website, if not, its time to go back to the drawing board.

Why Should You A/B Test?

A/B testing is a valuable resource for marketers because it can help them optimize the user experience and get those conversions! Testing small changes can help zero in on how the variation has an effect on the behavior of visitors. In the marketing program here at Western, we took consumer behavior which teaches us how to effectively understand what influences customers' buying decisions, with A/B testing we are able to statistically see how a single difference can influence behavior. Heck, that's really cool!

Now on to Conversion Rate Optimization (CRO)

Conversion Rate Optimization (CRO) is described as the "process of increasing the percentage of website visitors who take a desired action" (MOZ). I spoke about it before but I'll reiterate, a conversion is when a visitor of your site completes the site goal which can be a number of things. This can be anything from purchasing a product to signing up for the mailing list.

Examples of Conversions

Macro-conversions:

• Purchasing a product
• Requesting a quote
• Subscribing to a service

Micro-conversions:

• Signing up for email lists
• Creating an account
• Adding a product to a cart

A/B Testing and CRO Go Together Like Peanut Butter and Jelly

CRO, getting those conversions, is the overarching umbrella. A/B testing helps us set goals, hypothesize, and test variations that lead us to optimize our webpages and apps. This tool is essential for marketers to understand how to best serve consumers and reach their conversion goals. A/B testing is used all the time, even Obama's Analytics Director used it to help them raise money for his 2008 campaign. They tested four different action buttons and six media variations that led to 24 different combinations. Through testing and analyzing the data they found that the 11th combo worked the best and this gathered them an additional $60 million in donations. This is just one example of how A/B testing is a fantastic tool for marketers to have in their kit.