ReDrew Innovations

What is AB Testing? How do you do it?

AB Testing is a randomized controlled experiment used to compare the difference between two independent groups (A and B). The purpose of AB Testing in a marketing context is to improve user interface and user experience which will translate into a conversion. 

AB Testing first begins by using web analytics to identify areas of a site with high traffic but low conversion rates. The company then chooses KPIs that will be used to determine whether or not the change has been successful. A hypothesis follows next, explaining an A/B testing idea and hypotheses for why you think this change will be better than the current version. Then the company makes the desired change leaving one sample the same as when they initially started to be the controlled variable. Then the AB Test is performed and participants are randomly assigned to either the dependent or independent group. The interactions are measures to see how they compare. Lastly, your AB Testing software will tell you the difference between the version and whether there is a statistically significant difference.

A marketing example could be that a company noticed that their consumers are abandoning their carts during the checkout process. A KPI, in this case, could be shopping cart abandonment rate. An AB Testing variation could be changing from a multipage checkout to a single-page checkout. This variation would be better as there would be less friction in the checkout process. Once this variation is completed, the controlled (the current multipage checkout) and the variation (the single page checkout) can be performed, sending a randomized equal amount of participants to each. Then the results are analyzed to see if the data supports the hypothesis. At the end of this AB Testing, your company will either find that a single page checkout leads to more conversion or that there were no statistical significance (significant difference) and more testing needs to be conducted.

What is CRO? How do you do it?

A conversion rate tells you the percentage of users who perform a desired action on a website. For an eCommerce website, a conversion may be the percentage of website visitors who purchased from your online store.

Conversion rate optimization from a user-centric (qualitative) position is the process of understanding what drives consumers to a website, the barriers that make consumers leave a website, and the hooks that persuade consumers to convert. This gives users the best user experience possible and leads to a greater conversion rate. 

Companies can also improve conversion rate optimization by using quantitative tools to track interactions on their website. For example, web analytic tools like GA4. However, the downside to just using a quantitative tool is that you lose the reasoning behind a consumer’s behavior. This is why it’s best to use a quantitative tool like GA4 and a qualitative tool like an exit survey. Mouse tracking and heat maps are great examples of a quantitative and qualitative CRO tool in one.

How are AB Testing and CRO similar, and how are they different

AB Testing and CRO are similar in the sense that both strategies are used to increase conversion. They are different in the sense that CRO is a strategy to increase conversions and AB Testing is a tactic that can be used to better understand CRO.

Amazon is an example of a company that continues to utilize AB Testing to improve CRO. For example, Amazon AB Tested the removal of the “Buy Now” button on specific product pages by randomly assigning consumers to one variation with the “Buy Now” button and one variation without the “Buy Now” button. Later the “Buy Now” button was added back to those pages. This means that Amazon found that consumers preferred shortcut buttons. Overall the AB Test to keep the “Buy Now” button will help to increase CRO.