What is AB Testing (A/B testing)?

AB Testing, also known as split, multivariate or bucket testing, is a method that consists in comparing two (or more) versions of a page/app in order to gather data on which one performs best.

All businesses have a reason to exist, an ultimate goal. That wouldn’t be different for websites or apps. For example: if you own an eCommerce, your success metric is based on how many visitors end up buying from you or if you run a SaaS app, you’ll want users that signed up for a trial converting into paid users.

What this means is that ultimately all businesses want simple visitors converting into something else (as mentioned above, buyers or paid customers).

When you compare and contrast two or more versions of a website/app you’ll be analyzing which of the versions allows you to convert more users into your success metric.

AB testing is basically an experiment where two (or more) versions of a page/app are shown to users at random followed by statistic analysis to determine which performs better for the previously established conversion goal.

Directly comparing a new version to your current version using AB testing is a great asset for focusing on how (and how much) will the changes impact your users.

Deciding towards AB testing is the first step of going from “I guess” to “the numbers show that”. By enabling data-informed decisions it’s a precious asset towards website optimization and ensuring all changes will produce positive impact.

AB testing can be used to comparing simple aspects of a page/app, such as the position or the color of a button or even complete redesigns.

This testing method works by automatically sending proportional amounts of traffic to each version and collecting data to measure which one your users prefer.

When is AB Testing a good idea?

I can safely say “always”. AB testing is an amazing method that allows you to make more with less.

More than just allowing you to improve your site/app user experience, testing will allow you to build carefully developed hypothesis and learn a lot about how each simple modification can impact your user’s behavior.

AB testing is a safe and cost effective way for continually improving your overall user experience as well as one (or more) business goals such as conversion rate.

Acquiring paid traffic can cost a lot of money while AB testing can cost as little as $49/month (if you decide to go with VWO which is the tool we use here at Pipefy).

Improving your conversion rates while spending a lot less with lead acquisition can result in significant increases in sales and revenue, skyrocketing your ROI.

It’s essential that you always keep in mind that testing small changes at a time will allow you to identify much more precisely which of the changes affected your visitor’s behavior the most.

Recommend this article

Profile photo for Isabelle Salemme

Written By

Isabelle Salemme

Head of Customer Support & Education @Pipefy. She uses her extensive Pipefy knowledge to write informative pieces teaching users to make the best of Pipefy. Besides being in charge of product knowledge, she's an avid reader, a coffee lover and a professional photographer.