Kissmetrics Blog

A blog about analytics, marketing and testing

Built to optimize growth. Track, analyze and engage to get more customers.

Creating an Effective Conversion Optimization Process (Infographic)

Many companies don’t have success with A/B tests for one simple reason.

There is no process or system set up to run A/B tests. Marketers blindly throw tests up and see what happens. They have no process for creating tests, running tests, or learning from tests. They run tests on things that don’t impact the bottom line, and they stop tests too early.

The solution to this problem is an easy one.

Create an A/B testing process. It is one of the best things you can do to increase your success. Have a method for finding what to test, creating hypotheses, setting parameters on the test, and writing a recap of what you learn from each test, and you will be on the road to conversion optimization success.

Below you’ll find an infographic that outlines how to create an effective CRO program in your company.

The Perfect Execution to Conversion Rate Optimization
Courtesy of: Quick Sprout

Kissmetrics A/B Test Report

The Kissmetrics A/B Test Report empowers you to see how an A/B test impacts any part of your funnel. You can still create your tests in Optimizely, VWO, or any other platforms we integrate with. Then track the results in Kissmetrics with the A/B Test Report, and see how the test impacts the entire funnel. No longer are you limited to testing to the next conversion step. See how a headline test impacts sign-up rates with the A/B Test Report. As long as you’re tracking it, you can get the data that matters to you. Here’s a video demo of the report:

About the Author: Zach Bulygo (Twitter) is the Blog Manager for Kissmetrics.

  1. Marston Gould Nov 16, 2015 at 5:38 pm

    I think the biggest reason that most A/B tests fail is that most testers have forgotten that the biggest variable they are often impacting is change itself.

    There is a reason why many tests outperform a control for some period of time and then fade back to the norm…..

    The trick then becomes, what is the appropriate level of change…

  2. Can someone please explain to me the history of A/B testing? I don’t understand how this method of testing could actually provide factual data when there are so many other variables that influence the way people use the internet.

    Obviously, A/B testing works or people wouldn’t use it… This test is definitely not a controlled experiment so I don’t understand why people read the corresponding data as if it were fact. More info would be great! :-)

    • Hi Hannah,
      This is why it’s important to reach statistical significance when A/B testing. You need to make sure you have enough people in your experiment and enough conversions to be able to confidently move forward with the results. If it’s a winner at 99% significance, chances are good it’s a permanent improvement. If it’s a loser at 99% significance, chances are it’s a permanent loser.

      The big issue for many websites is that they simply do not have enough traffic to warrant A/B testing. Only sites that get thousands of conversions every month should be a/b testing…the rest will just be wasting their time (and money).


Please use your real name and a corresponding social media profile when commenting. Otherwise, your comment may be deleted.

← Previous ArticleNext Article →