A/B testing: more proof you should always be testing.

Written by Guido Jansen in
October 2010

Last month I performed a simple A/B test on another (Dutch) website of me (dutchento.org) using Visual Website Optimizer (VWO). It was a simple A/B test looking to improve newsletter and RSS subscribtion rates by adding 'persuasion' (three reasons to sign up) to excising subscription links. The results were somewhat surprising to me and apparently also to the people behind VWO because they liked to use my test as a case study on their blog. Read on to learn what I learned by using split testing: A/B testing with competing goals: newsletter CTR increased by 190%, but clicks on RSS feed…. [disclaimer] I'm a happy customer of VWO, It's one of the systems I use to analyze and test websites. I like to see myself as an objective person in talking and advising you about that. If I recommend bad software you won't come back here and you won't tell anyone about me so there's really no use for me to misinform so I never will. I do however signed up for the partner program of VWO which means that if you click the links to VWO in this blogpost I get a 15% commission if you buy anything from them within sixty days. This of course will make me huge amounts of money and I will have even more incentive to blog great things for you. If you don't like that just click this plain link to visualwebsiteoptimizer.com without the referral code. Thanks! [/disclaimer]

Recent posts
Announcing a new Magento community project
Announcing a new Magento community project

I would like to share a new project with you. I’ve had this idea for over 5 years now and I’ve finally been able to make it a reality in the last couple of weeks.

Can you afford not to know? Can you afford to be wrong?
Can you afford not to know? Can you afford to be wrong?

In rapid changing online environments, continues business experimentation is a great way of constantly learning what works and what doesn’t. But then the question might arise: how far do you go? What do you test? And is it sometimes ok NOT to test something?