How not to run an A/B test
 

A great writeup by Evan Miller on the principles of A/B testing:

How Not To Run An A/B Test

Re: How not to run an A/B test
 

To me, this type of article mentioning stats formulas scares off too many people from doing A/B testing. Much more important is how you generate test ideas - always use a test hypothesis, and always create insights from your web analytics tools and visitor feedback.

Doing that will ensure you get better test results that get test significance much quicker, rather than getting bogged down by understanding statistics and potential 'analysis paralysis'.

And don't forget, doing A/B testing is just a small piece of the CRO puzzle! And many sites don't even have enough traffic to do A/B testing - but they can still make improvements based on CRO best practices and do better analysis of their website and visitors.

Re: How not to run an A/B test
 

Google has a simple a/b split testing platform. You can test multiple different pages, you just need to pump a lot of traffics to the test pages.