Before we start, 'A/B' is not an abbreviation or acronym for anything. It is just A/B.
Now that that's out of the way, let's get into the meat of it.
What is An A/B Test?
An A/B Test can be applied to any format of content intended to produce a traceable conversion event.
This could be an email, online display ad, paid search ad and even your homepage. If you are wondering if you can apply an A/B test to something, ask yourself if the end result is a conversion you can track (such as a demo sign-up or even a click to view a blog post) and you are ready to go.
How Do I Run An A/B Test?
So now you have your item to A/B test. You need to decide what you want to test. A common mistake is testing more than 1 variable at a time. For instance, if you are testing an email, you would not want both emails to have different subject lines AND different layouts. The results are not going to be reliable because you really won't know if it was the subject line or the layout that determined the results.
Decide what you want to test and apply ONE variable and run it.
How Long Do I Run An A/B Test?
You A/B test email is likely a one and done. But when you are testing online display ads or landing pages, you want to consider a longer period of time to get good data. I like to use 14 days as a best practice when I am doing this.
Set it up, split your inventory and let it run. Don't touch it for 14 days! I'm confident you will keep checking the results (I am very guilty of this because I'm anxious to see which one worked) but don't declare victory for the winner until your test period is complete.
TIP: I love A/B tests but you can get caught up in them quickly. Pick the winning subject line, now test the image, now test the button color, etc. As you prioritize where you spend your time really focus on the conversions that mean the most to you. It's fun to play around with different tactics but it's about revenue in the end.