Thanks for stopping by and commenting. I may take you up on that offer for more guidance (and possibly a tour of Publicaster at some point, too).

Regarding your first question, there are tools out there made just to make testing easier for you (we have one in our Publicaster application). These tools are made so it’s really dummy-proof to set up: just tell the tool what you want to test, to which group of subscribers, and what you would consider to be a success. Once you have that set up, you really can sit back and forget about it (i.e. move onto your next task!).

I would suggest starting small with a date/time or subject line test. These are the two most basic tests. Test a Wed morning against a Wed afternoon, or a Thursday morning against a Saturday morning. Or try two different subject lines, one with branding and one without. These two simple tests will help you to find out what your subscribers respond to best, helping you to increase your overall response rates. From there, you can get more involved (when you have time) with creative or landing page tests.

Regarding timing, we typically suggest clients wait at least one full day to let their test results propagate. So, send the test to X% of your list, wait at least one day, then send the winning variable to the remainder of your list. But again, it’s about knowing your list. If you know that 90% of your subscribers take action on your email within the first 3 hours after send, then that might be enough time to let your test ride. If you don’t know this, I’d recommend sticking with the one day rule.

I hope this information helps! I also wrote a blog post about the steps of setting up and conducting a test that you might find helpful: http://blog.blueskyfactory.com/general/email-ab-testing-and-chemistry/

Feel free to reach out if you need more guidance … and good luck!

Appreciate the ideas. I find myself a bit overwhelmed these days, and all you hear in the space is “Test, test, test.” Just looking for an easy place to start and you’ve given that to me.

Thanks!
Scott

I would start with a simple test – a survey of your recipients. Find out what they like, don’t like, love to hear about, issues with HTML rendering, etc. The results are immediate for the most part and you’ll address the issues most important – the one’s that deal directly with your audience.

Don’t worry about A/B testing an image if you haven’t had a candid conversation with your list. Just a thought…

This post was mentioned on Twitter by scottwritesall: The Bane of Testing: Allow me to vent for a bit. A common frustration of mine when working in the email marketing … http://bit.ly/8WCQbW...