Did you know that I was a chemical engineering major as an undergrad? While I don’t mess with test tubes in a lab anymore, I do believe in continually testing marketing efforts. For example, you are receiving this newsletter on a Wednesday and last week’s was on Wednesday too. I’m testing open rates on Tuesdays (the old regular newsletter day) to Wednesdays. Of course, just one week isn’t good enough for a reliable test. There can be other factors that skew a single observation so it’s better to have multiple observations to compare. I’ll probably try Wednesdays for a few months to match my Tuesday data.
Does my testing need to meet the requirements for statistical significance that would be required if I were to try and publish it in an academic paper? Not necessarily. But, knowing the theory behind running a good experiment is certainly useful so that you draw the right conclusions (and know their limitations).
What else can you easily test? With newsletters- you can test subject lines(which I do also… did you get “Testing, Testing, 1.2.3.” or “A.B.T. Always Be Testing.”? You can compare open rates as I’m doing or click through rates. With ads (Google, Facebook, etc.) you can test the copy, the images, the target segments and again, compare click through rates or even conversion rates if you drive the traffic to different pages.
What are you testing? Do you need help to implement some marketing testing? I have helped several clients to test and enhance their ongoing marketing efforts, give me a call if you need some help!