Compare the open rates of 2 different subject lines.
1. Determine two email subject lines you want to test. Let's name one "A Control" and one "B Test". Usually the current subject line already in use is called the "control", but you could be testing 2 new subject lines. It doesn't matter.
2. You could just take your entire list and split it in half. However, usually you want to do a little test before you pick a winning subject line so you can send the winner to the remaining people on the list.
Using the chart below or a sample size calculator, figure out the number of people you need to email as a sample test to get valid results.
Picking the size of your test depends on your predicted open rate, total number in your mail list, and the confidence level you want in your results. (To simplify a bit, let's just assume a 95% confidence level with a margin of error of 5%.)
For example, if you have a mailing list of 5000 and an average open rate of 50%, you would send the "A Control" subject line to a random sample of 357 people and send "B Test" subject line to a random sample of another 357 people.
3. Randomly select your 2 mailing samples.
4. Email your tests at the exact same time to eliminate as many variables as possible.
5. You can then compare the open rates from samples and pick the winner to use as the subject line for the remaining people in your full list.
Not sure how to randomly select email names? One trick I use is to export the emails to Excel, use the RAND() function in a column, and then sort by that column. You can then select the first x number of rows to get a random sample.