From the outside, it's like A/B or multivariate testing: choose any number of subject lines or from names to test. That's all you have to do.
But after you send, MiraPost uses a special algorithm, based on data about your subscribers, to choose which version of your email to send to each subscriber.
So instead of one version of your email being chosen as the winner and sent to all of the remaining subscribers on your list, our algorithm chooses multiple winners - because different users respond to different emails.
You're sending an email with a few coupons inside to 50,000 subscribers. Which subject line should you choose?
Our algorithm sends three emails, each with a different subject line, to representative subsets of your list - say, 1,000 users each.
Once these subsets start opening and clicking the sample emails, our algorithm analyzes the attributes of what type of user clicks on each version.
Maybe Hector B, a 27-year-old, male, single banker in New York making $150k / year, responds to the email with the subject line "25% for Men's Watches". And Jasmine H, a 40-year-old married woman from Chicago, responds to the email with the subject line "30% off Jewelry and Handbags".
After analyzing the subsets for users like Hector and Jasmine, MiraPost splits up the remaining 47,000 users. People like Hector get the first subject line, and people like Jasmine get the second subject line.