How our predictive frequency model works

The recommendations and graphs in our predictive frequency screen are the result of complex AI algorithms, and aim to support you in deciding which email send frequency you may want to test.

See also: Testing predictive frequency

To do that, we need to make some important assumptions:

1. We rely on your past email performance numbers to predict future revenue. If you find that your email revenue performance varies a lot from one month to the next, you should consider this when reviewing your predictions. 
2. The data you see in the predictive frequency dashboards is refreshed at regular intervals throughout the month, so you might see some results changing slightly as the model continues to learn from your data.