KPIs provide you a way to track your performance against a dynamic or static target.
Once you are tracking a KPI, Jepto fetches new data every hour and builds a Machine Learning model using your historic data to generate a prediction for the rest of the reporting period. Along with the prediction a calculation for pacing is made to determine where you are, compared to where you should be in order to meet your target.
Using these two values a KPI will be assigned a status.
Example: You are tracking Website Form submissions, with a target of 100 per month and you are currently 14 days into the month with 30 form submissions submitted so far. Even pacing would dictate that if you are 14 days in you should be around 45% of your target, however your submissions are below this and therefore would be considered below pacing. The Machine Learning model predicts that you will only reach 85 submissions by the end of the month, and therefore the prediction below is below the target. The combination of these two will result in the status being Critical.