Natero's advanced machine learning models look across many factors to decide if an account is likely to churn, expand or convert. When a predictive alert triggers, it'll come with a list of reasons the model believes are most relevant to the situation predicted.
Below is the description of some common measures you are likely to see in your predictive alerts.
|Age||Time since the "join_date" of the account.|
Counts how many times each specific feature event was logged for each account, per week.
|Module Usage||Tracks time spent in each tracked module, per week.|
|Total time spent in all tracked modules, per week.|
|Counts how many days the account recorded activity per week, from 0 to 7.|
|Open Support Tickets||Counts how many support tickets are still open, per week.|
|Distinct Features||Counts how many distinct features (tracked by Natero) were used in a given week by each account.|
We also look at the relations hip between these measures across different weeks, in particular:
|Momentum of...||Compares a weekly metric, such as total usage of a specific feature or total session time, against historical weeks. Typically compares the most recent week against 2 or 3 weeks ago.|
|Current Time Since...||An activity index that is higher the more recently an account has done something, such as using a specific feature, or having a Salesforce interaction logged.|