Smart Alerts

Smart Alerts allow users to receive alerts on specific objects, whenever we detect an outlier (from the normal distribution) of the number of records that have been added, removed, or changed.

An outlier is a data point that differs significantly from data expected in a normal distribution. Outliers are major deviations from the general trend of your data. Significant changes in the added, removed, or changed record counts could indicate a data loss or a data corruption event that you should be aware of.

Statistical Outlier Detection

User can be alerted when outlier data appears. Smart Alerts support using either (or both) of the following mechanisms:  

1. Detecting statistical outliers in your data, based on significant deviation from calculated key values (based on a set of previous backups)
2. Detecting data that exceed thresholds set, based on “More Than” and/or “Less Than” scenarios (values set by the user). 

To accurately capture the data trends, distribution, and growth, we use a sliding window of the last 60 days. Backups during this period will be taken into consideration when calculating outliers. Setting alerts based on statistical outliers for a particular service requires a minimum of 14 backups. This is to ensure enough backups were accumulated to capture the trend and distribution of data (number of records added, removed, and changed).

Setting Up Smart Alerts

For more information on how to set up a Smart Alert in a Salesforce service, see here.

For more information on how to set up a Smart Alert in a ServiceNow service, see here.

For more information on how to set up a Smart Alert in a Dynamics service, see here.