The Alerts feature is available in Countly Lite, Countly Enterprise, and built-in Flex. However, Countly Lite offers limited access.
Getting access to data on-demand via comprehensive reports and dashboards is great, and that's what we excel at. However, you sometimes need to be aware of changes in important metrics even if you are not looking at your dashboard.
The Alerts feature is a highly beneficial reactive tool that sends you an email when metrics related to a specific condition on metrics like crashes, cohorts, data points, events, NPS, online users, rating revenue, sessions, survey, users, or views are met.
Benefits of Alerts
Alerts provide information about immediate changes in your data in relation to a given condition, triggering a notification when something needs your attention. It is a simple but effective mechanism to know right away about changes such as an increase/decrease in new users, occurrence count of an event, metrics, new crashes/errors, data points, online users, or changes in user ratings for a new feature or page.
With Alerts, you can quickly identify where to intervene and solve problems that your users are experiencing. For example, you released a new version but there may always be bugs that may have run out of sight even though the tests went smoothly. Some bugs may even be so critical that they may prevent your users from using the application. Setting an Alert that allows you to identify such situations can be a lifesaver.
Getting Started
Before using Alerts, you need to ensure that the feature is enabled. To do so, go to the Sidebar > Management > Feature Management and enable the Alerts toggle. Once this is done, you will be able to start using Alerts by going to the Sidebar > Management > Alerts.
Using Alerts
The Alerts view displays a dashboard with current statistics, a +Add New Alert button, and a list with the alerts created and which your user has permission to view. Each alert has a toggle button to disable it, as well as a 3-dot ellipsis menu to Edit or Delete it.
You can choose which application's alert data you want to explore with the Dropdown menu at the top of the data table. Here, you can select the specific application or, to see all applications' alert data, select All Applications (selected by default). The data table and widgets will update with the relevant information.
Creating an Alert
Click on the + Add New Alert button a drawer window will appear. You can create Alerts by filling the data in the drawer window.
- Enter the name of your Alert.
- Select the application for which you want to create an Alert.
- Choose the data type from which you want to create the Alert. You can select from options that include Crashes, Cohorts, Data points, Events, NPS, Online Users, Ratings, Revenue, Sessions, Survey, Users, and Views. When you select data types other than Users and View, you will have to add widgets or any data that will appear below this field.
- You can define the trigger using Metric, Variable, Value, and Time.
-
- Metric: You can select any metric from the list of metrics that are available for your app.
- Variable: You can define a variable whose value will be set.
- Value: The value of the variable, that you set will act as a threshold. By threshold, it means that, if the value exceeds the threshold, the alert will get triggered.
- Time: You can select the time period in which the value has been triggered.
5. You can select if you want to send alert to a specific email address, and then enter the email address in the below field.
6. You can select if you want to send alerts to a user group which has already been created. In this case, you will have to enter the group name in the below field.
7. When you select this option, the alerts will get sent when a trigger is encountered. A trigger is basically created on the basis of a hook. Once the trigger is encountered, the hook gets executed, and the alerts are sent based on that.
Click on save.
Your new alert will be displayed in the Alerts table.
Understanding Alert Data Types
There are various data types on the basis of which, the alerts are sent. The data types include Crashes, Cohorts, Data points, Events, NPS, Online Users, Rating, Revenue, Sessions, Survey, Users, and Views. The reason behind including this is to make the usage of alerts more extensible. To do so, you can use hooks and triggers with any data type. Whenever a variable defined encounters a value bigger than the threshold value, the triggers will use the hook to send alerts to the users based on the usage of the data types.
FAQ
Why can we add only one level of filtering?
The community edition doesn't have granular data; it doesn't actually let you segment anything by multiple levels. Because of this, to be able to make this work in both the community edition and enterprise edition, we'll allow one level of filtering.
When you select that one level of filtering, the hour option will not be selectable because of the same reason. In the Community edition, we don't store hourly data or segment value in the aggregated data format. So this won't be available as an option.