Tech Note: Track event properties and reporting best practices
UPDATE:
This Tech Note is no longer being updated. Please refer to these Help Articles instead:
---------------------------------------------------------------------------------
Welcome to part two of a two part series on Track events! For general information and best practices related to track events, please see my first Community post here: Track events best practices
Additionally, the information presented here is also relevant for Click Event Properties.
Reporting within the Pendo UI
When it comes to event properties, reporting options within the Pendo UI are more limited.
Track Event Details Page / Feature Details page
On the individual track events details (or feature details) page, the table will display up to 5000 entries, with each row corresponding to an event with event properties received within the selected filter and date range.
CSV download is also an option and does not have the same limit as the UI - the file will include all events.
Visitor and Account Reports
Please note that event properties cannot be added to visitor or account reports. For visitor and account reports, only ‘event count’ and ‘days active’ can be used for track events. Only 'clicks' and 'days active' can be used for features.
Data Explorer
When selecting track event or feature click events within Data Explorer, you can filter by specific event properties. Read more about this here: Event Properties - Data Explorer.
The ‘Group By’ function is another popular Data Explorer feature and is described here: Data Explorer - Group By. Up to five Group By options can be added.
However, for Group By, the event property type must be either boolean or string. An event property cannot be used for ‘Group By’ in Data Explorer if the event property type is ‘number’ or ‘time’.
In this below example, the track event ‘Record Created’ has an event property type of 'string'. Thus, ‘Group By’ can be used.
The chart will only display the top ten event property values (in my example above, there are only 6 event property values). The top ten is based on the Row Summary column in the breakdown table as shown below. Using the 'Visibility' icon in the breakdown table, you can control which ten event property values to display in the chart. Up to 10,000 event property values will be displayed in the breakdown table.
Recommendations on reporting within the Pendo UI
Event properties can be used to create extremely targeted segments for data analysis within Pendo. Additionally, if you are just looking for a snapshot of data, the event properties table on the individual track event detail (or feature detail page) page within the Pendo UI can be helpful.
If you are looking for just specific event properties, Data Explorer can be helpful.
If the event property is a boolean or string format, Data Explorer and the ‘Group By’ option can be helpful.
As a reminder, if you are sending an event property to Pendo that is a number or time field, you won’t be able to use that event property for ‘Group By’. However, you could send an additional event property that is a string and use that for ‘Group By’. I’ll explain this in more detail with the example below.
Below is an example track event that is being sent whenever someone completes a particular workflow. The event property ‘completion_time’ is a number (measured in seconds) and can’t be used for ‘Group By’. However, the event property ‘completion_speed’ is a string and can be used for ‘Group By’. You could send a limited number of values for ‘completion_speed’ (e.g: ‘slow’, ‘normal’, ‘fast’) .
pendo.track("Workflow1", {
completion_time: 300,
completion_speed:"slow",
result: "success",
});
Event properties and the Pendo API
There will still be situations where you need more robust reporting options than what is available in the Pendo UI. Common use cases are:
- Using ‘Group By’ for event properties with number or time type.
- Reporting on an event property with more than 10,000 values
- Reporting on multiple event properties for a specific track event (or click event) at the same time
- Complete export of historical data
In those situations, you would need to use the V1 API, specifically the aggregations endpoint with the source trackEvents (or source featureEvents). There are third party tools that can be used to pull Pendo data via the API if you don't want to develop the API calls on your own. Some additional resources on the API are below:
Comments
Post is closed for comments.