Actor columns are specified during the data onboarding process. An actor is one of the mandatory fields for every logged event, along with a timestamp. You can associate a property with an actor using one of the following methods:
- Show—The most common actor properties aggregate events or flows. For example, an actor property of count number of sessions for sessions with total time >5 min.
- Filter and label—If you have several actor properties, you might want to perform a bucketing operation, for example, if "Num5MinSessions > 10" then "heavy user" otherwise "light user".
- Calculate—You can perform a mathematical function and apply it as an actor property. For example, if you have Num5MinSessions and Num60MinSessions, you can apply a function that returns a ratio of the two for the actor property.
Defining an actor property
This section demonstrates how to create an actor property using the Show method. In our example, we create an actor property that determines the movies watched by each user. You can apply the principles from this example to create your own actor property.
To create an actor property, do the following:
- Choose a dataset from the drop-down list at the top of the window. In this example, we chose the movies dataset. The options available to you will reflect your data.
- In the left navigation bar, click Data. Click the Actor properties tab, then click New Actor Property at the top right of the window.
- At the top of the page, enter a unique name for the actor property. We named our property Movies watched.
- Select an actor from the drop-down list. In this example, we selected user. The options available to you reflect your data.
- Next to Method, make sure that Show is selected. Choose the appropriate options from the drop-down lists. In our example, we chose count of events.
- Make selections for Filtered to options. We chose events with action that matches watch_movie.
- Click Time options to specify a trailing window and time range (for more information, see specify relative time in a query). We specified a trailing window of 1 week, and accepted the default time range values of Beginning of time until Now.
- Click the Save in the upper right corner of the page, and then click GO. We received the following results for our example actor property.
Our results are rather busy. We could filter over a shorter time range, or click an individual line to filter the results and further explore the behavioral trends in the data.