Skip to main content
Interania

Creating a 2.x cohort as a 3.x actor property

0votes
23updates
54views

Cohorts are a way to group actors that behave in some similar way. Cohorts segment users based on their behavior and attributes within a time period, enabling both behavioral and demographic segmentation.

3.x actor properties are similar in function to 2.x cohorts, but much more flexible. Once you create an actor property, you can use it as a filter in an infinite number of queries. To attain the same results in 2x, you have to create individual cohorts for every different use case.

  • In Interana 2.x, you define a cohort of actors based on their attributes and associated events. You can then use the cohort in filters to ask targeted questions about subsets of the actors. 
  • In Interana 3.x, you create an actor property based on specific actor events over a set time. You can then use the actor property as a query filter for a group of actors.

Both cohorts and actor properties can be used to understand retention, cashflow, and lifetime value; to compare ROI for various customer acquisition channels; to map device batches to errors and defects.

Our example uses a movies dataset, and we discover users who watched movies over a specified time range and posted comments on the online movie service web site.

Goal

The following sections demonstrate how to accomplish these tasks with Interana 2.25 and Interana 3.0 for comparison, so you'll understand the differences:  

  1. Create a 2.x cohort for use in queries.
  2. Create a 2.x cohort query.
  3. Create a 3.x actor property for use in queries.
  4. Create a 3.x actor property query.

To get exactly the same results across Interana versions, we need to carefully match the way time is selected. In Interana 2.x, we will choose "180 days ago to today" which is calendar-aligned to the beginning of the current day. In Interana 3.x, we will choose "180 days ago to now" which is also calendar-aligned to the beginning of the current day.

Step 1: Create a 2.x cohort for use in queries

This step shows you how to create a 2.x cohort that we'll then use in a query. 

Cohorts serve as shorthand for a group of actors that you're interested in studying. Instead of constantly having to describe the actors to include in a query, they're described once in the cohort definition. Interana cohorts are focused on behavior. Actors are included in cohorts based on their actions, not just their demographics. 

In the following example, we create a cohort that filters for all users who watched at least one movie over the last 180 days, ending today. 

To create a 2.x cohort, do the following:
  1. Click the Cohorts icon in the left navigation bar, then in the upper right corner click New Cohort.

WB_Cohorts_icon.png

  1. In the New Cohort dialog, enter a unique Name and select a Dataset for the cohort. We entered Movies watched as the name and chose the movies dataset for our example.
  2. For each user, select a Measure. We selected Count Unique.
  3. Choose a filter and specify a value. We chose a Movies Watched and specified at least 1 for the value.

For information on how to create a per-actor metric like Movies Watched, see Creating a 2.x per-actor metric as a 3.x actor property.

  1. Specify a Start and End time range for the cohort. We entered 180 days ago and ending today for our example.
  2. Enter a Description for the cohort and then click Save.

WB_New-Cohort_dialog1.png

Step 2: Create a query with a 2.x cohort

This step demonstrates how to use our newly created 2.x cohort and 3.x actor property in similar queries. We use the Movies Watched cohort to construct a query that filters for users (count unique) who watched at least one movie (Movies watched cohort) and made a comment (Compare) on the online movie service site.

To create a query that filters for movies watched and comments made, do the following:
  1. From the cohorts page in the previous step, click the compass icon to the right of the Watched Movies name. The Explore window appears, showing the cohort values in the query builder fields on the left.

  1. Accept the cohort values that automatically populated the query builder fields.
  2. In the Compare section, we selected comment from the drop-down list. This will filter on all users who have commented on a movie.

WB_2x_Cohort-query_GO.png

  1. Click GO.

Step 3: Create a 3.x actor properties for use in queries

This step demonstrates how to create a 3.x actor properties that we'll then use in a query.

You can use actor properties to compare how actors behave in the same situation, or how the same actors behave in different situations. Wherever you have a group of actors that are similar to each other, you can benefit from actor properties.

In the following example, we'll create two actor properties:

  • Actor property that counts the number of movies watched over the last 180 days. 
  • Actor property creates the cohort of users who watched any number of movies over the last 180 days.
To create an actor property that counts movies watched over the last 180 days, do the following:
  1. Click the Actors icon in the left navigation bar, then in the upper right corner click New Actor Property.

3.x_Actors-icon.png

  1. In the upper left corner choose a dataset, then enter a unique Name for the actor property. We chose the movies dataset, and entered Movies watched over the last 180 days for the name of the actor property in our example.

WB_ActorProperty_dataset2.png

  1. Choose the Actor for your property. We chose user.
  2. On the Show tab, accept the default Show count unique values of movie and select from the drop-down list. Then Filtered to events with action that matches watch_movie.

WB_ActorProperty_matches-watch_movie.png

  1. Click the Time range radio button and specify a Starting and Ending time. We clicked the Starting text and entered 180 days ago, then clicked inside the Ending text and entered now

WB_ActorProperty_time-range2.png

  1. In the upper right corner of the window, click Save.

WB_ActorProperty_Save.png

To create an actor property that's a cohort of users who watched movies over the last 180 days, do the following:
  1. Click the Actors icon in the left navigation bar,
  2. then in the upper right corner click New Actor Property.

3.x_Actors-icon.png

  1. In the upper left corner make sure the dataset you used for the previous task is still selected, then enter a unique Name for the actor property. We chose the movies dataset, and entered Cohort of users who watch movies over the last 180 days for the name of the actor property in our example.

WB_ActorProperty_dataset2.png

  1. Click the Show tab and select Show count of events Filtered to all events. Then click the User time range from explorer check box.

WB_ActorProp_cohort-of-users1.png

  1. Click the Defined Value tab, and enter a value of Users who watched a movie in the last 180 days. Then select user actors with Movies watched in the last 180 days. This provides a filter similar to a 2.x cohort.

WB_ActorProp_Cohort-of-users-defined-value.png

  1. In the upper right corner of the window, click Save.

WB_ActorProperty_Save.png

  1. You can now construct a 3.x query with the 3.x properties.

Step 4: Create a query with the 3.x actor properties

This step demonstrates how to use our newly created 3.x actor properties in a query to generate results similar to the one used for the 2.x cohort.

In the following example, we construct a query using the Movies watched in the last 180 days actor property and Cohort users who watched movies in the last 180 days to that filters to users who watched at least one movie in the last 180 days.

To create a query that filters for movies watched and comments made, do the following:
  1. Click the Explore icon in the left navigation bar.

WB_3.x_Explore-icon.png

  1. In the first line query builder, select Show count unique of user actors.
  2. In the second line of the query builder, for Filtered to user actors with, enter Cohort of users who watch movies in last 180 days that matches Users who watched a movie in the last 180 days.

WB_3.x_actorprop_cohort_query.png

  1. Select the Starting text and enter 180 days ago. Then select the Ending text and enter now.

WB_ActorProperty-query_TimeRange.png

  1. Click GO

We received the following results for our example.

WB_3.x_actorprop_cohort_query-results1.png

  1. You can Split by user actor to view results for individual users.

We received the following results for our example.

WB_3.x_actorprop_query_splitby-results.png

  • Was this article helpful?