Skip to main content


Interana Docs

Example: Analyze user sessions with flows

This article explains how you can use flows to explore the details of user sessions, such as the number of comments each user makes, how often they make a purchase or perform other conversions, or how long they stay engaged on your site.

Using flows to analyze user sessions

This section demonstrates how to explore user sessions for actor events, such as the number of comments made, the number of errors per session, as well as periods of inactivity. To accomplish this, you first create a new flow, select the actor (shard) with the data you want to analyze, then run the query.

After reviewing the results, you explore the data with more granularity by creating a flow property and use it as a filter on the flow.

To analyze user sessions with a flow, do the following:
  1. Open Explorer, select a dataset in the upper left corner of the window, click Flows, then enter a name for the flow at the top of the page. We chose the SaaS dataset for our example.


  1. Click New Flow in the upper right corner, select an Actor (shard) from the drop-down list. Optionally, you can enter a name for Step 1. We entered user_id for our example.


  1. Click inside no timeout for Flow ends after inactivity of and enter 30 minutes. Optionally, you can click the minus  to delete Steps you won't need.


  1. Click GO and review the results. We decided we want to refine our results by creating a property filter.


  1. To create a flow property, click Properties in the upper right corner. Then click New Flow Property, enter a property name, specify property details, and click Save in the upper right corner. Our flow property filters for errors per session.


  1. To apply the flow property, click Explore and in the left panel choose to Show the new flow you created, filtered to the new flow property, specifying an appropriate value. We filtered for accounts with greater than 2 errors per session.


  1. Click GO, and then click the Sankey View icon. We got the following results for our example.


  • Was this article helpful?