The Spark AI reports give you visibility into how participants are using Spark in your programs, including who has gotten started, who still needs a nudge, and which role-play simulations people are practicing. Follow these steps to access the reports, filter the data, and learn more about what the columns mean.
There are two Spark AI reports:
-
Spark AI Engagement: A summary with one row per participant. Use it to track adoption and see who hasn’t started using Spark yet.
-
Spark AI Activity: A detailed log with one row per Spark conversation or role-play session. Use it to see when participants are using Spark, how long their sessions last, and which role-plays they’re practicing.
Note: To protect participants’ privacy, neither report includes conversation content, transcripts, conversation themes, or role-play skill feedback. You’ll only see activity data, like dates, session counts, and session length.
Accessing the reports
From the left-hand navigation, click Reports > Spark AI Engagement or Reports > Spark AI Activity.
You’ll see data for the programs and cohorts you have admin access to. Admins with owner-level access will see all Spark AI programs and cohorts in your organization.
Filtering the data
By default, you’ll see data for all the Spark AI programs and cohorts you have admin access to. To focus on specific information, use the Programs & Cohorts filter at the top to select the programs and cohorts you want to see.
In the Spark AI Activity report, you can also use the Activity date range filter to narrow the results to a specific timeframe. Choose a preset range, like the past 28 days, or pick custom start and end dates from the calendar. Dates are based on your own time zone.
The report data will update to reflect your filter selections. Each report shows 10 rows per page, and you can click a column name to sort the full table in ascending or descending order.
Torch tip: Want to save the report data? Click Download data in the top-right to download a CSV file of the report. The download includes all rows that match your current filters, not just the page you’re viewing.
Keep in mind that report data refreshes a few times a day, so very recent activity may take several hours to appear.
Understanding the Spark AI Engagement report
Each row of the table represents a participant in a Spark AI cohort. Everyone in the cohort is included, even participants who haven’t accepted their invitation or used Spark yet, so you can easily see who may need a reminder. A participant may show up more than once if they’re in more than one Spark AI cohort.
For each participant, you’ll see:
-
Organization name
-
Program name
-
Cohort name
-
Participant name
-
Participant email
-
Status: Where the participant is in the program. See the status definitions below.
-
Invited date: The date the participant was invited to the program.
-
Total Spark conversations: The number of coaching chats the participant has started with Spark.
-
Average conversation length: The average length of the participant’s Spark conversations, in minutes.
-
Total role-play sessions started: The number of role-play simulations the participant has started.
-
Total role-play sessions completed: The number of role-play simulations the participant has finished.
-
Last login date: The most recent date the participant logged in to Torch, shown in your time zone.
Participant statuses
-
Invite scheduled: The participant has been added to the program with a future invite date, and their invitation hasn’t been sent yet.
-
Invited: The participant’s invitation has been sent, but they haven’t opened their Spark AI program in Torch yet. This includes participants who already have a Torch account from another program.
-
Active: The participant has opened their Spark AI program in Torch at least once. This doesn’t necessarily mean they’ve used Spark yet, so an active participant can still show 0 conversations or role-plays.
-
Completed: The participant was active when the program ended.
-
Did not participate: The participant was still in an invited status (they never opened the program) when the program ended.
If a program is reopened, participants move back to Active or Invited. Participants who have been removed from a program don’t appear in the report.
Torch tip: To find participants who need a nudge, sort by Status to group everyone still in an Invited status, or sort by Total Spark conversations to find Active participants who haven’t started a conversation yet.
Understanding the Spark AI Activity report
Each row of the table represents a single Spark conversation or role-play session. Participants will appear once for each session they’ve had, and participants with no Spark activity won’t appear in this report. Use the Spark AI Engagement report to see everyone in your cohorts, including those who haven’t started yet.
For each session, you’ll see:
-
Participant name
-
Participant email
-
Program name
-
Cohort name
-
Activity date: The date the session started, shown in your time zone.
-
Activity type: Whether the session was a Spark AI conversation or a Role-play.
-
Role-play title: The name of the role-play simulation the participant practiced. This is blank for Spark AI conversations.
-
Modality: Whether the participant used Text or Voice for a Spark AI conversation. This is blank for role-plays.
-
Status: Whether the session was Started or Completed. Role-plays are marked completed when the participant finishes the simulation. Because participants can always return to and continue a Spark conversation, conversations don’t have a set end point.
-
Total session time: The length of the session, in minutes.
Torch tip: Download the Spark AI Activity report and sort by Role-play title to see which scenarios are most popular with your participants. This can help you spot common development needs across your cohorts.