Event Tracking Reference
Every way an event can be scoped, aggregated, and broken down when you build an Insights, Funnel, or Retention report in Analytics.
Overview
Every Analytics report, Insights, Funnel, or Retention, is built around one or more events. This reference lists every option available for referencing those events: which population a report measures, how an event's occurrences get aggregated into a single number, and how you can break that number down into groups. For the step-by-step report builder walkthroughs, see Reports. For the full list of autotracked, custom, and source-created events, see Event types.
Profile types
Every report scopes its data to one profile type.
| Profile | API value | Scope |
|---|---|---|
| Users | users | All users, identified and anonymous. |
| Accounts | accounts | Grouped by the account (company or team) a user belongs to, instead of individual users. |
📘 Good to know
An account is an object a set of users can belong to, such as a company of customers or a team of users. Switching a report's profile to Accounts lets you analyze how those groups interact with your product instead of seeing individual users.
Insights: event aggregation types
An Insights report plots one of these aggregations for each event series you add.
| Aggregation | API value | What it measures |
|---|---|---|
| Total events | total | The total number of times the event occurred in the selected period. |
| Unique sessions | unique_sessions | The number of distinct sessions in which the event occurred. |
| Unique users | count_unique | The number of distinct users who performed the event at least once in the period. Labeled "Unique accounts" when the report's profile is Accounts. |
| Count per user | total_per_user | The event's occurrences per user. Labeled "Count per account" when the report's profile is Accounts — it's the same aggregation, not a separate one. |
| Active users | active_users | The number of distinct users who performed the event within the report's active-user window. Labeled "Active accounts" when the report's profile is Accounts. |
| Property value | property_value | An aggregation of a numeric event attribute, using one of the functions below. |
Property value functions
When an event series uses the property_value aggregation, choose one of these functions to combine the attribute's values:
| Function | API value |
|---|---|
| Total | total |
| Sum | sum |
| Average | average |
| Median | median |
| Minimum | minimum |
| Maximum | maximum |
| 25th percentile | percentile_25th |
| 75th percentile | percentile_75th |
| 90th percentile | percentile_90th |
| 95th percentile | percentile_95th |
| 99th percentile | percentile_99th |
📘 Good to know
Property value aggregation only works on numeric event attributes.
Funnels: measurement types
| Measurement | API value | What it measures |
|---|---|---|
| Conversion rate | conversion_rate | The rate at which users complete the selected funnel steps. |
| Total conversions | total_conversions | The total number of times users reached the selected step, allowing a user to re-enter and convert more than once. |
| Unique users | unique_users | The number of distinct users who completed the selected steps, based on their first attempt at entering the funnel within the report's timeframe. |
| Time to convert | time_to_convert | How long users took to move through the funnel, aggregated with average, median, minimum, maximum, or a percentile. |
A funnel always measures conversion across every step you've defined, from the first step to the last — there's no way to narrow it to a sub-range of steps.
Retention: cohort windows
A Retention report groups its cohorts and measures return activity using one of these windows:
| Window | API value |
|---|---|
| Daily | day |
| Weekly | week |
| Monthly | month |
Measuring is set separately to either Retention rate (the percentage of the original cohort that returned in each period) or Unique users (the raw number of users who returned).
Breakdowns
A breakdown splits a report's chart and table into one line or row per group.
| Breakdown type | API value | Description |
|---|---|---|
| Attribute | attribute | Splits by a user or account attribute's values. |
| Event property | collection | Splits by a property recorded on the event itself. |
| Segment | segment | Splits by membership in one or more segments. |
📘 Good to know
Insights, Funnel, and Retention reports all support up to two breakdowns at a time.
Time grouping
Insights and Retention reports bucket their X-axis into one of these intervals:
| Interval | API value |
|---|---|
| Day | day |
| Week | week |
| Month | month |
| Year | year |
Use cases
- Tracking active engagement. Use the Active users aggregation on your core action event instead of building a separate active-user calculation.
- Comparing user-level vs. account-level adoption. Switch a report's profile type from Users to Accounts to see how companies, not individuals, use your product.
- Measuring revenue per event. Use the Property value aggregation with the Sum function on a price attribute to total revenue instead of counting events.
- Choosing conversions vs. re-entries. Use Total conversions when users can legitimately complete a funnel more than once, and Unique users when you only want each user's first pass counted.
- Matching retention windows to usage cadence. Use a Daily window for products used every day, and Weekly or Monthly for less frequent ones.
- Analyzing session-level frequency. Use the Unique sessions aggregation instead of Unique users when a single user's repeat visits should each count separately.
- Comparing two dimensions on one retention report. Add two breakdowns, for example plan tier and acquisition channel, to see cohort return rates across both at once.
Where to go next
- Analytics overview for what reports and dashboards do.
- Creating an insights report to walk through building one of these.
- Creating a funnel report for the step-by-step funnel builder.
- Creating a retention report for the step-by-step retention builder.
- Event types for the full list of autotracked, custom, and source-created events.