Use Gemini Enterprise’s Analytics dashboard to measure adoption and agent activity, Cloud Monitoring to track operational performance, and Cloud Logging to investigate detailed events when needed. These answer different questions: activity is not the same as business value, and detailed logs can expose sensitive information.
For a useful report, combine seat reach, active users, task-relevant engagement, agent sessions, quality signals, reliability, and a workflow outcome your organization defines. Interpret each measure using its counting rules, reporting period, and access limits.
As an Amazon Associate I earn from qualifying purchases.
Start with the question each data source answers
| Source | Best for | Important qualification |
|---|---|---|
| Gemini Enterprise Analytics | Adoption, engagement, search and answer quality indicators, and agent usage | Google says Analytics metrics refresh about every six hours; metrics can take several hours to appear after app creation. Google’s Analytics documentation |
| Cloud Monitoring | Application and agent traffic, latency, errors, sessions, turns, and tool activity | Metric data refreshes every minute. Agent-level metrics require the relevant instrumentation setting. Google’s metrics documentation |
| Cloud Logging | Investigating detailed usage events | Prompt and response content can be sensitive, so access and logging settings need governance controls. Google’s usage audit log documentation |
The Analytics documentation describes its purpose as providing “insight into the usage trends, search quality, and end-user engagement of your app.” Its metrics are indicators to interpret, not a universal adoption target or proof that usage caused a business result.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Measure adoption and license reach
In the app, open Analytics → Adoption. Google documents total users; users calling assistant, answer, or search APIs; daily, weekly, and monthly active users; 7-day and 28-day retention; growth and churn rates; and seats purchased and claimed. A unique user who calls multiple API methods counts once in total users.
#1 Best Overall
Interpret active users carefully
Gemini Enterprise defines an active user broadly: a user qualifies after signing in and making any API call. Landing on the home page can trigger such a call. Therefore, active users are a reach or activity signal—not a count of people who completed a meaningful task. Pair the figure with narrower measures such as assistant, search, or answer users and with a locally defined workflow outcome.
Compare seats and usage without mixing time bases
Purchased and claimed seats are cumulative since registration, while active users are counted for a selected period. Show them as a rollout funnel—seats purchased → seats claimed → active users—but label the different time bases and report the active-user date range. A claimed seat indicates assignment or acceptance, not meaningful use.
Google’s weekly active-user week runs Monday through Sunday. When comparing retention, growth, or churn, state the reporting window and keep periods and populations consistent; otherwise, a difference may reflect the comparison window rather than a change in adoption.
Recommended Free Tools
Read search, answer, and feedback metrics as signals
The Usage and Quality view includes search count, answer count, action count, click-through rate per search, successful search rate, successful answer rate, and feedback like/dislike counts and reasons. One important scope limitation: the dashboard’s answer count excludes answers generated by an agent, so it is not a complete count of agent-generated answers.
Use these figures to spot patterns worth examining, then evaluate them against your team’s definition of task completion and a suitable review or sampling process. Google’s dashboard definitions do not prescribe an organization-wide quality standard or business outcome.
Measure agent adoption without overstating it
The Agent view includes monthly active agent users, monthly agent chat sessions, monthly total chat sessions, monthly agents used, and monthly agents created. These are useful for understanding reach and repeat interaction, but their definitions affect how to read them.
Know what counts as agent activity
A user counts as active for an agent after an Assistant call in agent, research, or idea forge mode. The active-agent-user measure can include the agent creator’s Workflow Builder preview and test calls because they use the same Assistant path. Preview and test sessions also count in session measures; unsaved drafts without an agent ID are excluded.
Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Clear out junk files and repair common Windows errors3Scan for outdated or missing drivers - takes under a minuteUnderstand sessions and short-lived calls
A session is a user interacting with an agent in the same chat thread on a single day. If the conversation continues in that thread the next day, it counts as a new session. Monthly agents used can occasionally exceed monthly total chat sessions when short-lived calls—such as cancelled turns or one-shot calls—do not produce a persisted session ID.
When your data permits, separate builder testing from employee usage, annotate preview activity and agent lifecycle changes, and report both unique active users and sessions. Compare agents over the same period and audience; raw counts are misleading when exposure, audience size, preview activity, or reporting intervals differ.
Rank #4
Use Cloud Monitoring for operational performance
App-level Engine metrics
Use the Gemini Enterprise Engine resource for app-scoped traffic and latency. The documented request-count metric can be broken down by API method, response code, query type, and model. The total-latency metric includes requests that return an error. Time-to-first-answer is recorded only when an answer token is produced, while time-to-first-token is recorded only when a token is streamed.
Agent-level metrics and Observability
Use the Gemini Enterprise Agent resource for per-agent sessions, turns, tool invocations, and latency. For supported agents, the console’s Observability view provides overview, tools, latency, and error-rate views. Tool measures include call counts, P95 duration by tool, error rate, and the rate at which no tool is called. Google documents Observability tab availability for Core Assistant, Workflow Builder, and Deep Research agents.
Enable the required instrumentation and access
App-scoped metrics are collected without the agent instrumentation toggle. Agent-scoped metrics require Enable instrumentation of OpenTelemetry traces and logs at the app level for Core Assistant, or at agent level for the documented employee-made Workflow Builder and Deep Research agents. Metrics Explorer also requires Monitoring Viewer access in addition to the documented Gemini Enterprise roles. Check the app and agent configuration before treating an absent agent metric as zero activity.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Use logs and analytics access with appropriate controls
Usage audit logs can identify activity and expose detailed fields. Google classifies prompt or query content as sensitive in the documented service paths. With prompt/response logging disabled, text values are elided and other sensitive values omitted; enabling it logs full prompt and response content, which may include personal information. Limit Cloud Logging access to authorized roles and enable content logging only when company policy supports it. Aggregate adoption reporting generally does not require prompt-content logging.
The Analytics User Level tab includes user IDs and interactions but requires allowlisting. For US and EU multi-region apps, Google documents search analytics as Public Preview; search analytics are generally available for global apps. Availability can depend on app configuration and location, so verify access and regional status with your Google Cloud administrator.
Build a report that supports fair comparisons
For each reporting period, capture the date range and timezone, app location and edition, and the measures below. Keep population and period consistent when comparing teams, agents, or time windows.
- Reach: purchased and claimed seats, total users, and task-relevant active users. State that the active-user definition includes any API call.
- Repeat use: retention, monthly active agent users, and sessions, with the same-day session boundary in mind.
- Breadth and depth: agents with any active user, plus users and sessions by agent; annotate preview and test activity and short-lived calls where possible.
- Usefulness and quality signals: successful search and answer rates and feedback, paired with a task-specific outcome your organization defines.
- Reliability and speed: relevant error rates and latency distributions, including tool-specific errors where available.
- Governance and coverage: user-identifiable views, access roles, regional availability, data refresh and retention, instrumentation state, and whether prompt/response logging is enabled.
The documented Gemini Enterprise Monitoring metric series use the discoveryengine.googleapis.com/ prefix and fall under Google’s “all other Google Cloud metrics” tier, which has six weeks of default retention. Older data is automatically deleted under that default. Confirm your current retention settings and export needs before relying on long-range trend reporting.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
