---
title: "Analyze user sessions | Grafana Cloud documentation"
description: "Learn how to analyze real user sessions recorded through Frontend Observability"
---

> For a curated documentation index, see [llms.txt](/llms.txt). For the complete documentation index, see [llms-full.txt](/llms-full.txt).

# Analyze user sessions

User sessions help you understand the complete user journey through your application, from initial page load through interactions, errors, and backend API calls. Each session groups all events from a single user’s visit, so you can trace what led to errors or performance issues.

Browser, user-journey, and UX related events recorded by the Faro Web SDK are associated with a unique session ID to enable deeper insights into erroneous conditions and why they occurred.

## Before you begin

- Your application must be instrumented with the Faro Web SDK
- For trace correlation features, add the web-tracing package to Faro (included in the Frontend Observability plugin setup code)

You can find all recorded user sessions on the **Sessions** page. The sessions list can be filtered by session properties.

## Sort and filter sessions

You can filter sessions by properties and meta data in the session panel.

For example, filter on the following meta data to get quick insights into users’ behavior:

- `page_url` - the full url where the event was recorded
- `browser_*` - meta attributes about the browser and the operating system
- `event_*` - meta attributes related to a recorded event

Click on the column headings to sort the data. Subsequent clicks cycle through the sort order: descending, ascending, and not sorted. For example, click the “Errors” heading to order the sessions by error count in descending order.

Use the search field to narrow the scope of the table to only show rows which data includes the given search term.

## Analyze a single user session

Click on a **Session Id** to view the session details.

On the Session Details view, you can learn more about: the user session attributes, user journey events, and backend traces that were created as part of the user session.

## Inspect user journey events

The user journey table contains all events that were automatically recorded and manually created:

- `session_start` - A new user session was started
- `session_resume` - A user session was resumed after a navigation
- `sessions_extend` - A user session reached its maximum activity time or maximum lifetime and a new session was automatically created
- `view_changed` - Indicates that the user transitioned into another view
- `faro.performance.navigation` - Contains data about the performance of a navigation such as network and render timings
- `faro.tracing.fetch` - A trace event recorded for a fetch request send when the web-tracing package is added
- `faro.tracing.xml-http-request` - A trace event recorded for a xml-http-request request send when the web-tracing package is added
- `route_change` - A React Route change event send when the faro-react is used

Signal types:

- `measurement` - Represents a measurement signal. For example measurements for the different web-vitals
- `exception` - Represents an error that occurred
- `log` - Represents a log message

Click on a single event to view the raw event log or a single error view, depending on the event kind.

## Continue investigation in other parts of Grafana Cloud

To gain end-to-end insights the User Journey table provides several actions to continue the investigation in other parts of Grafana Cloud and to follow a requests throughout the system.

> Note
> 
> It is mandatory to have the web-tracing package added to Faro. The Faro setup code provided by the Frontend Observability plugin has it integrated already.

The User Journey table provides different actions to continue the investigation in other places:

- The `Trace ID` column.
  
  If there’s a trace ID available for a signal, it shows up in the `Trace ID` column. Following the trace ID link opens the Grafana Cloud Trace Explore view.
- The `Explore Column`
  
  The `Explore Column` provides the most relevant action fo a specific row.
  
  - `Traces` opens Trace Explore with the respective trace ID selected
  - `Logs` opens Logs Explore filtered to show only logs for the specific signal
  - `Errors` opens the specific error details in the **Errors** application
  - `Services` opens the Traces view for the specific trace in Application Observability
    
    > Note
    > 
    > The `Services` action is only visible if there are server spans found for the given trace ID. If it is a client only span, the `Traces` action shows up instead.

## Inspect traces produced as part of a user journey

If the tracing instrumentation is activated, the Faro Web SDK records interactions with backend services as trace data. Those interactions have an additional `session_id` attribute and allow for easy discovery of slow backend transactions.

The traces inspect table provides a full search editor if you need to narrow the scope of the displayed items.

## Inspect and understand navigation performance

An overview about the time spent in the different phases of loading and rendering a page is available in the user journey table under the `faro.performance.navigation` event. Press the chevron on the left side of the row to reveal the navigation timings panel.

The navigation timings panel provides insights to the load and rendering performance of the page alongside other useful information.

#### Summary

The summary row provides an overview about the whole duration of the navigation. This is useful to see at a glance about how much time each part takes in the whole navigation.

Metrics shown in this row:

- `Navigation Duration` - how long the whole duration takes
- `Page Load Time` - the time to fetch and render the resource
- `Time To First Byte (TTFB)` - the time between the request for a resource and when the first byte of the response arrives
- `Network` - the time it takes from making the request till the response is loaded
- `Render` - time to parse and build the page (contains load handler execution time too).

It also shows useful metadata about the resource load like:

- `Cache Hit Status` - tells if a resource is directly loaded from cache, loaded by a conditional fetch or is a full load.
  
  Possible values are:
  
  - `cache`: direct cache hit
  - `conditionalFetch`: loaded via a 304
  - `fullLoad`: loaded from the server
- `Protocol` - The network protocol used to fetch the resource
- `Body (Compressed | Uncompressed)` - The size of the request body compressed and uncompressed
- `Visibility State` - whether the navigation happened in the foreground or background.
  
  This useful because browsers prioritize visible/foreground work and tabs loaded in the background are usually slower.
  
  Possible values are: `visible` and `hidden`

#### Network

The network row provides insights into how long each step involved loading a resource takes.

Metrics shown in this row:

- `Redirect` - the time to follow all redirects
- `Service Worker` - Service Worker processing time
- `DNS Lookup` - DNS lookup time
- `TCP Handshake` - TCP handshake time
- `TSL Negotiation` - TLS negotiation time
- `Request` - the time it takes send the request
- `Response` - how long it takes to read the response

#### Render

- `Document Parsing Time` - the time to parse the HTML document and to construct the Document Object Model (DOM)
- `Dom Processing Time` - the time to load and process all resources after the document has been parsed and the Document Object Model (DOM) is build.
- `OnLoad Time`: the time it takes to process all registered load event handlers

## Next steps

After analyzing sessions:

- [Review errors](/docs/grafana-cloud/monitor-applications/frontend-observability/visualize-data/errors/) to see which issues are affecting the most sessions
- [Monitor performance](/docs/grafana-cloud/monitor-applications/frontend-observability/visualize-data/performance/) to identify slow page loads or interactions
- [Analyze HTTP requests](/docs/grafana-cloud/monitor-applications/frontend-observability/visualize-data/http-insights/) to troubleshoot backend integration issues
- [Filter by geolocation](/docs/grafana-cloud/monitor-applications/frontend-observability/visualize-data/geolocation/) to understand regional user behavior patterns