Traffic
See how many people read your site, which pages they land on, what they search for, and what they think of your content
The Traffic report covers the people reading your documentation. Machines reading alongside them, such as LLMs, agents, and crawlers, are counted separately in Agents & MCP, so the figures here reflect human readers.
To open it, go to Analyze → Traffic in your site’s sidebar.
At the top, you’ll find page views, visitors, sessions, and searches for the period you selected, each compared to the period before, followed by a chart of readers over time. See Site analytics for filters, time periods, and comparing periods.

Pages and feedback
Top pages ranks the pages your readers open, so you can see what your site is used for.
Next to it, Page feedback shows how readers rate your content once you’ve enabled page rating in the Customize menu, along with the comments they left. Low-rated pages, read together with their comments, are usually the fastest place to start improving your docs.
Filter and group feedback data
Use filters to focus on a subset of ratings and comments, and groups to compare that data across a dimension. You can filter or group feedback by these dimensions:
Content: section, variant, and page.
Visitor dimensions: country, language, device, browser, referrer, and authenticated visitor status.
Visitor claims configured for adaptive content, such as a customer, plan, role, or feature access.
Content filters identify where feedback comes from. Visitor dimensions add the same audience context available elsewhere in site analytics. For example, filter a page’s ratings by country, or group its feedback by device or authenticated visitor status.
The Authenticated visitor filter shows whether GitBook identified the visitor when they left feedback:
Authenticated: The visitor signed in or arrived with signed visitor claims.
Anonymous: The visitor arrived without signed identification. Claims passed through URL parameters or a public visitor cookie are unsigned.
Unknown: GitBook recorded the event before tracking visitor authentication began in August 2026.
Feedback filters help you turn ratings and comments into focused updates. For example, you can:
Identify pages or sections with lower ratings, then review their comments.
Compare feedback from authenticated and anonymous visitors to find audience-specific gaps.
Check ratings for pages that vary by customer claim, then improve the content for that segment.
To use or analyze this data outside of GitBook, open the card’s menu and click Download CSV. The export has one row per rating, with the comment, a link to the page, and the time the rating was left.
Search
The Search card shows what readers typed into your site’s search. Use it to find out what people expect to exist, and which searches your content doesn’t answer yet. That is a good signal for what to write next, or for content that exists but is hard to find.
Where your readers come from
The rest of the report breaks your audience down:
Platform, Countries, and Languages: who your readers are and what they read on.
Traffic sources: the domains and URLs sending you visitors.
Campaigns: traffic broken down by
utm_source,utm_medium,utm_campaign,utm_term, orutm_content.Active hours: when your site is busiest, as a heatmap across the week.
Bots: page views from crawlers, kept out of the reader figures in this report.
Sites using adaptive content also get a Visitor attributes card, showing traffic by the claims you pass, for example a customer, plan, or role.
Last updated
Was this helpful?