How to Use Google Analytics Behavior Flow

Edited on Jun 23, 2026.

Google Analytics is the standard web analytics platform across the majority of websites globally. The Behavior Flow report inside Google Analytics is one of the most useful — and most underused — visualizations in the platform. It maps how users actually move through a site, where they drop off, and where they convert.

This is the working reference on what Behavior Flow shows, how to read it, and how to use it to actually improve site performance.

What Behavior Flow does

Behavior Flow is a visualization inside Google Analytics that tracks the path users take through a website. The report shows the sequence of pages users view, branching to display where they continue and where they exit. The visualization is read left to right, with the entry pages on the left and the subsequent interaction nodes flowing to the right.

Each node in the visualization represents either a single page or a group of related pages. The connections between nodes show how many users moved from one to the next. The red drop-off bars on each node show how many users exited the site at that point.

How to read the report

Three patterns are worth looking for.

Drop-off concentration on specific pages. If a single page in the user flow shows a consistently high drop-off rate, the page is either failing to answer the user's question, failing to lead the user to the next action, or producing a technical experience that makes users leave. The drop-off itself is the diagnostic — the cause requires further investigation.

Flow asymmetry. If users entering on landing page A consistently flow to conversion, while users entering on landing page B consistently exit, the entry-page experience is producing different downstream behavior. The asymmetry indicates an opportunity to either improve B or redirect traffic toward A.

Unexpected paths. Behavior Flow regularly surfaces user paths that the site team did not anticipate. Users entering on product pages and flowing to the blog. Users entering on the homepage and immediately going to the FAQ. The unexpected paths often reveal a mismatch between how the site is structured and how users actually want to use it.

What Behavior Flow is good at

• Identifying high-traffic drop-off points. Pages where many users exit produce the highest leverage for improvement.
• Comparing entry-page performance. Side-by-side comparison of which entry pages produce the strongest downstream engagement.
• Surfacing path patterns. Common sequences of pages that users follow, which the site team can use to optimize navigation.
• Tracking changes over time. Comparing Behavior Flow before and after site changes reveals whether the changes actually affected user behavior.

What Behavior Flow is not good at

• Granular conversion attribution. The visualization shows path patterns, not specific conversion attribution. For that, use the Multi-Channel Funnels report or Attribution Modeling.
• Cross-device journeys. Behavior Flow tracks sessions, not users across devices. Cross-device behavior requires User ID implementation and the cross-device reporting that depends on it.
• Specific event tracking. Behavior Flow tracks pageviews and grouped content categories, not specific button clicks, video plays, or form submissions. Event Flow is the parallel visualization for those interactions.
• Small-sample analysis. Behavior Flow becomes less reliable as the sample size shrinks. For low-traffic sites or low-traffic page groups, the visualization can produce misleading patterns from small numbers of users.

Setting up Behavior Flow for useful analysis

The default Behavior Flow report is often noisy. Four configurations make it more useful.

1. Apply an advanced segment. Filter the visualization to specific user segments — new users, returning users, mobile users, converters, non-converters — to see how the flow patterns differ across populations.
2. Group content into categories. Content groupings let you visualize flow at a higher level than individual pages. Group blog posts together, product pages together, support pages together. The category-level flow is easier to read than the page-level flow on sites with many URLs.
3. Filter by traffic source. The flow patterns for organic search traffic, paid search traffic, email traffic, and social traffic are usually different. Filtering by source reveals the source-specific patterns.
4. Define your own starting and ending nodes. The starting Node configuration lets you focus the visualization on flows that begin or end at specific points — useful for understanding a specific conversion funnel or a specific entry-page experience.

What to do with what Behavior Flow reveals

Two operational patterns produce the most value from Behavior Flow analysis.

Address the highest-traffic drop-off points first. Improvement leverage is proportional to traffic volume. A page where 10,000 monthly visitors drop off is a higher-priority improvement target than a page where 100 monthly visitors drop off, even if the percentage drop-off rate is similar.

Test the unexpected paths. If users are flowing to a page the site team did not expect them to visit, the site team should investigate why. Sometimes the path reveals an information need that the existing structure doesn't serve. Sometimes it reveals an opportunity to add navigation or links that align with how users actually use the site.

FAQ

How is Behavior Flow different from the regular Behavior reports?
The regular Behavior reports show aggregate metrics for each page (pageviews, time on page, bounce rate). Behavior Flow shows the sequence of pages users visit, which is information the aggregate reports do not provide.

Does Behavior Flow work for mobile sites?
Yes. The report works the same way regardless of device type. Filter by device category to see mobile-specific flow patterns.

What does the red drop-off bar mean?
The red bar on each node represents the percentage of users who exited the site at that page. Higher red bars indicate higher drop-off; lower red bars indicate users who continued to the next page.

Can Behavior Flow track events as well as pageviews?
Events have their own parallel visualization called Event Flow, which works the same way as Behavior Flow but for tracked events rather than pageviews.

How often should I review Behavior Flow?
At least monthly, with additional reviews after any significant site changes. The patterns evolve over time as content, traffic sources, and user behavior change.