
Every analytics platform uses a session window to group user actions into a single visit. When someone reads three articles in ten minutes, that’s one session. When they return six hours later, that’s a new session—even if they never closed the tab.
The timeout threshold is arbitrary, platform-specific, and rarely examined. And when it expires mid-journey, attribution breaks, conversion funnels fragment, and you optimise for ghosts.
What resets the session clock
Google Analytics 4 uses a 30-minute inactivity window by default (checked November 2026). If a user stops interacting for 30 minutes, the next pageview starts a new session. The timer resets with every event—scroll, click, page load.
Plausible Analytics doesn’t use session cookies at all; it infers sessions by grouping pageviews from the same visitor within a continuous window, also defaulting to 30 minutes of inactivity (checked November 2026).
Fathom Analytics mirrors the pattern: 30 minutes idle, new session (checked November 2026).
Matomo lets you configure the timeout in settings, with a default of 30 minutes (checked November 2026).
The pattern is consistent: 30 minutes is the industry anchor. But real user behaviour doesn’t respect it.
Where journeys split silently
A reader opens your article at 9:00 AM, skims the intro, switches tabs to Slack, returns at 9:35 AM, finishes reading, and subscribes. Two sessions. The subscription gets attributed to a direct visit with no referring source, even though the original session came from organic search.
A podcast listener clicks your link at lunch, leaves the tab open while they eat and walk the dog, returns 90 minutes later, and buys your course. Three sessions, depending on background tab behaviour and whether the browser suspended the page.
Long-form content, tabbed research, mobile interruptions, and email-to-web handoffs all push past the 30-minute threshold. And every break severs attribution.
If your content is designed for deep reading or your audience multitasks heavily—common for B2B operators, researchers, and remote workers—the default timeout undercounts assisted conversions and inflates direct traffic.
When to adjust the window
Google Analytics 4 lets you extend the session timeout to a maximum of 7 hours and 55 minutes via the data stream settings (checked November 2026). Navigate to Admin → Data Streams → [your stream] → Configure tag settings → Adjust session timeout.
Matomo exposes the timeout in Settings → Websites → Manage → Edit → Global list of Query URL parameters to exclude, under the advanced section where you can also set the visit standard length (checked November 2026).
Extending the window reduces session fragmentation but introduces new problems. A four-hour timeout will group a morning research session with an afternoon purchase, even if the user genuinely left and returned with fresh intent. You’re trading one attribution error for another.
The right threshold depends on content length and format. If your average article takes 12 minutes to read and your audience rarely background-tabs, 30 minutes is fine. If you publish 3,000-word guides that readers save for later, consider 60 or 90 minutes. If you send a weekly email that drives traffic in waves throughout the day, a longer window helps—but you’ll still see session breaks across email clients that strip or rewrite parameters.
What to watch instead
Rather than guess at the ideal timeout, track metrics that surface the problem. In GA4, compare Engaged sessions (at least 10 seconds or two pageviews) against total sessions. A widening gap suggests users are landing, leaving, and returning—each return counting as a new, unengaged session.
Check your Source/Medium report and filter conversions by “direct / none.” If direct traffic accounts for more conversions than top-funnel sources like organic or email, session timeouts are likely severing the attribution chain.
In Matomo, review the Visitors → Visits Log for individual user journeys. Look for the same visitor ID across multiple short sessions with identical referrers or UTM parameters—evidence that one continuous journey was artificially split.
Session timeout is a reporting construct, not a law of physics. Adjust it when your data shows fragmentation, not because a blog post told you to. And remember: no timeout setting will fix attribution if your users switch devices, clear cookies, or click links from privacy-focused email clients that strip query strings entirely.
One Two Three Send covers analytics, attribution, and the infrastructure behind online businesses every week. Subscribe here to catch the next one.