Exclude bot and AI bot traffic from all stats

Open in ChatGPT Open in Claude

Bot traffic is non-human activity on your website or app: search engine crawlers, SEO research tools such as AhrefsBot or SemrushBot, scrapers, uptime monitors, and — increasingly — AI crawlers and AI assistants. Left unfiltered, it inflates visitor and pageview counts and distorts engagement and conversion metrics.

Usermaven gives you a set of independent exclusions you can combine, so you can remove automated traffic without also removing AI bot traffic, and the other way round. This article explains each option on the Exclude Traffic page: what it does, how it behaves, and when to use it.

How to reach the Exclude Traffic page

Log in to Usermaven.
Click Workspace Settings in the top-right corner.
Select Exclude Traffic from the left-hand menu, under WORKSPACE SETTINGS.

The page is at /env/<workspace>/settings/exclude_traffic, and its browser tab title is Exclude Traffic. You need the workspace settings view permission to open it, and edit permission to change any option.

Usermaven Workspace Settings screen showing the left-hand settings menu with the Exclude Traffic entry selected under the WORKSPACE SETTINGS heading. — Exclude Traffic sits in the Workspace Settings menu, under Workspace Settings.

What is on the page

The page has five stacked cards. They are independent — turning one on does not turn another off. The AI bot card sits at the top, because it is the one people most often want to change on its own.

Card	What it filters
Exclude AI bot traffic from stats	AI crawlers and AI assistants, by category
Exclude other bot traffic from stats	All other detected bots, including historical bots with no AI classification
Exclude proxy traffic from all stats	VPN, TOR, datacenter, public proxy, web proxy and search engine robot traffic
Exclude server side events from website analytics	Events your server sends through the API
Exclude internal traffic and users	Your own team, test accounts, and development environments

The complete Usermaven Exclude Traffic settings page showing all five stacked cards with their toggles and options. — The full Exclude Traffic page: AI bot exclusion, other bot exclusion, proxy exclusion, server side events and internal traffic.

Exclude AI bot traffic

Turn on Exclude AI bot traffic from stats to keep selected AI bot activity out of your reports.

With the switch on, a panel expands with five checkboxes under Select AI bot activity:

Category	What it covers	Examples
Training	Crawlers collecting pages to train a model	GPTBot, ClaudeBot
Search	Crawlers indexing content for AI search and AI answers	OAI-SearchBot, PerplexityBot
User fetcher	A page fetched because a person asked an AI assistant to read it	ChatGPT-User, Perplexity-User
Agent	AI agent traffic	—
AI Other	Traffic attributed to a recognised AI provider without a more specific category — see the AI Other note below	—

Usermaven's Exclude AI bot traffic from stats card showing the enabled toggle, the helper text about the AI Traffic dashboard and billable usage, and five checkboxes labelled Training, Search, User fetcher, Agent and AI Other. — The Exclude AI bot traffic card, with the switch on and the five activity categories: Training, Search, User fetcher, Agent and AI Other.

How the switch and checkboxes interact

With the switch on, only the checked categories are excluded. The switch alone excludes nothing.
If you uncheck all five, no AI bot traffic is excluded.
With the switch off, AI bot traffic stays in your reports.

Things worth knowing

The AI Traffic dashboard is unaffected. It keeps showing crawler activity whichever categories you exclude here. This setting applies to your other reports, so you can strip AI traffic from your headline numbers while still monitoring what AI crawlers read.
Recognised bot traffic never counts toward billable usage, whether or not you exclude it.
Defaults: workspaces that existed before the two bot switches were split keep the bot preference they already had. New workspaces start with the switch on and all five categories selected.
Some traffic shown as AI Other is controlled by Exclude other bot traffic, not the AI Other checkbox. Bot identification improved after these visits were recorded, so older traffic can now display a bot name and provider that were not stored at the time. Its original category was deliberately left unchanged so past visitor counts stay the same. The checkboxes here read the category stored when an event was recorded, so that older traffic stays under Exclude other bot traffic. Traffic classified as AI Other from now on is controlled by the AI Other checkbox.
CCBot and similar generic crawlers are filtered by Exclude other bot traffic, not by the AI Training checkbox.

Exclude other bot traffic

Exclude other bot traffic from stats removes every other detected bot from your reports, including historical bot traffic recorded before Usermaven classified AI bots.

This is the renamed original switch. Older instructions and older versions of this article call it "Exclude bot traffic from all stats" — it is the same setting under a clearer name, and the AI categories now sit in their own card above it.
Leave it on unless you specifically want non-AI bots counted. It is the switch that keeps your historical numbers consistent, because it also covers traffic recorded before AI bots were classified.

Exclude proxy traffic

Turn on Exclude proxy traffic from all stats to filter traffic that arrives through intermediary servers rather than directly from a visitor's device. With the switch on, you choose which proxy types to exclude:

Proxy type	What it is	Why it affects your analytics
VPN	Services that encrypt a connection and route it through servers elsewhere	Masks a visitor's true location, so geography can be wrong. Also removes legitimate privacy-conscious users and remote employees.
TOR	A volunteer-run anonymity network	Origin is effectively untraceable, so these visits rarely represent typical behaviour.
Datacenter	IP ranges belonging to cloud providers and data centers	Often automated processes, bots, or large-scale scraping. Some legitimate B2B visitors arrive through corporate networks on datacenter IPs.
Public Proxies	Openly available proxy servers anyone can use	Frequently used for non-genuine traffic; skews geography and inflates unique visitor counts.
Web Proxies	Browser-based services that hide a visitor's IP	Similar to public proxies; often used to bypass geo-restrictions or workplace filters.
Search Engine Robots	Automated crawlers from search engines indexing your site	Essential for SEO, but inflates traffic and engagement metrics. Keep it visible if you monitor crawl rate.

Usermaven's Exclude proxy traffic from all stats card, showing the enabled toggle above six checkboxes: VPN, TOR, Datacenter, Public Proxies, Web Proxies and Search Engine Robots. — The Exclude proxy traffic card, with the switch on and the six proxy types you can select.

Things worth knowing

Proxy exclusion is independent of the bot switches. It filters the types you select whether or not bot exclusion is on.
This behaviour changed. Proxy exclusion used to apply only when the general bot exclusion was also on. It now works on its own, so if you had proxy exclusion enabled and bot exclusion disabled, your reports may show lower traffic than before. That is the fix working, not a loss of data.
Bot status can mask proxy type. Traffic already flagged as a bot has its proxy type hidden at ingestion, so filtering cannot recover that detail. The existing server-side event exception also still applies.
When in doubt, start with the most obvious non-human sources — datacenter and public proxies — and review your numbers before adding more.

Exclude server side events

Exclude server side events from website analytics keeps events your server sends through the Usermaven API out of your website analytics.

Turn it on when server-side events duplicate what the browser already reports, or when back-end activity would otherwise be counted as website traffic. Leave it off if your server-side events are the only record of conversions you care about — excluding them removes them from website reports.

Exclude internal traffic and users

Exclude internal traffic and users filters out traffic from internal sources — team members, test accounts, and development environments. The card also links to the Usermaven - Traffic Exclusion Chrome Extension, which excludes your own visits in one click.

Using the Chrome extension

Install Usermaven Traffic Exclusion from the Chrome Web Store.
Activate the extension.
Visit your website and click the extension icon to enable exclusion for your browser.

Adding rules

Turn the switch on, then click Add rule. Each rule combines three parts:

Type — IP Address, Hostname, or Email
Operator — Equals or Contains
Values — one or more values; you can type new ones directly into the list

Rules are saved individually and can be edited or deleted afterwards. Until the switch is on, the rules area stays disabled.

Example rules from the page

Rule	Effect
`Email contains @yourdomain.com`	Excludes all events from your company's team members
`Hostname contains localhost`	Excludes all events from local development environments
Cookieless tracking mode active	Change the last octet of the IP address to `.1` — for example `95.10.177.38` becomes `95.10.177.1` — to exclude visits when cookieless mode strips the identifier you would normally match on

Which exclusions do I need?

Goal	What to turn on
Remove AI crawlers and assistants, keep search engine bots	Exclude AI bot traffic, with the categories you want checked
Remove search engine bots and other crawlers, keep AI bot traffic	Exclude other bot traffic
Remove all automated traffic	Both bot switches
Keep automation out of reports but still monitor what AI crawlers read	Exclude AI bot traffic — the AI Traffic dashboard still shows crawler activity
Remove trips through VPNs and data centers	Exclude proxy traffic, with the types you want
Remove your own team's visits	Exclude internal traffic and users, plus the Chrome extension for your own browser
Stop server-side events inflating site traffic	Exclude server side events

Why your counts can change

Every exclusion on this page lowers reported traffic compared with leaving it off, and a step change in pageviews or visitors right after you enable one is expected.

AI data specifically moves in both directions: AI requests previously counted as ordinary visits drop out of your reports, while crawlers that were not recognised before — Apple's, for example — now appear on the AI Traffic dashboard. The proxy fix described above can also lower traffic for workspaces that already had proxy exclusion switched on.

If you compare Usermaven against another analytics tool, note which exclusions each one has enabled before treating a gap as a problem — see Why your Usermaven numbers differ from GA4.

Was this article helpful?