AI traffic and crawler analytics

AI traffic shows two things: the visits people make to your site from AI answers, and the requests AI crawlers make for your pages. Open AI traffic from the Measure group of the console rail.

AI traffic has five tabs: Overview, Sessions, Pages, Crawlers and Setup. The figures cover the whole workspace until you pick a product.

Two kinds of AI traffic#

An AI referral visit is a session that arrives from an AI product. Praised counts these sessions under one model, direct AI referral.

AI referral visits are identified by their referrer. Stripped referrers, copied links and native apps are invisible to any analytics, which makes the count of AI-referred sessions a lower bound.

An AI crawler visit is a request for one of your pages from an AI bot, such as GPTBot or ClaudeBot. Praised names the crawler from the request’s user agent and keeps only AI-crawler requests.

Crawler requests are grouped by purpose: training crawlers collect pages for model training, search crawlers build an answer engine’s index, and a user fetch loads a page while an AI tool answers a person.

A user agent can be forged. Praised checks whether each request came from the crawler the request names, and counts Verified, Can’t verify and Failed requests apart.

Connect your data#

AI referral visits come from a connected analytics account: Google Analytics, Matomo or Plausible. Praised reads the account with read-only access, and the first sync reads the past 28 days.

Connecting Google Search Console adds search clicks for each page to the Pages tab.

Crawler visits come from a log feed that your site or CDN sends to Praised. Praised holds no account with the provider, and a feed counts from the day the feed starts.

The Setup tab lists each provider with what the provider’s own plan needs and a link to the install steps.

Only an Owner can connect a source. Other members are shown what is missing and who can connect the source.

What each tab shows#

### Overview

Overview shows AI-referred sessions, key events, the value of those key events and crawler fetches for the window. Overview also compares AI-referred sessions with the rest of the site.

### Sessions

Sessions has one row for each AI source: sessions, share of AI sessions, change against the window before, landing pages, key events, event rate, value and value per session.

### Pages

Pages has one row for each page on your site: crawler requests, the last verified crawl, AI sessions, key events, value, search clicks and the number of answers that cite the page.

### Crawlers

Crawlers has a Summary and a Request log. The summary lists each crawler with its purpose, its verified, can’t-verify and failed requests, the paths requested, the verified requests your site refused, and when the crawler was last seen.

### Setup

Setup is a checklist of seven groups: Site and properties, Traffic analytics, Crawler logs, Sitemaps and page identity, Business outcomes, AI source classification, and Privacy and retention.

A figure appears only where its source covers the window. Outside that coverage, the tab says why instead of showing 0.

Limits by plan#

Raw crawler requests are kept for a number of days set by the plan. Older days are kept as daily totals, without page paths.

The address a crawler request came from is used to verify the request and is not stored.

When figures are missing#

An empty request log says which of two things happened: no matching requests from a feed that is delivering, or a feed that is not delivering.

Setup names the cause on the group that owns the cause, for example Not connected, Waiting for first data, No fresh data or Sync failed.