跳到导航
Metrics

Web analytics

以 Markdown 格式查看

Fern automatically collects aggregated web analytics for every published Fern docs site in the Web Analytics tab of the Fern Dashboard.

Select a 7, 14, 30, 90, or 180-day time window. Data isn’t real-time and typically lags by three to four hours.

Traffic and audience

Traffic and audience metrics include:

  • Page views
  • Unique visitors
  • Sessions
  • Top pages (a ranked subset, not every page)
  • Top countries
  • Device types

Acquisition

Acquisition breaks down traffic channels (direct, organic, and referral) and referring domains.

Developer intent

Developer intent metrics include:

  • API Explorer request volume, split into successful and failed requests
  • Views of llms.txt and Markdown files
  • LLM and other AI bot traffic by provider

How bot traffic is classified

The bot traffic chart counts requests for Markdown pages, llms.txt, and llms-full.txt. Each request’s user agent is matched against a list of known providers: Googlebot, OpenAI (GPTBot and ChatGPT), Anthropic (Claude), Bingbot, Perplexity, Cohere, Slackbot, Meta, Amazonbot, Baidu, Yandex, DuckDuckBot, Applebot, etc. Requests that match none of them fall into one of two buckets:

  • Other Bot: the user agent contains bot, crawler, or spider, or the request was flagged as a likely bot when served. A request is flagged as a likely bot when its user agent doesn’t look like a browser (it lacks mozilla, or doesn’t match Chrome, Firefox, Safari, Edge, Opera, or Brave). This bucket includes generic HTTP clients such as curl and Python requests, plus less common AI crawlers.
  • Unknown: the user agent looks like a browser and matches no bot pattern.