Skip to content

Audience analytics

Ordinary traffic counters count everyone: real people are mixed with bots, scrapers, and fake activity. WebShield sits in front of all traffic and sees who passed the browser check and who didn’t — so it can show an honest picture of your audience: how many were live people and how much was automation.

Where to look: the Visitors tab of domain statistics.

“Unique visitors” are unique verified browsers: clients that passed the browser check and behaved like a real browser. This is the most accurate measure of how many live people you had.

Note: verification is counted on hosts where bot protection (browser check) is enabled. Without it this metric stays at zero — then rely on the other cards on the tab. See “Bot and click-fraud protection” for how to enable it.

Next to it is “Unique IPs” — a rough estimate by address. It is always higher than the real number of people: whole networks sit behind one IP (NAT), and bots forge their signals. That is exactly why this metric diverges from external counters like Google Analytics, and why “verified visitors” is the number to trust.

The Automated traffic card shows what percentage of all requests came from recognised bots (search engines, AI crawlers, scrapers, monitoring, scanners). A high share signals that a large part of your load is generated by non-humans.

The AI crawler requests card isolates traffic from language-model and AI-search bots (GPTBot, ClaudeBot, PerplexityBot and others). This shows how actively AI systems crawl your site. You can control their access in “Known bots access”.

The People and bots chart breaks traffic into categories (people, search, AI, social, SEO tools, monitoring, scanners) and shows their dynamics. A spike in a bot category over a flat “human” layer usually means a scraper pass or an attack, not a surge in popularity.

  • Measure traffic by verified visitors, not by IPs or requests.
  • Watch the automated traffic share: growth while the number of people stays flat means resources are being spent on bots.
  • If AI crawlers make up a noticeable share, decide whether you want visibility in AI search, and restrict them in “Known bots access” if needed.

Recognising people relies on passing the browser check and on client behaviour. No method is 100% accurate: some advanced bots pass the check, and some real people with unusual browsers fail it. WebShield gives a more honest estimate than “by IP” counters, but does not claim absolute precision.