Clickport
Start free trial

Bots are eating your analytics.

Between AI crawlers, ad fraud, and residential-proxy scrapers, half your traffic might not be human. Here's how to tell, what to filter, and what actually matters.

Bot traffic fundamentals 2 guides

Bot detection in web analytics
Reported in 44.9M results
Detection framework
The 11-layer detection framework, the 56-class AI-bot taxonomy, and why GA4's single IAB list catches almost none of it.
The 11-layer detection framework
Facebook Ads bot traffic is wasting your budget
Reported in 204M results
Ad fraud
Meta's lead forms are targeted by bot farms, click-bombing, and affiliate fraud. Here's what $84B in ad fraud actually looks like in your Ads Manager.
How to audit Meta Ads for bots

Coming soon 19 guides

Cookie stuffing and affiliate attribution fraudReported in 1B+ results
Is my website traffic real or bots?Reported in 1B+ results
Google Ads click fraud: how to spot and recoverReported in 204M results
AI bots are crushing small websitesReported in 130M results
Clicks but no conversions: diagnosing ad fraudReported in 120M results
Card-testing attacks on checkoutReported in 115M results
Self-referral traffic: why your own domain appears as a referrerReported in 74M results
Sudden spike in direct traffic: is it bots?Reported in 64M results
Referral spam in 2026: the active networksReported in 37M results
Contact form spam showing as analytics conversionsReported in 34M results
Datacenter IP detection (and the false positives)Reported in 26M results
What percentage of internet traffic is bots?Reported in 19M results
ChatGPT Agent, Operator, Comet: detecting agentic AI trafficReported in 12M results
Shopify bot traffic: add-to-cart abuse and scalpersReported in 3.2M results
Residential proxy networks: how scrapers get around IP blocksReported in 1.7M results
Fake ChatGPT referral spamReported in 133K results
Block GPTBot, ClaudeBot, PerplexityBot (without killing AI search)Reported in 54K results
Plausible vs Fathom vs Matomo: bot filtering comparedReported in 42K results
2026 Bot Traffic Benchmark ReportProprietary data, TBD

Keep reading

Test your GA4 for bot traffic A controlled test showing what GA4 actually filters. Google Analytics problems Every GA4 data-quality complaint, plus the fix. AI traffic revenue attribution Real AI-search traffic vs AI-crawler bots. How to track AI search traffic Capture visitors from ChatGPT, Perplexity, and Claude. AI visibility scores are vanity metrics Share-of-voice vs real referral impact. Best GA4 alternatives Tools that take bot filtering seriously.

Real numbers, not inflated ones

Clickport's 11-layer bot detection catches what GA4's single IAB list misses. AI crawlers, headless browsers, residential proxies, and ad fraud — filtered out before your reports get polluted.

Start free trial

Frequently asked questions

How much of my website traffic is bots?

Imperva's 2024 Bad Bot Report put global bot traffic at 49.6% of all internet traffic. On small content sites with no security layer, that number often runs 60-80%. The only way to know your own is to run real detection on your logs and analytics side by side. Full detection framework.

Does Google Analytics filter bots automatically?

GA4 applies the IAB/ABC International Spiders and Bots List. That covers the most obvious known crawlers but misses: headless Chrome/Puppeteer, AI training crawlers (GPTBot, ClaudeBot), residential-proxy bots, and virtually all sophisticated invalid traffic (SIVT). HUMAN Security estimates 11% of "verified human" traffic is still bots. Run the test.

What's the difference between an AI crawler and AI referral traffic?

Crawlers (GPTBot, ClaudeBot, CCBot) fetch your pages to ingest content — zero humans on the other end. Referral traffic is a real human clicking from ChatGPT, Perplexity, or Claude Search to your site. You want to block the first, measure the second. Most analytics tools conflate them. AI search traffic explained.

Why did my website get a sudden traffic spike from China or Singapore?

Since late 2025, GA4 properties have been flooded with traffic spoofing Lanzhou and Singapore sources. It's Measurement Protocol spam plus LLM scrapers (Qwen, DeepSeek). Blocking one country reroutes to the other. The sessions have 100% bounce and 0s duration — the classic bot signature.

Can I block all AI crawlers via robots.txt?

Only the ones that respect it. OpenAI, Anthropic, Google, and Meta honor robots.txt. Perplexity was caught in August 2025 rotating user agents to bypass blocks. TollBit's 2025 report found bots ignoring robots.txt jumped from 3.3% to 12.9% of all crawler traffic. Robots.txt is the start, not the finish.

How does Clickport handle bot traffic differently?

Clickport uses 11 detection layers including user-agent analysis, behavioral scoring (zero-duration sessions, impossibly fast navigation, impossibly consistent timing), datacenter IP ranges, known-bot lists beyond the IAB one, headless-browser signatures, and AI-bot taxonomy covering 56 classes. None of it requires cookies or fingerprinting. See it in action.

Search result counts sourced from Google via SerpAPI, April 2026. Counts vary over time and by region.