
13,828 Pages Crawled, 5 Visitors Sent Back
AI Traffic is live in WebDecoy. See crawls and observed AI referrals for your property, then freeze the figures into a shareable report.
bot-detectionSee which AI crawlers fetch your site and where observed AI referrals land. A real WebDecoy dashboard walkthrough, with setup steps and sensor coverage.
Published · Updated
See which AI crawlers fetch your site, then inspect the visits AI platforms send back. WebDecoy puts those observations in AI Traffic, alongside the sensor coverage behind them.
On our own webdecoy.com property, the seven-day dashboard captured on November 10, 2026 UTC showed:
| Dashboard measure | Captured value |
|---|---|
| Pages crawled by AI-only crawlers | 3,789 |
| Additional crawling shared between search and AI | 629 |
| Observed AI referral visits | 6 |
| Distinct referral landing pages | 5 |
These are one property’s dashboard observations, not an industry benchmark or a result caused by a marketing campaign. Crawling volume does not measure visitors acquired. The capture includes the current UTC day, which was incomplete.
Install WebDecoy on your site to inspect your own bot traffic, AI crawling and observed referrals.
Silent walkthrough built from the actual app captures below, with English captions. The seven-day snapshot is dated; it is not a live counter.
Transcript: AI crawlers found your site. See what came back. Track crawling and observed referral visits separately. Our own seven-day dashboard showed 3,789 AI-only crawls and six observed referral visits, with 629 additional crawls shared between search and AI. Five landing pages received observed AI referrals. Open AI Traffic → Referrals to inspect destinations. A page’s platform label shows its leading source; its visit count combines sources. Install WebDecoy, confirm sensor coverage, and use server or edge collection for non-JavaScript crawlers.

Actual own-site app capture. The collection panel includes a warning about sensors reporting without proving their type; these figures do not establish verified edge coverage for every source.
Start with the selected property and time window. Then read three different questions:
The landing-page row’s platform label names its leading platform. The visit count combines platforms for that page, so summing row counts by their displayed labels will not reconstruct the platform totals.
A crawl does not prove a citation. A referral does not reveal the user’s prompt, prove the visitor is human, or identify the exact answer they saw. The app’s crawl-to-referral ratio is an observed comparison, not a conversion rate.
| What you want to observe | Collection to consider |
|---|---|
| Browser visits and supported referral signals | Browser tag |
| Requests handled by your application, including non-JavaScript crawlers | Supported server integration or WordPress plugin |
| Requests handled at the edge, including cache-served requests | Supported edge integration configured for that traffic path |
A browser tag cannot observe a crawler that never executes it. An origin plugin cannot observe a cached page request that never reaches WordPress. Check WordPress and Cloudflare cache coverage before interpreting an empty origin log.
Use the collection panel to confirm which sources reported and when. A source name alone is not proof of an authenticated sensor or complete traffic coverage.
Keep your existing Cloudflare protection and acquisition analytics. WebDecoy adds bot and crawler evidence; it complements those tools.
Google Analytics can identify traffic when source information is available. It does not need to classify every AI arrival as Direct. For example, OpenAI documents that ChatGPT search referral URLs include utm_source=chatgpt.com.
Before comparing dashboards, align the property, date range, timezone and counting unit. A WebDecoy observed visit, a GA4 session and a page view are different measurements. Missing referrers, redirects, consent choices, collection coverage and campaign parameters can also change what a tool records. See Google’s Traffic acquisition documentation.
For the next content decision, look at the actual destination. A technical tool receiving an observed visit gives you a specific page to inspect for a useful installation path. A crawler fetching that tool gives you a separate observation about automated access. Neither alone establishes sales or signup impact.
WebDecoy’s value starts with seeing what reaches your site: bots, crawlers and the referral traffic you can identify. Install the right collection, confirm it reports, and inspect your own activity.
Choose your WebDecoy integration →
Updated November 2026 with an actual dashboard walkthrough. This replaces the original launch copy’s blanket claims about other analytics tools and guaranteed human referral traffic.
Yes. AI Traffic shows observed crawling and recognized AI referrals for a selected property and time window. Collection depends on the installed sensors. Server or edge coverage can observe non-JavaScript crawlers; the page tag observes browser activity.
No. A crawler request, a citation and a referred visit are separate events. Crawling alone does not establish that your page appeared in an answer or that someone clicked it.
No. Use WebDecoy's bot and crawler evidence alongside your acquisition analytics and existing Cloudflare protection. Compare the same property, period and attribution definitions before comparing counts.

AI Traffic is live in WebDecoy. See crawls and observed AI referrals for your property, then freeze the figures into a shareable report.
bot-detection
Keep Cloudflare. Add visibility into the bots reaching your site: decoy interactions, request evidence, AI crawlers and the referral visits they send back.
bot-detectionNine crawlers no bot list named yet, from an undocumented Alibaba bot to an anonymous crawler on 158 addresses in one day. What each sent and how to handle it.
bot-detectionLike this post? Share it with your friends!
Get a personalized demo from our team.