About the project

AI discovery needs a commercial protocol people can inspect.

CrawlClick is built by Skelf Research around a simple tension: answer engines need publisher material, publishers need a sustainable exchange, and advertisers need a way to test discovery without disguising paid claims as editorial fact.

Why build this now?

Crawler traffic and human referral traffic are different evidence classes. Treating both as “AI traffic” hides the economic question. CrawlClick measures them separately, but only network-verified eligible crawler requests can carry a billable insertion; forgeable navigation signals remain analytics.

What principle governs the product?

Claims follow evidence. The code can insert, meter, bill and settle. That does not prove an answer engine will repeat the insertion. The public pilot is designed to create that evidence while live money remains gated.

Why publish a crawler reference?

Publishers cannot make a sensible allow-or-block decision if training crawlers, search indexers and live user agents are collapsed into one category. The crawler directory keeps those purposes, verification methods and vendor sources explicit.

What is the company posture on disclosure?

Paid language is labelled in the visible passage, bound to a reviewed HTTPS destination, and excluded from training-only traffic. The product does not insert model instructions or rely on hidden metadata as its disclosure mechanism.