Practical guide · Search topic: AI referral source attribution
AI referral source attribution: build an honest channel rule
An AI referral channel is a grouping rule, not a new measurement source. It should classify only source values your analytics actually receives and make it possible to inspect the original value behind each grouped visit.
What the query can—and cannot—measure
Start with a sample of referrer and campaign rows. List included hostnames, exclusions and the date the rule was checked. Keep Google organic separate unless your data distinguishes a specific AI surface. If an assistant link carries a UTM value, retain it alongside the referrer rather than replacing one with the other.
A practical way to do it
Test the rule on historical rows and compare the grouped total with the raw source list. When changing a rule, note the effective date so a classification change is not mistaken for traffic growth.
A useful checklist
- Document included source values and exclusions.
- Preserve raw source/medium fields in the report.
- Annotate rule changes and compare like-for-like dates.
Limits to keep in mind
No rule can recover a referrer that was never sent. Grouping assumptions should be visible, not hidden inside an unexplained metric.
Questions
Should every AI-related domain be included?
Only include sources you have verified as relevant to assistant referrals.
Can I classify Google organic as AI?
Not without a reliable signal that separates the relevant Google surface.
Sources and further reading
Related pages
Sources and channels · quietstats cookie-free analytics AI search referrals · quietstats cookie-free analytics Methodology: how quietstats counts visitors, visits, bounce rate and duration
Put your traffic in context
Explore the product, its measurement method and pricing before choosing a tool.