Conductor's 2026 benchmark of 13,770 enterprise domains put AI referrals at 1.08%, ranging 0.25% (Communication Services) to 2.80% (Information Technology), and attribution leaks across apps and surfaces; there is no defensible multiplier. Evidence confidence: hypothesis, not yet confirmed.

Rule aeo-8ddecd48102e0de5 · measurement · hypothesis, not yet confirmed

Treat measurable AI referral traffic as a floor, not a ceiling, and never apply a universal multiplier to estimate the missing share.

Applies to: all

Hypothesis: the reasoning is published, but no independent evidence confirms it yet. Test it on your own pages before you rely on it.

Why this matters

Conductor's 2026 benchmark of 13,770 enterprise domains put AI referrals at 1.08%, ranging 0.25% (Communication Services) to 2.80% (Information Technology), and attribution leaks across apps and surfaces; there is no defensible multiplier.

Implementation

Type: validator

Anti-pattern

Quoting the 1% blended average for planning in a specific industry with an 11-fold sector spread.

Get the full rule + automated checks

Subscribe to access evidence, remediation steps, and configurable check descriptors for this rule.

Check a page free · Subscribe

← Back to measurement rules

Updated