
Search with Candour Debunking AI Search Myths: Measurement, “Direct” Traffic, and Location Signals (with André Pitì)
10 snips
Nov 2, 2026 André Pitì, a Spanish-born SEO consultant and founder of SEOritmo, joins the conversation to challenge popular AI search assumptions. They explore why visibility metrics are modeled estimates, why direct-traffic spikes do not prove AI referrals, and how geography, URL patterns, logs, and analytics can reveal stronger signals. They also discuss crawler readiness, brand representation, cross-team alignment, and André’s life in music and martial arts.
AI Snips
Chapters
Transcript
Episode notes
AI Visibility Scores Are Modeled Estimates
- AI visibility scores are modeled estimates, not direct measurements of actual traffic or appearances.
- Tools may inherit Google keyword volume, apply platform multipliers, and use stale snapshots or proprietary datasets.
Direct Traffic Spikes Do Not Prove AI Referrals
- Do not treat direct-traffic spikes as proof of AI referrals because dark traffic can combine apps, privacy restrictions, typed URLs, and AI visits.
- Inspect direct sessions by URL and preserve UTM parameters through redirects and canonicalization.
Cross Reference AI Signals Before Claiming Attribution
- Cross-reference direct sessions with server logs, AI crawler activity, URLs, and page clusters before inferring AI-driven visits.
- Treat overlapping signals as evidence for an inference, not as definitive attribution.
