Methodology, legal, customer guides, and engineering reference - all published as the source of truth for how Rezolts works. Versioned with the code, intended to be argued with.
How we measure AI visibility. The maths, the sampling design, the things we can and cannot detect. Read these before trusting any number we put on a dashboard.
Overview
The four signals, the Visibility Score formula, per-engine playbooks
Visibility Score
Exact formula, weights, justification, interpretation guide
Prompt panels
Sampling design, account hygiene, why we don't scrape
Bot tracking
The 17 AI bot patterns we match, what we log, what we don't
AI referral attribution
Why GA4 misses ~90% of AI traffic and how we capture it
Citation extraction
How we parse and classify cited sources from each engine
Cohort benchmarks
The n≥10 suppression rule, opt-out, percentile computation
Limitations
What we explicitly do not claim to measure, plus changelog
Public commitments on what we will and will not do, plus how we handle data. Designed to be reviewable by procurement, security and DPO teams.
Setup walkthroughs and how-to documentation for everyone using the platform.
For developers integrating with or contributing to the platform. Architecture, conventions, runbook, known issues.
Architecture overview
System map, data flow, tech stack rationale
Engine adapter authoring
How to add a new AI engine in under a day
Coding conventions
TypeScript, file naming, API patterns, do/don't
Runbook
Deploy, incident response, on-call procedures
Known issues
Phase 1 → Phase 2 gaps, environmental gotchas, engine quirks
Disagree with us? If you read the methodology and think we're measuring the wrong thing, write to hello@rezolts.com. The methodology is versioned and revised in public.