Visibility Score
A single comparable score per model family, with published weights and full component drill-down.
The AI Visibility Platform
Measure. Understand. Improve. One system for how AI assistants see, cite and recommend your brand.
Overview
From prompt to recommendation
Every AI answer that names a brand is the output of a pipeline: a buyer's prompt is interpreted, evidence is retrieved, entities are selected, and a recommendation is framed with citations. The platform instruments that entire pipeline. It continuously replays the prompts your market actually asks across every major model family, parses each response into entities, recommendations and citations, and turns the results into scores, trends and prioritised actions.
The result is a closed loop: measure how you appear, understand why, ship the fixes, and watch the same panel confirm the movement — quarter after quarter, model update after model update.
Architecture
How a prompt becomes an insight
Measure
01Prompt
Panels of real buyer questions, versioned per category and locale.
02AI Models
Sampled across GPT, Claude, Gemini and Perplexity under consumer defaults.
Understand
03Entity Detection
Brand and product mentions resolved against a curated entity registry.
04Recommendation Analysis
Position-weighted scoring of who gets recommended, and how strongly.
05Citation Analysis
Every cited source classified: official, community, encyclopedic, editorial.
Improve
06Visibility Scoring
The published Visibility Score formula, computed per model family.
07Insights
Trends, gaps, competitor deltas and post-model-update alerts.
08Recommendations
Prioritised GEO actions mapped to the funnel layer where you leak.
Core modules
Prompt Monitoring
Track the buying prompts that matter in your category and how each model answers them over time.
Competitor Tracking
Side-by-side recommendation share, movement and share-of-voice against any competitor set.
Citation Analysis
Where your citations come from, which assets earn them, and where competitors earn theirs.
Entity Analysis
Resolution rate, naming consistency and knowledge-graph coverage for your brand entities.
GEO Recommendations
Prioritised actions with expected impact, mapped to the layer of the Entity Authority Model you're leaking from.
Research Engine
The benchmark corpus and methodology behind every metric — the same one we publish publicly.
Reporting
Board-ready reports and exports aligned to the quarterly State-of-GEO format.
Alerts
Notifications when a model update, competitor move or citation loss shifts your visibility.
Historical Trends
Every metric as a time series, so movement is attributable to actions — not anecdotes.
Models supported
ChatGPT
GPT family
The broadest recommender — wide brand sets with a community-aware tilt.
Claude
Claude family
The most canonical — narrow consideration sets anchored on documentation.
Gemini
Gemini family
Ecosystem-aligned — structured data and entity panels weigh heavily.
Perplexity
Perplexity
Citation-dense and freshness-weighted — a retrieval engine at heart.
What ships next
Future models
New assistant families join the panel as they reach meaningful usage.
Workflow
A quarter with KernelX
Baseline
We build your prompt panel, resolve your entities and establish per-model Visibility Scores — the numbers everything else moves against.
Diagnose
The AI Visibility Funnel locates your deepest leak: recognition, retrieval, recommendation, preference or trust.
Act
Prioritised GEO recommendations — entity fixes, citable assets, comparison content, community signals — with owners and expected impact.
Verify
The same panel re-runs continuously; movement shows up in trends, alerts fire on regressions, and the quarter closes with evidence.
Why it matters
Invisible to AI means invisible, increasingly
A growing share of buying journeys now starts — and often ends — inside an AI answer. Those answers concentrate attention on a shortlist of two or three names per category, and the shortlist is decided by signals most brands have never measured: entity consistency, citation footprints, community consensus, content freshness. The brands instrumenting this layer today are compounding an advantage that will be expensive to chase later.