Reading the VeritasScore (company score) | Veritas Links | GEO, AEO, AI Focus Groups - analyzes
VeritasLinks

Help center

Documentation

Guides for GEO reports, AI Focus Groups, scores, sharing, and the Partner API.

Reading the VeritasScore (company score)

VeritasScore is the composite company score that blends a completed GEO analysis with a linked AI Focus Group into one FICO-style number (300–870). It is a management benchmark for AI-facing perception—not a credit rating, SEO rank, or financial score. Live UI examples below use sample data so you can recognize the same controls in your report.

What you need before a score exists

  1. A completed GEO (AI Brand Analysis) run for the brand or URL.
  2. A completed AI Focus Group linked to that GEO (or available in the same brand context).
  3. The scoring job then aggregates both into VeritasScore, breakdown components, narrative, and recommendations.

Related

The headline VeritasScore

The hero gauge shows the overall score on a 300–870 scale (same shape as a FICO-style dial). Higher means stronger modeled AI perception in your category for this run. A zone badge labels the band; a delta badge appears when a previous score exists for the same domain.

VeritasScore

VeritasScore

Overall score

Northline Analytics

612Good

Scale 300–870 (FICO-style) · higher = stronger AI perception

Calculated June 4, 2026 · vs previous run +8

Example UI: VeritasScore overall gauge with zone badge and scale caption.

Score bands

  • Needs Improvement — 300–499: large gaps in visibility, preference, or risk; prioritize GEO + FG fixes before leadership storytelling.
  • Good — 500–699: competitive mid-band; use breakdown to pick the weakest levers.
  • Excellent — 700–870: strong AI-facing position; defend with proof assets and re-runs after major site changes.

Score bands

420

Needs Improvement

300–499

612

Good

500–699

742

Excellent

700–870

Example UI: three VeritasScore bands — Needs Improvement, Good, Excellent.

How the number is built

Base starts at 300. Positive breakdown components add points; Risk subtracts. The result is clamped to 300–870. When multiple LLM model scores exist, Overall is a weighted average across the model set (weights from prompt/response volume). Selecting a model in the UI swaps the gauge, breakdown, narrative, and recommendations to that model’s slice.

Score by LLM model

When per-model scores are available, mini gauges show Overall plus each assistant (ChatGPT, Claude, Gemini, Perplexity, Grok, DeepSeek, etc.). Click a model to inspect how that engine alone scores your brand. Locked models may show as ready but paywalled depending on plan—the pipeline often still ran them.

Score by LLM model

Models

Score by LLM model

Each gauge uses the same 300–870 scale. Overall is a weighted average — select a model to inspect its slice.

Example UI: Score by LLM model — Overall plus per-model mini gauges (clickable).

Score breakdown (components)

Breakdown rows explain which levers moved the headline. Each row shows points contributed (or deducted for Risk) toward the 300–870 total. Lowest positive rows and largest Risk deductions are usually the fastest levers.

  • Brand Visibility (up to ~150) — GEO AI visibility and share-of-voice blend.
  • Market Position (up to ~120) — competitive framing, SWOT strengths/opportunities, rival SOV.
  • Customer Perception (up to ~100) — Focus Group Confidence/Idea minus objection pressure.
  • Content Presence (up to ~80) — GEO content/visibility signals for owned narrative.
  • Growth Potential (up to ~80) — headroom from opportunities and missed-opportunity style signals.
  • Brand Preference (up to ~40) — Focus Group competitor-choice share for your brand.
  • Risk (up to ~40 deducted) — SWOT weaknesses/threats, FG objections, high risks.

Score breakdown

Components

Score breakdown

Formula: 300 + positive components − Risk → clamp 300–870

  • Brand Visibility+98

    of 150 pts max

  • Market Position+72

    of 120 pts max

  • Customer Perception+64

    of 100 pts max

  • Content Presence+48

    of 80 pts max

  • Growth Potential+52

    of 80 pts max

  • Brand Preference+22

    of 40 pts max

  • Risk-44

    of 40 pts max

Example UI: Score breakdown component rows with points and progress bars.

Trend over time

When prior VeritasScore runs exist for the same domain, a history chart plots scores on the 300–870 axis. Direction after real changes (site, proof, GEO re-run, new FG) matters more than single-point noise. Day-7 style banners may highlight a follow-up delta when that pipeline ran.

Trend

Trend

All past VeritasScore runs for this domain (6 analyses).

Example UI: VeritasScore trend line across past analyses.

Score explanation (narrative)

The narrative section is model-authored prose that ties the headline and breakdown into a readable story—why you sit in a band, which rivals dominate, and what proof gaps matter. Treat it as structured interpretation of this run, not a financial filing. On some plans the full narrative may be teaser-locked.

Score explanation

Narrative

Score explanation

Northline Analytics sits in the Good band (612). Brand Visibility is solid across models, but Customer Perception and Brand Preference trail because Focus Group personas still default to SignalPeak without stronger proof. Risk deductions come mainly from pricing opacity and compliance questions. Closing case-study and comparison gaps is the fastest path toward the Excellent band (≥700).

Example UI: Score explanation narrative block.

Recommendations

Action cards list prioritized improvements (high / medium / low) with short impact notes. Use them as a backlog: assign owners, map each card to a GEO experiment or Focus Group re-run, then recalculate VeritasScore to prove movement. Titles alone may appear when recommendations are paywalled.

Recommendations

Actions

Recommendations

Publish mid-market case studies

high

Close the proof gap that keeps Customer Perception and Brand Preference below peers.

Impact: Lifts perception and preference components

Ship SignalPeak alternatives page

high

Own comparison queries where Market Position lags the category leader.

Impact: Improves visibility and market position

Clarify pricing on the homepage

medium

Reduce Risk deduction from pricing opacity objections in Focus Groups.

Impact: Lowers risk penalty

Example UI: Recommendations cards with priority badges.

Recommended services

Suggestions map score gaps to next steps such as blog or FAQ generation. They are optional accelerators—not required to improve the score. Generate buttons appear for blog and FAQ when a deep link is available.

Recommended services

Marketplace

Recommended services

Suggestions from your score — may deep-link to content generators.

AI Blog Generation

high

Targets content gaps that drag Content Presence.

+ Content Presence

FAQ Generator

medium

Answers objections that inflate Risk.

− Risk

Example UI: Recommended services marketplace cards.

How to read breakdown vs headline

  • Headline first for the leadership slide; breakdown for the working session.
  • A “Good” overall with one weak component is a focused project; “Needs Improvement” with high Risk means fix trust/pricing/compliance before chasing vanity visibility.
  • If models disagree widely, open GEO Performance by LLM and Mentions—do not average away a hostile engine your buyers use.
  • Recommendations are ordered suggestions—not guaranteed ROI.

Sharing, PDF, and recalculate

From an owned score report you can download PDF, generate an easy-read summary, share a read-only link, jump to the source GEO or Focus Group, and recalculate after new inputs. Shared views may hide gated sections depending on plan.

Sharing

How to interpret for action

  1. Pick the two weakest breakdown rows (or largest Risk).
  2. Fix them with GEO content/citations and/or a Focus Group–driven proof or messaging change.
  3. Re-run GEO → Focus Group → VeritasScore and compare the trend chart.
  4. Export PDF or combined report for board packs when the trajectory is clear.