Transparent & independent

How we measure AI visibility

One Signal Score, built from eight independent signals. No pay-to-rank. Here's exactly how it works — the weights, the maths, and the grade tiers.

Citation Score — 22%Trust Score — 18%Entity Strength — 15%Coverage Score — 13%Source Diversity — 10%Consistency — 9%Freshness — 8%Expert Mentions — 5%
100
Signal Score
8 weighted signals
The weighting

Eight signals. One score. Fixed weights.

Being cited by credible sources as a recognized entity is the foundation of durable AI visibility — so those three signals carry 55% of the score. The remaining five refine and stress-test that foundation.

  • Citation Score22%
  • Trust Score18%
  • Entity Strength15%
  • Coverage Score13%
  • Source Diversity10%
  • Consistency9%
  • Freshness8%
  • Expert Mentions5%
The eight signals

What each signal actually measures

Every brand receives eight sub-scores from 0 to 100. Each captures a distinct dimension of AI visibility.

0122%

Citation Score

How frequently the brand is cited across AI answers and authoritative web sources.

22% of the Signal Score
0218%

Trust Score

The authority and reputation of the sources doing the citing.

18% of the Signal Score
0315%

Entity Strength

How clearly AI models recognize the brand as a distinct, well-defined entity.

15% of the Signal Score
0413%

Coverage Score

The breadth of topics, questions, and categories where the brand appears.

13% of the Signal Score
0510%

Source Diversity

The range of independent domains and publishers referencing the brand.

10% of the Signal Score
069%

Consistency

How consistently the brand is described across engines and over time.

9% of the Signal Score
078%

Freshness

The recency of citations, mentions, and updated references.

8% of the Signal Score
085%

Expert Mentions

Presence in expert, editorial, and analyst-grade sources.

5% of the Signal Score
The math

Scored in the open

No black box. Each sub-score is multiplied by its fixed weight, and the eight contributions are added together. That sum is the Signal Score.

The formula

Signal Score =Σ (sub-score × weight)

Worked through for a representative brand. Every number here is produced by the same scoring engine that grades every profile on this site.

Citation Score
92
× 22%
20.2
Trust Score
88
× 18%
15.8
Entity Strength
90
× 15%
13.5
Coverage Score
81
× 13%
10.5
Source Diversity
76
× 10%
7.6
Consistency
85
× 9%
7.6
Freshness
72
× 8%
5.8
Expert Mentions
64
× 5%
3.2
Signal Score — sum of contributions
Example Brand · Strong Signal
84.3/ 100
Grade tiers

Where a score lands on the scale

The Signal Score maps to one of five tiers. The bands are fixed — a brand moves up only by moving its score.

0
60+
70+
80+
90+
0 — no measurable signal100 — maximum signal
Signal Leader
90–100 Signal Score

Dominant, consistent AI visibility across engines.

Strong Signal
80–89 Signal Score

Well-established presence with broad citation coverage.

Established
70–79 Signal Score

Recognized entity with solid, reliable visibility.

Emerging
60–69 Signal Score

Growing footprint; gaining citation momentum.

Developing
Below 60 Signal Score

Early-stage visibility with room to grow.

The pipeline

From an AI answer to a published score

  1. 01

    Observe

    We run category prompts across AI engines and answer engines, and track how brands surface in the authoritative web those engines are grounded on.

  2. 02

    Attribute

    Every answer is parsed for brand mentions and the sources cited alongside them — who was named, who was quoted, and who was linked.

  3. 03

    Weight

    Sources are graded by authority, independence and recency, so a citation from a credible publisher counts for more than an anonymous mention.

  4. 04

    Score

    Each of the eight signals is normalised to 0–100, the fixed weights are applied, and the composite Signal Score and grade tier are published.

Independence

No pay-to-rank. Ever.

Brands can claim, verify and enhance their profile. None of it touches the score.

Payment never moves rank

Ranking position is determined only by the Signal Score. No brand can buy a higher score or a better position.

Sponsored is always labelled

Any paid placement is clearly marked as sponsored and sits outside the ranked order.

Verification is not a boost

Verifying a profile corrects the facts. It never adds points — though accurate facts do help Consistency over time.

In full

The complete methodology

What Top Citation Signal measures

Top Citation Signal (TCS) is an independent research platform. We measure one thing: AI visibility — how visible, trusted, cited, and consistently represented a brand is across AI engines, answer engines, search engines, and authoritative web sources.

We do not measure a product's quality, its revenue, or its customer satisfaction directly. We measure the signal a brand sends into the systems that increasingly decide which brands get recommended: large language models, AI search, and the authoritative web they are trained and grounded on.

Important & independent. The scores on this site are Top Citation Signal's own proprietary methodology. They are not official metrics from Google, OpenAI (ChatGPT), Anthropic (Claude), Perplexity, Gemini, or any other platform, and we never represent them as such.

The eight signals

Every brand receives eight sub-scores from 0 to 100. Each captures a distinct dimension of AI visibility.

1. Citation Score

How frequently the brand is cited across AI-generated answers and authoritative web sources. This is the volume dimension — presence at scale.

2. Trust Score

The authority and reputation of the sources doing the citing. A mention in an established industry publication counts for more than an anonymous forum post. Trust weighs who is talking, not just how many.

3. Entity Strength

How clearly AI models recognize the brand as a distinct, well-defined entity — with a stable identity, category, and set of associated attributes. Strong entities are less likely to be confused, misattributed, or omitted.

4. Coverage Score

The breadth of topics, questions, and categories where the brand appears. A brand that surfaces across many relevant queries has broader coverage than one that appears in a single niche.

5. Freshness

The recency of citations, mentions, and updated references. AI systems favor current information; freshness rewards brands that maintain an up-to-date footprint.

6. Consistency

How consistently the brand is described across engines and over time. Conflicting descriptions — wrong category, outdated pricing, mismatched positioning — weaken visibility. Consistency rewards a coherent story everywhere.

7. Source Diversity

The range of independent domains and publishers referencing the brand. Diversity protects against fragility: a brand cited by fifty independent sources is more resilient than one cited fifty times by the same source.

8. Expert Mentions

Presence in expert, editorial, and analyst-grade sources — the kind of authoritative references that carry disproportionate weight in how AI systems assess credibility.

How the Signal Score is calculated

The eight sub-scores are combined into a single Signal Score (0–100) using a fixed, transparent weighting:

Signal Weight
Citation Score 22%
Trust Score 18%
Entity Strength 15%
Coverage Score 13%
Source Diversity 10%
Consistency 9%
Freshness 8%
Expert Mentions 5%

The weighting reflects our view that being cited (Citation) by credible sources (Trust) as a recognized entity (Entity Strength) is the foundation of durable AI visibility, with the remaining signals refining and stress-testing that foundation.

The Signal Score maps to a grade tier:

  • 90–100 — Signal Leader: dominant, consistent AI visibility across engines.
  • 80–89 — Strong Signal: well-established presence with broad citation coverage.
  • 70–79 — Established: a recognized entity with solid, reliable visibility.
  • 60–69 — Emerging: a growing footprint gaining citation momentum.
  • Below 60 — Developing: early-stage visibility with room to grow.

Independent vs. sponsored

Ranking position is determined only by the Signal Score. Payment never changes a brand's score or rank.

Brands may claim and verify their profile, and may purchase enhanced features — richer media, lead capture, review collection, monitoring, and certification badges. None of these affect the Signal Score or ranking order. Any sponsored placement is always clearly labeled as such.

How verification works

Verification confirms that a real, authorized representative of the brand controls the profile. A verified brand can correct factual details (links, pricing, positioning), which can indirectly improve Consistency over time — because accurate, coherent information is exactly what strong AI visibility rewards.

Certification is a separate, higher bar: a brand that has both verified its profile and reached a qualifying Signal Score may license a TCS Certified or Gold mark. Certification signals earned credibility; it is never a shortcut to a higher rank.

How brands can improve their profile

The fastest, most durable way to raise a Signal Score is to become genuinely more visible and consistently represented across the authoritative web:

  • Claim and verify your profile so the facts are correct everywhere.
  • Earn citations from credible, independent sources (Trust + Source Diversity).
  • Strengthen your entity with clear, consistent positioning and structured information (Entity Strength + Consistency).
  • Keep your footprint current with fresh, updated references (Freshness).
  • Broaden coverage into the full range of questions your category answers (Coverage).

We publish practical guidance on all of the above in our resources.

Want to improve your Signal Score?

Claim your profile, get your full AI Visibility Report, and see the fastest, most durable ways to climb.