Brand Rating: Measuring a Company’s Position Inside the AI Evidence Graph

Answer Capsule

Brand Rating is a proposed framework for measuring how strongly a brand is positioned within the network of sources, claims, and evidence that influence AI-generated answers. Unlike raw brand mentions, Brand Rating weights where a brand appears, how often it appears in influential sources, whether those mentions are positive or negative, how close they are to recommendations, and how consistently the brand is represented across multiple AI systems and over time.

In simple terms:

Brand visibility measures how often a company is mentioned. Brand Rating measures whether the company is present where machine opinion is actually formed.

A brand can be mentioned everywhere and still lose inside AI search.

That sounds counterintuitive until you stop counting mentions and start examining where those mentions occur.

Consider two competitors.

Brand A

• 1,500 web mentions
• hundreds of backlinks
• broad social visibility
• strong traditional SEO presence

Brand B

• 600 web mentions
• fewer referring domains
• less overall media volume

At first glance, Brand A appears stronger.

But now inspect the information sources repeatedly used when AI systems answer high-intent questions in the category.

Suppose Brand B appears positively in:

• eight of the ten most influential comparison sources,
• three independent testing publications,
• the leading industry association,
• several high-intent buyer guides,
• and the sources most frequently cited near final AI recommendations.

Brand A appears frequently across the web, but only weakly inside those sources.

Which company has the stronger position inside the machine evidence environment?

Possibly Brand B.

That is the problem I call Brand Rating.

What Is Brand Rating?

Brand Rating measures how strongly a brand is represented within the influential information network surrounding AI-generated conclusions.

It asks:

Is the brand present inside the sources that matter most when machines evaluate the category?

A brand with high Brand Rating may:

• appear across highly central sources,
• receive repeated positive comparison language,
• be supported by independent evidence,
• show up near recommendation-oriented claims,
• be represented across multiple AI systems,
• maintain that position over time,
• and occupy a strong relative position against competitors.

Brand Rating is therefore not simply another visibility score.

It is a network-position measurement.

Why Mention Counts Are Not Enough

Most brand-monitoring systems begin with volume.

How many times was the company mentioned?

How many articles discussed it?

How many backlinks does it have?

How many social posts referenced it?

How much share of voice did it earn?

Those are useful measurements.

But they assume that mentions are roughly comparable.

They are not.

A positive mention in a page repeatedly used by AI systems for purchase recommendations may matter more than twenty generic mentions on low-influence pages.

Likewise, one negative statement inside a highly central source can potentially matter more than dozens of favorable but peripheral references.

Brand Rating attempts to account for where the mention sits inside the evidence network.

Visibility Is Not Position

This distinction is similar to the difference between being present in a city and owning property in the center of it.

A brand can appear across hundreds of peripheral sources.

Another brand can occupy a smaller number of highly strategic information nodes.

Both are visible.

Only one may be central.

This matters because AI systems do not synthesize the entire internet evenly.

They retrieve, rank, filter, and use subsets of information.

The commercial question becomes:

Which brand is best represented inside the subset that repeatedly contributes to machine conclusions?

That is Brand Rating.

Brand Rating Depends on Citation Rating

Brand Rating is the brand-side counterpart to Citation Rating.

Citation Rating asks:

Which sources disproportionately shape AI answers?

Brand Rating asks:

Which brands are most strongly represented inside those sources?

The relationship is straightforward.

If Source A has low influence and Source B has high influence, a brand mention in Source B should probably not be treated as equivalent to a mention in Source A.

This is why a useful Brand Rating model should weight brand presence by the centrality of the source containing that presence.

Read: Citation Rating— Why the Sources That Shape AI Opinion Matter More Than the Sources That Simply Rank

A Simple Example

Imagine a category with ten highly influential sources.

Citation Rating analysis produces this rough hierarchy:

1. Source A
2. Source B
3. Source C
4. Source D
5. Source E
6. Source F
7. Source G
8. Source H
9. Source I
10. Source J

Now compare two brands.

Brand Alpha

Appears on:

• Source F
• Source G
• Source H
• Source I
• Source J

Five central-source appearances.

Brand Beta

Appears on:

• Source A
• Source B
• Source C

Only three appearances.

A simple mention count favors Brand Alpha.

But Brand Beta occupies the three most influential nodes.

If those sources repeatedly support high-intent recommendations, Brand Beta may possess greater Brand Rating.

That distinction is the point.

Brand Rating Should Be Weighted, Not Counted

A basic Brand Rating framework could consider several dimensions.

1. Source Centrality

How influential is the source mentioning the brand?

A mention inside a highly central source should generally carry more analytical weight than a mention inside a peripheral source.

2. Mention Frequency

How often does the brand appear across the monitored evidence network?

Frequency still matters.

It simply should not be treated as the whole story.

3. Recommendation Proximity

How close is the brand mention to:

• a ranking,
• a recommendation,
• a buying decision,
• a product comparison,
• or a category-defining claim?

A mention in background context is different from:

“Brand X is the best option for seniors living alone.”

4. Sentiment or Direction

Is the brand being represented:

• positively,
• negatively,
• neutrally,
• comparatively,
• or conditionally?

Centrality without sentiment can be misleading.

A company can be highly central for the wrong reason.

5. Evidence Support

Are favorable claims about the brand supported by:

• independent testing,
• expert evaluation,
• customer evidence,
• regulatory data,
• technical documentation,
• or merely the brand's own marketing material?

6. Cross-Model Presence

Does the brand appear consistently across different AI systems?

Or is its visibility concentrated in one platform?

7. Query Coverage

Does the brand appear across multiple relevant query classes?

For example:

• “best overall”
• “best for seniors living alone”
• “best budget option”
• “best for active users”
• “most reliable”
• “best customer service”

A brand that dominates only one narrow query cluster may have different centrality from one appearing across the category.

8. Persistence Over Time

Does the brand maintain its position across repeated measurement periods?

A one-month spike differs from durable centrality.

Brand Rating Is Category-Specific

Just like Citation Rating, Brand Rating should not be treated as one universal score.

A company can be highly central for one decision and nearly irrelevant for another.

Consider a financial-services brand.

It may have high Brand Rating for:

“best business checking account”

but low Brand Rating for:

“best credit card for international travel.”

Likewise, a software company may dominate:

“best CRM for real estate teams”

while barely appearing for:

“best CRM for large enterprises.”

Brand Rating therefore needs to be measured within a defined market, query set, or decision context.

Brand Rating Is Also Claim-Specific

A brand can be central around one claim and peripheral around another.

For example:

Brand X

Frequently associated with:

“lowest price”

but rarely associated with:

“best reliability.”

Brand Y

Frequently associated with:

“premium service”

but rarely associated with:

“best value.”

These associations matter because AI systems often synthesize category recommendations from recurring claim patterns.

That means Brand Rating may eventually need to be measured at several levels:

Category-level Brand Rating

How strongly is the company positioned across the category overall?

Query-level Brand Rating

How strongly is it positioned for a specific question?

Attribute-level Brand Rating

How strongly is it associated with a particular product or brand attribute?

Recommendation-level Brand Rating

How strongly is it positioned close to final purchase recommendations?

This turns AI visibility into something more diagnostic than a generic share-of-voice percentage.

The Evidence Graph Is the Real Object

In the previous article on Search Authority vs. Machine Authority, I described what I call the evidence graph.

Traditional SEO maps links.

An evidence graph maps relationships among:

• sources,
• claims,
• companies,
• categories,
• AI systems,
• and conclusions.

Read: Authority Is Splitting in Two — Search Authority vs. Machine Authority

Brand Rating measures where a company sits inside that graph.

The question is no longer merely:

“How many sites mention us?”

It becomes:

“How centrally positioned are we inside the evidence network that explains our market?”

Brand Rating and Network Science

The idea of measuring importance based on network position is well established.

Linton Freeman's foundational work on centrality formalized concepts such as degree, closeness, and betweenness centrality.[1]

Phillip Bonacich developed measures in which the importance of a node is affected by the importance of surrounding nodes.[2]

Google's PageRank applied a recursive version of this logic to hyperlinks, recognizing that links from important pages should not automatically be treated the same as links from less important pages.[3]

Brand Rating borrows from this intellectual tradition.

The hypothesis is:

A brand's influence should partly depend on the importance of the sources in which it is represented.

This does not mean Brand Rating is PageRank for brands.

It means the same core insight applies:

network position matters.

Brand Rating Is Not Brand Awareness

A company can have enormous awareness and mediocre Brand Rating.

Consider a household-name brand entering a new product category.

Millions of people know the company.

But influential specialist sources may consistently favor established competitors.

The brand has:

high awareness

but potentially:

low category-specific Brand Rating.

Conversely, a relatively unknown specialist brand may dominate the evidence network for a narrow category.

It can have:

low general awareness

but:

high Brand Rating within that decision environment.

Those are very different assets.

Brand Rating Is Not Share of Voice

Share of voice usually asks:

What percentage of overall category discussion belongs to us?

Brand Rating asks:

What percentage of influential category evidence belongs to us?

That is a major difference.

Suppose:

Brand A

35% total web share of voice.

But only 12% presence across the highest-centrality sources.

Brand B

18% total web share of voice.

But 48% presence across the highest-centrality sources.

Traditional monitoring favors Brand A.

Machine-evidence analysis may favor Brand B.

That gap could help explain why AI systems recommend Brand B more often.

Centrality Can Be Positive or Negative

A high Brand Rating score is not automatically good news.

Imagine a company appears constantly across the most influential sources in its market.

Unfortunately, those sources repeatedly associate the company with:

• poor customer service,
• difficult cancellations,
• product failures,
• regulatory problems,
• or higher prices.

The brand has high centrality.

But the centrality is negative.

A mature framework should therefore separate:

Brand Rating

from:

Brand Direction or Sentiment.

For example:

Brand Rating: 82/100
Direction: Negative
Recommendation Proximity: High

That would signal a major reputational problem.

The company is highly visible exactly where it does not want to be.

The Most Important Question May Be “Why?”

Suppose a company discovers that a competitor has much stronger Brand Rating.

That tells us something is happening.

It does not yet tell us why.

We then inspect the source network.

Perhaps the competitor is central because:

• it has more independent reviews,
• it performs better in testing,
• its product data is clearer,
• it has stronger customer evidence,
• it appears in more category roundups,
• it is endorsed by an industry association,
• or one highly influential review site repeatedly favors it.

Now the competitive problem becomes actionable.

Without Brand Rating, the team knows:

“AI likes our competitor.”

With Brand Rating, it can begin to answer:

“Which parts of the evidence network are causing that preference?”

Brand Rating Can Reveal Narrative Ownership

Another important use is understanding who owns specific category narratives.

For example:

Attribute: Best Value

Brand A appears across 70% of high-centrality sources.

Attribute: Best Reliability

Brand B appears across 61%.

Attribute: Best for Beginners

Brand C dominates.

Attribute: Best Premium Option

Brand D dominates.

Now the market is no longer represented as one generic ranking.

It becomes a machine-perceived positioning map.

That is extremely useful for brand strategy.

Because brands rarely need to win every category.

They need to own the right ones.

Brand Rating Can Show When Positioning Has Failed

Marketing teams frequently describe themselves with words that the market does not use.

A company may say:

“We are the easiest platform for small businesses.”

But if high-centrality third-party sources rarely associate the company with ease of use, the positioning has not propagated.

The brand's own website may contain the message hundreds of times.

That does not mean the market evidence network agrees.

This creates a useful comparison:

Owned Positioning

What the company says about itself.

Central Positioning

What influential external sources repeatedly say about the company.

The gap between those two may be one of the most important diagnostics in LLM optimization.

First-Party Claims Should Be Discounted

This is especially important because machine-generated recommendations often rely on a mixture of first-party and third-party information.

A brand can publish:

“We have the best customer service.”

That is evidence of what the company claims.

It is not independent validation.

If five external publications repeat the same statement but all trace it back to the company's own marketing, that should not necessarily count as five independent pieces of evidence.

Brand Rating should therefore incorporate source independence.

A company whose position is supported by:

• independent testing,
• customer data,
• expert review,
• and institutional sources

should be treated differently from one whose apparent centrality depends mostly on repeated first-party claims.

Brand Rating Can Be Manipulated Poorly

There is an obvious temptation once marketers understand this concept.

Find the most central sources.

Get the brand mentioned there at any cost.

That would be a mistake.

A central source is valuable precisely because it has influence.

Destroying its independence destroys the value of the network.

The sustainable approach is to improve the evidence available to the source.

That can include:

• original research,
• verifiable product data,
• independent testing,
• expert access,
• transparent pricing,
• corrections to outdated information,
• meaningful customer data,
• public documentation,
• and genuinely newsworthy findings.

The goal is not to manufacture Brand Rating.

It is to earn a stronger position within the evidence network.

This Changes the Definition of Digital PR

Traditional digital PR often optimizes for:

• publication reach,
• Domain Rating,
• backlinks,
• brand awareness,
• and referral traffic.

Brand Rating adds another dimension:

Does this placement improve the brand's position inside the sources that influence machine recommendations?

That changes campaign prioritization.

A niche publication may matter more than a giant publisher if it has far higher Citation Ratingfor the exact category being targeted.

This is one reason I believe digital PR is evolving toward what I call Machine Relations.

Read on CiteWorks Studios: Machine Relations — Why PR Is Becoming Citation Engineering

Brand Rating Creates a Competitive Matrix

A useful Brand Rating report could eventually look something like this:

Now the CMO can see something much more useful than:

“Brand A has more backlinks.”

The company can see where competitors are structurally stronger inside the machine evidence environment.

The Brand Rating Gap

Just as Citation Rating can reveal a mismatch between conventional PR targets and machine-influential sources, Brand Rating can reveal a Brand Rating Gap.

The Brand Rating Gap is the difference between:

how strong a company believes its category position is

and:

how strongly it is actually represented across influential external sources.

A brand may lead:

• paid media,
• organic traffic,
• social reach,
• or traditional PR volume

while still trailing competitors inside the AI evidence network.

That gap is strategically important.

Because it can remain invisible until consumers begin asking machines instead of search engines.

Brand Rating Should Be Longitudinal

A single snapshot is useful.

A trajectory is better.

Suppose:

January

Brand A Rating: 41
Brand B Rating: 67

April

Brand A: 52
Brand B: 64

August

Brand A: 71
Brand B: 61

That movement tells a story.

Now we can investigate:

• Which sources changed?
• Which new reviews appeared?
• Which claims gained support?
• Did independent evidence increase?
• Did AI recommendation share move with Brand Rating?
• Was one major publication responsible?
• Did the change precede or follow conventional search growth?

This is where longitudinal AI citation intelligence becomes more useful than one-time monitoring.

Brand Rating Can Become Predictive

The most interesting possibility is whether Brand Rating can eventually help predict AI recommendation movement.

Suppose a brand begins gaining positive representation across highly central sources.

AI recommendations have not yet changed.

Three months later, its recommendation share rises.

If that pattern repeats across many categories, Brand Rating may become a useful leading indicator.

That is a hypothesis.

It needs testing.

But it is exactly the kind of question LLM Authority Index should investigate.

Brand Rating and the Consensus Index

Our Consensus Index measures what multiple AI systems currently conclude.

Read: The Consensus Index — What Happens When We Stop Asking One Reviewer Who Is Best?

Brand Rating helps explain those conclusions.

The relationship is:

Consensus Index → Who do the machines recommend?

Citation Rating→ Which sources influence those recommendations?

Brand Rating → Which brands occupy the strongest positions inside those sources?

This creates an analytical chain from output back to evidence.

Brand Rating and Machine Authority

In the article Authority Is Splitting in Two, I argued that conventional search authority and machine authority should be measured separately.

Read: Search Authority vs. Machine Authority

Brand Rating is one component of the brand-side Machine Authority picture.

A company with high Brand Rating may be more deeply embedded in the information environment from which machine recommendations are formed.

That does not guarantee an AI system will recommend it.

Other factors matter.

But it gives us a much more useful measurement than raw mentions alone.

Brand Rating and the Algorithmic Reciprocity Loop

Brand Rating also interacts with the Algorithmic Reciprocity Loop.

Read: The Algorithmic Reciprocity Loop

If machines repeatedly recommend a brand, companies, journalists, customers, and creators may reference those recommendations.

Those human references create new public evidence.

That evidence can then further strengthen the brand's presence across the information network.

The loop becomes:

high Brand Rating

→ greater machine recommendation likelihood

→ greater human attention

→ more external mentions and citations

→ stronger evidence-network presence

→ potentially greater future Brand Rating

Again, this should be treated as a testable hypothesis rather than a declared ranking mechanism.

Brand Rating Can Also Fall

The same loop can work in reverse.

A brand loses favorable coverage.

A major source changes its recommendation.

A new competitor begins receiving independent validation.

Customers increasingly complain about one issue.

Several highly central sources update their reviews.

AI systems begin reflecting the changed evidence environment.

The incumbent's Brand Rating declines.

That can happen even if:

• its backlink count remains enormous,
• its domain authority remains strong,
• and its traditional organic traffic remains relatively stable.

This is another reason traditional metrics alone may not reveal an emerging AI reputation problem.

A Working Brand Rating Model

There is currently no standardized formula for Brand Rating.

Conceptually, a future model might include:

Brand Rating = source centrality × mention strength × recommendation proximity × evidence independence × cross-model presence × query coverage × persistence × directional sentiment

Additional adjustments could account for:

• claim-level relevance,
• source duplication,
• negative centrality,
• evidence quality,
• competitor relative strength,
• and category-specific weighting.

The formula should not be fixed by intuition alone.

The correct weights should be determined empirically by comparing calculated Brand Rating against observed AI recommendation behavior.

Brand Rating Is Not a Secret AI Ranking Factor

This distinction matters.

Brand Rating is not:

• a Google ranking factor,
• an OpenAI metric,
• an Anthropic score,
• a Perplexity score,
• or a documented internal variable used by any commercial AI system.

It is an observer-side framework.

We measure:

• AI outputs,
• cited sources,
• brand mentions,
• source relationships,
• recommendation patterns,
• and changes over time.

Then we attempt to model which brand positions best explain the observed results.

That is the same fundamental logic used in many forms of market intelligence.

We cannot see every internal mechanism.

We can still measure the outputs.

Why This Matters to CMOs

A CMO does not ultimately care about a number called Brand Rating.

The CMO cares about questions like:

Why does AI recommend our competitor?

Which external sources matter most?

Where are we missing?

Which claims do machines associate with us?

Which claims do they associate with competitors?

Are those claims accurate?

Which influential publications need updated information?

What evidence are we failing to provide?

Is our reputation improving or deteriorating?

Which PR activities actually change machine perception?

Brand Rating is useful if it helps answer those questions.

If it becomes another vanity score, it has failed.

From Measurement to Machine Relations

This is where the commercial application becomes clear.

Once a company knows:

• what the machines say,
• which sources shape those conclusions,
• and where the brand sits inside those sources,

it can begin making rational communications decisions.

That is the foundation of Machine Relations.

Instead of spraying press releases across hundreds of websites, a company can identify the information environments that genuinely matter.

Instead of manufacturing mentions, it can improve the evidence available inside those environments.

Instead of monitoring AI answers as mysterious outputs, it can trace them back toward the sources and claims responsible.

That is a much more mature version of LLM optimization.

The Emerging Measurement Stack

At LLM Authority Index, I believe the emerging measurement stack looks like this:

Consensus Index

What do the machines conclude?

Citation Rating

Which sources disproportionately influence those conclusions?

Brand Rating

Which brands occupy the strongest positions within those influential sources?

Machine Authority

How much influence does a source or entity exert over generated answers?

Algorithmic Reciprocity

How does machine recognition create new human recognition and web authority?

Machine Relations

How should companies act on that intelligence?

Each layer answers a different question.

Together, they begin to map what conventional SEO tools cannot fully show.

The Most Important Shift

Traditional marketing measurement often focuses on volume.

More links.

More mentions.

More traffic.

More articles.

More impressions.

AI-mediated discovery introduces a different possibility:

position may matter more than volume.

A brand does not necessarily need to appear everywhere.

It needs to be credibly represented in the right places.

That is the principle behind Brand Rating.

The objective is not to flood the web.

It is to understand the evidence network surrounding the market and determine whether the brand occupies an appropriate, accurate, and defensible position within it.

The Question Brands Should Start Asking

The old question was:

“How much authority does our website have?”

The next question became:

“Which sources have the greatest Citation Rating?”

The brand-level question is:

“How strongly are we represented inside those sources compared with our competitors?”

That is Brand Rating.

And if AI systems increasingly mediate product discovery, vendor evaluation, reputation, and purchasing decisions, that question may become much more important than most companies currently realize.

Sources and Research

1. Freeman, Linton C. — “Centrality in Social Networks: Conceptual Clarification.” Social Networks, 1978/1979.
https://doi.org/10.1016/0378-8733(78)90021-7

A foundational work on network centrality. Brand Rating builds upon the general principle that the structural position of a node can provide more information than simple frequency or degree alone.

2. Bonacich, Phillip — “Power and Centrality: A Family of Measures.” American Journal of Sociology, 1987.
https://doi.org/10.1086/228631

Bonacich's work demonstrates how node importance can depend on the importance and structure of surrounding nodes. This is directly relevant to the hypothesis that a brand mention inside a highly influential source should be weighted differently from a peripheral mention.

3. Page, Lawrence; Brin, Sergey; Motwani, Rajeev; Winograd, Terry — “The PageRank Citation Ranking: Bringing Order to the Web.” Stanford InfoLab, 1999.
https://ilpubs.stanford.edu:8090/422/

PageRank provides a historically important example of weighting relationships according to network importance rather than simply counting them.

4. Kleinberg, Jon M. — “Authoritative Sources in a Hyperlinked Environment.” Journal of the ACM, 1999.
https://doi.org/10.1145/324133.324140

Kleinberg's HITS framework distinguishes between authorities and hubs and provides useful conceptual precedent for examining different forms of structural importance within information networks.

5. Lewis, Patrick et al. — “Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks.” NeurIPS, 2020.
https://arxiv.org/abs/2005.11401

Foundational RAG research provides technical context for why external information retrieval creates a source-selection environment that can be analyzed separately from conventional web ranking.

6. Aggarwal, Pranjal et al. — “GEO: Generative Engine Optimization.” arXiv:2311.09735; presented at KDD 2024.
https://arxiv.org/abs/2311.09735

This research treats source visibility within generative-engine answers as a measurable outcome. It does not validate Brand Rating specifically, but it supports the broader premise that generative visibility can be systematically studied.

7. Liu, Nelson F. et al. — “Evaluating Verifiability in Generative Search Engines.” 2023.
https://arxiv.org/abs/2304.09848

Research evaluating whether generative-search citations adequately support generated claims. It reinforces the importance of evidence quality and provenance rather than treating all mentions or citations as equivalent.

Methodology and Disclosure

Brand Rating is a working analytical framework proposed by Mark Huntley and LLM Authority Index.

It is intended to measure how strongly a company or entity is represented within a defined AI evidence network.

Brand Rating is not a documented ranking metric used by Google, OpenAI, Anthropic, Perplexity, Microsoft, or another AI provider.

The framework builds upon established concepts from:

• network centrality,
• citation analysis,
• information retrieval,
• PageRank,
• entity analysis,
• and generative-search research.

Candidate Brand Rating variables include:

• Citation Rating of surrounding sources,
• brand mention frequency,
• recommendation proximity,
• directional sentiment,
• evidence independence,
• source diversity,
• cross-model presence,
• query coverage,
• category specificity,
• and persistence over time.

The relative importance of these variables remains an empirical question.

LLM Authority Index intends to test whether Brand Rating correlates with, explains, or predicts changes in AI recommendation behavior across commercial categories.