Defining brands for the age of answers.
The internet was built around search. The future is built around answers.
Consumers increasingly ask AI what to choose, what to trust and what best fits their needs. In those moments, brands are no longer competing only for attention or visibility.
They are competing to be understood, trusted and recommended.
The SpokenOf Research Foundation defines the principles behind that new reality.
A new decision layer is emerging.
Search helped people find information. It presented links, products and possible destinations, leaving the evaluation largely to the user.
AI takes a more active role.
It interprets the question, considers the context, compares alternatives and forms an answer. It may explain why one product is more suitable than another, introduce considerations the user had not raised and narrow the field before a website is ever visited.
AI is therefore becoming more than a discovery channel.
It is becoming a decision layer.
This is the Recommendation Economy.
Brands are no longer competing to appear.
They are competing to be selected.
Recommendation is not retrieval.
An AI system does not recommend a brand simply because it can find it.
It evaluates the brand in relation to a specific person, need, moment and competitive set. A product may be highly relevant in one context and unsuitable in another. The same brand can therefore be understood correctly, represented accurately and still not belong in every answer.
That distinction matters.
The objective is not to appear everywhere. It is to be selected where the brand genuinely belongs.
Recommendation is a judgement made in context.
A brand now exists in two forms.
One is created and managed by the company.
The other is reconstructed by AI from the information available across the wider ecosystem.
The intended brand
The intended brand is the identity created and managed by the company. It includes the positioning, products, audiences, benefits, differences, values and voice the organisation wants people to understand.
The interpreted brand
The interpreted brand is the identity reconstructed by AI. It is formed from websites, product information, retailers, reviews, publishers, experts, communities, comparisons and every other accessible signal surrounding the brand.
These two identities are rarely identical.
Commercial influence increasingly depends on the distance between them.
When the gap is small, AI can represent the brand accurately. When the gap is large, AI may simplify it, misunderstand it or prefer a competitor whose evidence is clearer.
That gap is where SpokenOf begins.
AI does not recommend claims.
It recommends confidence.
Confidence cannot be created by one campaign, one product page or one perfectly written statement.
AI forms an understanding from the relationship between many signals. It looks for patterns, consistency and corroboration. It compares what the brand says about itself with what the wider ecosystem appears to confirm.
When those signals support the same conclusion, confidence grows.
When they are incomplete, contradictory or weakly validated, AI must fill the gaps. The resulting representation may be inaccurate, generic or shaped more strongly by external sources than by the brand itself.
Visibility may place a brand inside the answer.
Evidence determines whether it can be recommended.
How information becomes perception.
The SpokenOf Brand Intelligence Framework describes how information surrounding a brand becomes machine perception and how that perception influences recommendation.
Each stage affects the next.
A weak or undefined Brand Truth creates inconsistent signals. Inconsistent signals limit understanding. Limited understanding weakens representation. Weak representation reduces confidence and makes recommendation less likely.
The framework allows that process to be examined, measured and improved.
Define what must be understood.
Brand Truth is the clearest, most defensible expression of the brand.
It establishes who the brand is, what it offers, who it serves, which needs it solves and why it is meaningfully different. It also defines what the brand can prove and the situations in which it should - or should not - be recommended.
Brand Truth is not a slogan or a campaign idea. It is the strategic foundation that connects the brand’s positioning, products, evidence and communication to the same underlying reality.
Without a defined truth, consistency cannot be measured because there is no clear standard against which interpretation can be assessed.
Everything teaches AI.
AI does not learn a brand from one source.
It reconstructs the brand from the entire information environment surrounding it.
Owned content explains what the company wants to communicate. Retailers add category and commercial context. Reviews introduce experience. Experts, publishers and communities contribute validation and interpretation. Comparisons position the brand against alternatives.
Even missing information becomes a signal.
When the answer cannot be found in one place, AI must infer it from somewhere else.
Individually, these signals may appear insignificant. Together, they shape what AI believes to be true.
Visibility means AI found you.
Understanding means AI knows you.
A brand can appear frequently in AI-generated answers and still remain poorly understood.
AI may recognise the brand name but associate it with the wrong audience. It may identify the product category while missing the actual benefit. It may describe what the company sells without understanding when or why someone should choose it.
True understanding begins when AI can explain what the brand does, who it serves, which needs it addresses and how it differs from credible alternatives.
It must also recognise the contexts in which the brand belongs.
Without that clarity, visibility creates presence but not preference.
Every answer is a brand interaction.
Representation is the version of the brand communicated inside an AI response.
It is visible in the facts AI selects, the benefits it emphasises, the language it uses and the comparisons it makes. It determines which audiences are associated with the brand, which strengths are remembered and which limitations are introduced.
Accurate representation protects more than factual correctness.
It protects meaning.
A brand can be visible, correctly named and still be represented in a way that weakens its positioning. When AI reduces a differentiated brand to a generic category description, something commercially valuable has been lost.
Representation shows whether the intended brand survives inside the answer.
Authority is externally validated confidence.
Brands cannot create authority through claims alone.
Authority develops when relevant and credible sources repeatedly support the same understanding. Reviews, experts, publishers, retailers, research and independent comparisons help AI determine whether a brand’s claims are isolated assertions or recognised truths.
This does not mean every source carries equal weight.
Authority depends on relevance, credibility, consistency and context.
A single high-profile mention may create awareness. A coherent body of supporting evidence creates confidence.
Authority is therefore not simply popularity.
It is the strength of the ecosystem confirming what the brand represents.
Recommendation is the moment of selection.
AI recommendations are contextual.
The system considers the person asking, the need being expressed, the constraints surrounding the decision and the alternatives available. It then determines which brand appears most appropriate based on the evidence it can access and the understanding it has formed.
This means recommendation cannot be measured through one generic prompt.
It must be examined across different audiences, use cases, stages of the journey, product needs, markets and AI systems.
The objective is not universal recommendation.
It is justified recommendation.
A strong brand should be selected consistently in the conversations where its value is genuinely relevant.
Perception is the complete AI view of the brand.
Perception is not one answer.
It is the pattern that emerges across models, questions, products, markets and decision contexts.
It reflects what AI can find, what it understands, how it describes the brand, which evidence it trusts and when it chooses the brand over an alternative.
That pattern reveals the brand AI currently believes to be true.
This may align closely with the intended brand. It may also expose misunderstandings, omissions and competitive disadvantages that traditional brand tracking cannot see.
Perception is where every previous layer becomes observable.
Measure what determines whether AI recommends your brand.
Traditional marketing measures exposure, reach and response.
The Perception Score™ measures whether AI is equipped to understand, represent and recommend the brand with confidence.
Governance
Governance measures how clearly the organisation defines and manages the truth AI should understand. It examines whether positioning, product logic, audiences, claims and evidence are sufficiently aligned to create a consistent foundation.
Readiness
Readiness measures whether the brand’s information is complete, structured and interpretable across the products, markets and use cases that matter. It reveals where missing explanations or inaccessible information limit machine understanding.
Ecosystem
Ecosystem measures how strongly external sources support the brand’s relevance, credibility and claims. It examines whether the wider information environment confirms the intended brand or creates a different perception.
Influence
Influence measures how consistently AI systems surface, describe and recommend the brand inside real decision-making conversations. It shows what the accumulated signals currently produce in the answer.
Together, these dimensions reveal more than current visibility.
They show where the brand stands, what is holding it back and how far it can realistically move.
The method creates the value.
A number without explanation is not intelligence.
Every assessment must be traceable to the question asked, the model tested, the response received, the competitors considered and the evidence supporting the conclusion.
The Perception Score is therefore not designed to create another superficial marketing dashboard.
It is designed to support accountable decisions.
The score shows the current position. The underlying evidence explains why that position exists. The methodology identifies what should change next.
Understand how AI reaches its conclusions.
SpokenOf evaluates brands across structured sets of real-world questions and decision scenarios.
We compare what the brand needs AI to understand with what leading AI systems currently say, omit and recommend. We examine whether that representation remains consistent across models and whether the same evidence leads to similar conclusions.
The analysis also identifies which competitors are selected instead, which sources appear to shape the answer and where incomplete or contradictory signals reduce confidence.
The result is not a generic visibility report.
It is a map of how AI currently interprets the brand and what is required to improve that interpretation.
AI perception is not a one-time optimisation project.
Models change.
Sources change.
Products change.
Markets and competitors change.
The questions consumers ask change as well.
Brands therefore need an ongoing way to define, measure and govern how they are understood across the answer ecosystem.
Diagnose
Establish the current perception and reveal the gap between the intended brand and the interpreted brand.
Improve
Strengthen the owned and external signals that shape machine understanding, representation and recommendation.
Measure
Track whether those changes produce a stronger, more accurate and more competitive perception over time.
This is AI Brand Governance.
The systems will change.
The need for clarity will not.
The SpokenOf Research Foundation is not tied to one model, platform or moment in the development of AI.
Its purpose is to establish a durable way of understanding how brands are interpreted and selected across intelligent systems.
The framework will continue to evolve through diagnostics, controlled testing, category benchmarks, industry research and real-world improvement programmes.
As new systems and behaviours emerge, the methods will develop.
The central principle will remain:
AI can only represent and recommend what it can understand with confidence.
The future will not be won by the brands that publish the most.
It will be won by the brands that are understood.
Whose truth is clear.
Whose evidence is credible.
Whose ecosystem supports the same conclusion.
Whose value is relevant to the person asking the question.
Brands that AI can represent accurately.
Brands that AI can recommend confidently.
Brands that deserve to be SpokenOf.
Discover how AI understands your brand.
Measure your current perception, identify the gaps and see what is preventing AI from recommending your brand with confidence.