Method
A method to turn scattered signals into actionable decisions
AI presence is not managed with intuition, fast promises, or one-off actions. It requires evidence, diagnosis, prioritization, coordinated execution, and continuous review.
The SUAM method is designed to separate signal from noise, understand what's actually happening, and turn it into decisions a company can execute, measure, and defend internally.
Why this order
AI presence is built in phases, not in impulses.
Order matters. We don't start implementing changes without knowing what problem we're solving.
1. Audit / Blueprint: understand before acting
Before touching the website, creating content, or activating external authority, we need to know how the brand is being interpreted: where it appears, where it doesn't, who it competes with, which sources influence it, and which signals are missing.
Without that base, any action can be a bet.
2. Build: build on evidence
Once the diagnosis identifies the gaps, the priority improvements are activated. Some can be executed internally; others can be coordinated with current providers; others require specialized execution.
Implementation doesn't start from intuition — it starts from a map of priorities.
3. Always-On: sustain the position and respond to change
AI presence changes with models, sources, competitors, sentiment, and market behavior. That's why a one-off project isn't enough. After building, we measure evolution, analyze impact, detect risks, and prioritize new responses.
Decision system
Evidence → Diagnosis → Priority → Action → Validation
The SUAM method works as a decision system. Each phase answers a different question.
Analytical rigor
We don't draw conclusions from a single isolated answer
A one-off response from ChatGPT, Perplexity, or Google AI is not a strategy. It can be a signal, but it isn't enough to make decisions.
That's why we analyze patterns, not anecdotes.
The work doesn't consist of asking once "which brands do you recommend" and drawing fast conclusions. It consists of building sets of high-intent prompts, comparing models, reviewing competitors, observing sources, detecting repetitions, and documenting evidence.
AI presence works like a system of gears: prompts, sources, claims, content, authority, technical structure, reputation, and conversion. If one gear is missing or misaligned, the system loses strength.
Our work is to detect which gears are missing, which are out of place, and which need to be reinforced so the brand becomes more visible, understandable, reliable, and chosen.
Control, influence, and limits
Not everything is controlled. But many signals can be worked
A key part of the method is separating what the company can control, what it can influence, and what no one serious can guarantee.
Controllable
Web, content, claims, structured data, source pages, messages, information architecture, own assets, tracking, and official materials.
Influenceable
Media, rankings, reviews, communities, partners, mentions, comparators, external authority, and public reputation.
Not controllable
The exact response of a model, algorithm changes, third-party publications, total elimination of hallucinations, or a guaranteed recommendation.
The opportunity is in working sources, consistency, authority, and structure well. A Yext analysis of 6.8 million AI citations concluded that 86% came from sources within the brands' sphere of influence — websites, listings, and reviews. That doesn't mean controlling AI. It means there are real signals a company can act on.
Five layers of legitimate authority
Authority isn't forced. It's built in layers.
The SUAM method analyzes which layers exist, which are weak, which influence AI responses, and which should be reinforced without resorting to artificial signals or manipulative tactics.
Prioritization
It's not the flashiest action that wins. It's the one that makes the most sense
We don't work from a rigid menu of tasks. We activate the necessary capabilities based on each company's evidence, priority, and maturity.
Each action is evaluated by:
- potential impact;
- urgency;
- feasibility;
- cost;
- risk;
- internal dependency;
- reputational or regulatory sensitivity;
- execution capacity;
- expected effect on visibility, trust, conversion, or risk control.
The goal isn't to do more things. It's to decide which actions make the most sense now, which should wait, and which don't justify investment.
Impact measurement
How we connect AI presence with business signals
AI can influence a decision before the customer visits the website, completes a form, or talks to sales. That's why we don't measure direct clicks alone.
The method separates different evidence levels to understand what part of the impact is traceable, what part is assisted, and what part should be interpreted as an indirect signal.
1. AI referral direct
Sessions coming from AI engines or assistants with an identifiable referrer.
Examples: - visits from ChatGPT; - Perplexity; - Copilot; - Gemini; - other AI assistants or search engines when the referrer is traceable.
What it indicates: direct traffic attributable to an AI surface.
2. Assisted traffic
Branded searches, direct visits, or subsequent journeys that may appear after a previous AI interaction, even when there's no direct referrer.
What it indicates: possible previous AI influence in the discovery or comparison process.
This should be treated as an assisted signal, not absolute attribution.
3. Branded lift
Evolution of branded searches, mentions, engagement, direct traffic, or demand signals after improvements in presence, content, authority, or reputation.
What it indicates: growth in brand interest or recall, potentially influenced by better presence in discovery channels including AI.
4. Influenced leads
Opportunities where the user declares having discovered, compared, or validated the brand through ChatGPT, Perplexity, Google AI, or another AI system.
This can be captured via: - forms; - "how did you hear about us?" field; - sales notes; - CRM; - commercial interviews; - questions on calls.
What it indicates: declared influence in real opportunities.
5. Attributed or influenced revenue
Closed revenue where the commercial journey includes at least one identifiable AI signal: referrer, lead source, declared response, sales note, assisted search, or touchpoint registered in CRM.
What it indicates: business impact connected to AI when the company has sufficient data to build attribution or multi-touch.
Not every project allows the five levels to be measured from day one. Measurement quality depends on available data, analytics setup, CRM, forms, sales cycle, and collaboration with the commercial team.
That's why the method doesn't promise perfect attribution. It builds a progressive, defensible read of AI's impact on visibility, demand, conversion, and revenue.
Operational coordination
We adapt to the existing structure without breaking it
Many companies already have marketing, SEO, content, communications, PR, development teams, or external agencies.
The SUAM method doesn't seek to replace that structure. It adds a layer of specialized AI criterion to organize an area that usually sits between several teams without a clear owner.
Direction, CMO, or Growth
Defines business goals, priorities, and investment decisions.
Marketing, SEO, content, and communications
Execute or coordinate actions on messages, content, web, reputation, campaigns, and digital assets.
Legal, compliance, or technical team
Validates claims, regulated sectors, privacy, risks, integrations, or technical changes.
SUAM
Diagnoses, prioritizes, defines criterion, validates quality, and executes the parts that require specialized expertise in AI Presence.
We don't come in as another burden for the marketing department. We come in to organize a new layer: how AI interprets, compares, and considers the brand.
Sustainability
We build what holds up
There are methods that promise fast results: guaranteed positions, artificial authority, mass-produced content, link networks, forced mentions, or signal manipulation.
They may look attractive short-term, but they increase the risk of inconsistency, loss of trust, dependence on fragile tactics, and reputational damage.
SUAM does not play that game.
We work with real authority, verifiable sources, citable content, defensible claims, technical structure, and signals that can hold when models change.
Google maintains policies against practices like scaled content to manipulate rankings, link spam, and site reputation abuse, which reinforces the importance of building legitimate, sustainable signals instead of relying on artificial tactics.
This isn't ideology. It's protection of mid- and long-term investment.
Work cadence
A clear process from first contact to continuous improvement
Each project may vary by company, market, and scope — but the working logic follows an ordered cadence.
-
Step 01
Initial conversation
We understand the company, category, goals, markets, competitors, current team, and main concerns.
-
Step 02
Scope and proposal
We define the project perimeter: brand or unit analyzed, markets, languages, models, competitors, prompt families, sources, technical depth, metrics, deliverables, approximate timelines, and project conditions.
-
Step 03
Kickoff and context
We gather key information: assets, restrictions, approved claims, competitors, necessary accesses, and internal owners.
-
Step 04
Analysis and diagnosis
We work the Audit / Blueprint with evidence, prompts, sources, competitors, risks, opportunities, and priorities.
-
Step 05
Executive delivery
We present the diagnosis, evidence, conclusions, action plan, and roadmap. The meeting should include profiles with decision-making or influence on marketing, growth, leadership, technology, or communications.
-
Step 06
Implementation decision
The client decides what to execute internally, what to coordinate with current providers, and what to activate with SUAM as the Build phase.
-
Step 07
Build
If implementation is activated, we agree on scope, owners, calendar, deliverables, and quality criteria.
-
Step 08
Always-On
Always-On operates as a continuous layer of monitoring, analysis, and response. It lets us measure evolution, detect risks, interpret changes, prioritize countermeasures, and keep AI presence aligned with the market.
The team receives monthly reporting, relevant alerts, actionable priorities, and a quarterly strategic review. SUAM handles continuous intelligence, analysis, risk detection, and countermeasure recommendation. Additional executions are activated according to agreed scope.
What we need from the client
The better we understand the context, the better we can prioritize
To work with rigor we need to understand the company, its market, and its constraints. We don't always need everything, but these are the elements that typically help:
- business goals;
- priority markets and languages;
- main competitors;
- key products or services;
- customer profiles or buyer personas;
- brand assets;
- approved claims;
- commercial materials;
- current website and content;
- relevant external sources;
- analytics data if applicable;
- Search Console or similar tools if applicable;
- legal, regulatory, or communication restrictions;
- internal owners or involved providers.
Method quality depends on context quality. That's why kickoff isn't a formality — it's part of the diagnosis.
Frequently asked questions
About the method
How do you avoid analysis depending on a single response?
We work with sets of prompts, models, competitors, sources, and decision scenarios. A single response can be a signal, but decisions are based on documented patterns.
What part of AI presence can really be controlled?
You can control own assets like web, content, claims, structure, data, and messages. You can also influence external sources like reviews, rankings, mentions, media, or communities. What can't be guaranteed is the exact response of a model.
How do you prioritize actions?
We evaluate impact, urgency, feasibility, cost, risk, internal dependency, and execution capacity. Priority doesn't depend on what's most visible — it depends on what makes the most strategic sense.
What if our internal team can't execute?
The roadmap can separate internal actions, actions coordinated with current providers, actions led by SUAM, and actions executed directly by SUAM when they require specialized expertise.
How do you work with existing agencies?
We can work on top of the current structure. If an agency can execute part of the roadmap, SUAM can define criterion, priorities, and validation. If an action requires AI Presence specialization, it can be executed from SUAM.
How do you handle regulated sectors?
With more caution on claims, sources, language, evidence, and risks. SUAM can detect risks and flag points to be reviewed with legal, compliance, or the responsible team — but it does not substitute specialized legal advice.
How do you measure whether an action has worked?
We compare against the initial baseline: presence, mention quality, AI Share of Voice, cited sources, errors, sentiment, position against competitors, AI-sourced traffic, and conversions when data allows.
How often is evolution reviewed?
It depends on scope. In Always-On projects, monthly tracking is typical, with alerts on relevant risks and a quarterly strategic review.
What's the difference between AI Presence Management and SEO?
SEO optimizes a brand's presence in search engines and web results. AI Presence Management works a broader layer: how AI systems discover, interpret, compare, cite, and recommend a brand inside generated answers. SEO remains an important base, but it no longer covers the full discovery and comparison phase mediated by AI. SUAM works with the current SEO team, not against it.
Start by understanding which signals are influencing how AI interprets your brand.
The first decision isn't what to change. It's understanding what's happening, which signals are causing it, and which priorities make the most sense.