/ Glossary /
The words of AI Brand Governance
Every term we use on this site, explained in one sentence, with the everyday word it corresponds to and the page where it is used.
Request the AI Light AuditSix AI platforms · One method · Verifiable numbers
/ Concepts /
Concepts
The words that describe the problem. GEO is explained in full in the article what Generative Engine Optimization is; GEM on the Generative Experience Marketing page.
AI Brand Governance
The work of measuring and steering how generative AI systems describe, position, and recommend a brand.
Keeping what AI says about a company under control
AI visibility
How much and how a brand appears in AI answers: whether it is cited, in what position, in what tone.
Being present in AI answers
GEO, Generative Engine Optimization
The set of interventions on sources, structure, and clarity of information so that AI models understand and cite a brand.
SEO for AI answers
GEM, Generative Experience Marketing
The discipline that coordinates a brand’s organic presence in AI answers (GEO) with its paid presence in the same spaces, such as ChatGPT Ads and Google’s ads in AI Overviews and AI Mode.
Making the AI answer and the ad next to it say the same thing
AEO, Answer Engine Optimization
Preparing content so that it answers questions directly, in the format that answer engines and AI pick up.
Writing answers, not just pages
AI Overview
The AI-generated summary Google shows at the top of search results, with the sources it was built from.
Google’s answer before the links
LLM, Large Language Model
The language model behind ChatGPT, Gemini, Claude, and the other assistants: it generates text from what it read in training and what it retrieves at query time.
The engine of the AI assistant
AI Perception Gap™
The distance between how a company wants to be perceived and how AI systems actually describe it.
What AI gets wrong about you
/ Method and tools /
Method and tools
How we work. The four phases are described on the ThinkAI BrandOS™ page; the starting point is the AI Brand Audit.
ThinkAI BrandOS™
ThinkAI’s method in four repeating phases: Monitor, Diagnose, Orient, Correct.
Our way of working, in cycles
AI Brand Audit
The initial measurement: 80 questions across 6 AI platforms, recorded and traced back to their sources, with a competitor comparison and an action plan.
The starting snapshot
Light Audit
The free, short version of the audit: 16 questions on ChatGPT, Gemini, and Perplexity, results presented in a 20-minute call with a written summary.
A taste of the audit
IntentIQ
ThinkAI’s platform that asks several AIs the same questions, records the answers and sources, and tracks how they change over time.
The tool that asks the questions and keeps score
/ Documents /
The documents the client receives
Each phase produces a document. They are delivered with the audit and at every governance cycle; IntentIQ feeds them with the recorded answers.
Brand Health AI Dashboard™
The dashboard that gathers the indicators of how AI describes the brand, updated every cycle.
The control panel
AI Perception Gap Report™
The document that quantifies the distance between the intended positioning and the one perceived by AI, with the risk areas.
The report on what AI gets wrong
Source Intelligence Map™
The list of sources AI uses to talk about a brand, with the weight of each.
Where AI got what it says
RIM Matrix™
The matrix that ranks interventions with the formula Impact² / Effort: first those that change the brand’s representation the most with the least work.
The priority list
Content Architecture Plan™
The plan of content and structured data to build or fix so that AI finds clear, complete information.
What to write and where
Governance Cycle Report™
The end-of-cycle report: interventions made, AI answers before and after, next priorities.
The minutes of each cycle
GEM Readiness Report
How the brand appears in both layers of an AI answer, organic and paid, question by question, with sources and competitors’ ads.
The snapshot of answer and ads together
Paid Layer Plan
The plan for the paid layer: where to buy, on which questions, and with which message, with priorities and expected results.
The ads plan
/ Indicators /
Indicators
The numbers we use to measure results, cycle after cycle. They are also described in the metrics section of the method.
AI Perception Score™
A 0 to 100 index of consistency between how the brand wants to be perceived and how AI describes it.
The overall grade
AI Presence Rate™
Percentage of relevant questions in which the brand is cited correctly in AI answers.
How often you show up
Citation Quality Index™
Accuracy, tone, and competitive framing with which AI describes the brand when it cites it.
How they talk about you
Silent Warning Frequency™
How many times per cycle AI answers contain an implicit warning about the brand, such as “some users report issues.”
AI’s little digs
AI Readability Score™
How well the brand’s content is correctly understood by AI systems.
How well AI understands you
Correction Effectiveness Rate™
Percentage of interventions, organic and paid, that produce a positive change in the next cycle’s answers.
How many interventions work
Answer Consistency Score™
Consistency, from 0 to 100, between the brand’s organic description in the AI answer and the paid ad on the same question.
Do the answer and the ad agree?
Want to see these numbers for your brand?
The free Light Audit calculates them on 10 questions and 2 AI platforms, in five business days.
Request the AI Light Audit