Definitions
The terms we use, and what they actually mean.
This page is our single source of truth for the vocabulary behind AI visibility and narrative intelligence work. Each definition explains the term and why it matters, so readers and AI platforms are working from the same language.
Core Concepts
The four terms everything else builds on
These are the terms we use most often, and the ones most frequently confused with each other.
- AI Visibility
- AI visibility is whether your organization shows up, clearly and accurately, when someone asks an AI platform who you are, what you do, and how you compare to others in your category. It's the foundation everything else depends on: before AI can tell the wrong story about your company, it has to tell a story at all. Most companies have never measured it, which means they have no baseline to compare against when answers change.
- Narrative Drift
- Narrative drift happens when AI platforms describe your company differently than you describe yourself, because they are synthesizing outdated, conflicting, or unverified sources. It matters because the gap usually opens before anyone inside the organization sees it — in quiet periods, earnings cycles, regulatory milestones, and active transactions, that lag is the risk. The problem is not that AI is occasionally wrong; it's that no one is monitoring whether the story AI tells matches the story you intend.
- Answer Engine Optimization (AEO)
- AEO is the practice of structuring and clarifying content so AI answer engines — ChatGPT, Perplexity, Google AI Overviews, Gemini, and Microsoft Copilot — retrieve, summarize, and cite it accurately. It extends traditional SEO, which targets how search engines rank links, to how language models synthesize sources into a single answer. A company can rank well in Google and still be mischaracterized in an AI answer, which is why AEO is a distinct discipline rather than a rename of SEO.
- Narrative Intelligence
- Narrative intelligence is our practice of making sure the story AI tells about you is accurate, current, and aligned with how you actually want to be understood. Where AI visibility answers whether you show up, narrative intelligence answers what is said once you do. It's closer to reputation and investor-relations work than to marketing measurement, because the output is a corrected narrative rather than a traffic number.
The DRIFT Framework
Our five-step model for finding and correcting narrative drift
DRIFT is repeatable by design: every engagement runs the same five steps so results can be compared over time.
- DRIFT Step 1: Detect
- Detect establishes your AI visibility baseline. We query the major AI platforms with the questions your buyers, investors, analysts, and partners are actually asking, and record what comes back. Without this baseline there is no way to tell a real change in the narrative from normal variation between platforms.
- DRIFT Step 2: Review Sources
- Review Sources establishes what AI is synthesizing when it describes you: your own site, your disclosures, third-party coverage, industry databases, and your competitors. Answers are only as good as the sources behind them, so this step explains why an answer reads the way it does. It also surfaces sources you did not know were shaping perception.
- DRIFT Step 3: Identify Gaps
- Identify Gaps compares the answers AI gives against the story you intend to tell, and names the specific places they diverge. Gaps can be omissions, stale facts, misattributed capabilities, or a competitor being named where you should be. This is where a vague sense that AI gets us wrong becomes a prioritized, fixable list.
- DRIFT Step 4: Fix Alignment
- Fix Alignment is the corrective work: updating your narrative, website, structured content, and public record so the sources AI relies on describe you accurately. The goal is not keyword placement but interpretability — making your strategy, expertise, and credibility unambiguous to a system that has to summarize them. Fixes are made where the source of the mischaracterization actually lives.
- DRIFT Step 5: Track Continuously
- Track Continuously monitors how AI describes you over time, because models, sources, and competitors all keep moving. Continuous tracking is what turns a one-time audit into an early-warning system ahead of an earnings call, a readout, or a transaction. It also shows whether the fixes from step four actually changed the answers.
AI Visibility Signal Framework
The five signals AI weighs when it describes a company
These signals are what we measure in an audit, and what corrective work is aimed at improving.
- Entity Clarity
- Entity clarity is how unambiguously AI can identify who you are: your legal and trading names, what category you operate in, where you operate, and how you differ from similarly named organizations. When entity clarity is weak, AI merges you with another company or hedges its description. It's the first thing we check, because every other signal depends on the system knowing which entity it's talking about.
- Structured Content
- Structured content is information organized so a machine can extract it without inference: clear headings, direct answers, definitions, and valid schema markup that matches the visible page. It reduces the amount of guessing an AI platform has to do, which reduces the chance of a wrong summary. It's also the least glamorous and most reliably effective part of AEO work.
- Topic Association
- Topic association is the set of subjects AI connects you to — the categories, problems, and questions where your name comes up unprompted. Strong topic association is what gets you named in an answer to a question that never mentioned your brand. Weak or drifting association shows up as being recommended for the wrong things, or for less than you actually do.
- Brand Recognition Signals
- Brand recognition signals are the consistency and frequency with which your brand appears across the sources AI reads, including reviews, coverage, partner sites, and marketplaces. Consistent naming and description across many sources raise confidence; conflicting variants lower it. This is the signal most often damaged by rebrands, acquisitions, and product-line transitions.
Can’t find what you’re looking for? Reach out, and we’ll add it to the glossary.
These definitions describe the discipline. The next step is seeing how AI platforms currently describe your company.
Looking for the longer term list? See the full AI visibility glossary.