Glossary
AI visibility for business aviation: the terms
Short, checkable definitions of the words we use. Where a term has a regulator or standards body behind it, we say what it does and does not cover.
Definitions of measured quantities are the same strings the Index and Methodology pages use.
- AI visibility
How often, and how accurately, a company is named or cited when buyers ask an answer engine the questions that lead to a purchase. It is measured as a share of repeated runs on a fixed prompt panel, per engine and measurement surface, never as a single screenshot or a single rank. Being visible is not the same as being chosen; it is the precondition.
- Answer engine
A system that answers a question in prose, with or without sources, such as ChatGPT, Perplexity, Gemini, Google AI Mode and AI Overviews, Claude and Copilot. The answer is assembled from a retrieval step, a reading step that extracts entities and claims, and a generation step that writes the paragraph; each step has measurable preferences. The same product can differ from its API, which is why every number on this site names the surface it was measured on.
- Answer engine optimization (AEO)
The practice of increasing how often, and how accurately, a company is named or cited in answer-engine responses to its buyers' questions. It overlaps with SEO because engines retrieve from search indexes, but the levers differ: third-party lists, breadth of mention, quotable facts and entity clarity matter more than keyword volume. It does not include automated querying of consumer apps, llms.txt as a lever, or buying placement.
- Generative engine optimization (GEO)
A synonym for answer engine optimization used by some agencies and by the 2024 Princeton study that named it. We use AEO because the buyer's question is answered, not generated, and because the word keeps the focus on being cited in an answer rather than on producing content at volume. The two terms describe the same practice.
- Named / Cited
Named — The company is mentioned by name in the answer text. Cited — A page on the company's domain is listed as a source. A company can be named without being cited (the engine took the fact from a third-party page) and cited without being named (a page on its domain is listed as a source for someone else's claim). We record both, per run.
- Position credit
1.00 first named, 0.85 second, 0.60 third, 0.35 fourth or later, 0 not named — reported alongside, never instead of, share of voice. It rewards being named first without pretending that order is stable; identical ordering across repeated runs is rarer than identical lists, so the credit is published with its interval and after share of voice.
- Accuracy flag
A logged instance of an engine misstating the company's fleet, rating, base or certificate. Flags are logged per run with the wrong statement and the correct one, reviewed by an analyst, and reported alongside share of voice, because being named wrongly (a retired aircraft, a lapsed rating, a base you left) can cost more than not being named.
- Variance (run-to-run)
The spread of "named" across the five repeats of the same prompt; published per prompt. A prompt where a company is named in five of five runs and one where it is named in one of five both count as "named"; variance is what tells them apart. We publish it rather than hide it, and we do not publish before/after results from one run.
- Measurement surface
The specific system a run was made on — an official API with web search for the ranked engines; a hand-run check in a clean browser profile for the calibration sample; a licensed Google AI Overviews check in client work only. We name which, and we never pool them. API answers differ from the consumer apps: different system prompts, model routing, no personalization, no logged-in state.
- Calibration sample
A quarterly, hand-run set of 20 panel prompts × 2 runs entered by an analyst in a clean US browser profile in the ChatGPT app, Google AI Overviews/AI Mode and Microsoft Copilot, logged with screenshots; reported only as the rate at which the app named the same companies as the API engine, never as a ranking. It shows how far the apps diverge from the APIs; it does not close the gap and it is never a ranking.
- Entity facts
Checkable statements about a company: certificate, approvals, ratings with dates, fleet or capabilities, locations, contacts. They are the statements an engine reaches for when a prompt includes "safe", "based at" or "can they"; stated precisely, dated and source-linked, in plain HTML rather than in a badge image or a PDF, they get repeated. Fixing them is the first job in every engagement.
- Placement
An appearance of a company on a third-party page (list, directory, article) that an engine retrieves. Placements are the retrieval layer: "best X" lists, directories, trade-press articles, OEM service-center lists, comparison sites, forum threads. We report every placement attempted, secured and declined, and we never buy one that is presented as editorial.
- Third-party list
A page on someone else's domain that lists companies for a buyer question — a "best charter operators for X" article, a directory category, an OEM's authorized service centers, a comparison table on a review site. Lists were the largest single page type ChatGPT cited in a 750-prompt study, and fresh ones were cited more, so keeping them current is measurable work. 1
- Prompt panel
A fixed, public set of buyer prompts run repeatedly on a schedule so that changes over time mean something. The Index panel is 120 prompts across seven segments and four buying stages, tagged by region, frozen per quarter, with changes proposed monthly, applied only at quarter boundaries and logged with reasons. Clients and subscribers may not propose prompts.
- The Citation Ledger
Our working method for client engagements: prompt panel → logged runs → gaps → fixes → placements → re-runs, kept on one shared dashboard the client can open at any time, with run logs downloadable where the engine's terms allow. It is a method, not a trademark or a product; the methodology page defines each step, and a before/after is only ever the same panel, surfaces and repeats.
- Part 135
The FAA rules for commuter and on-demand operations. In the United States, an air charter flight for hire is flown by a company holding a Part 135 air carrier certificate, which sets crew, training, maintenance, duty-time and operational requirements. Holding the certificate alone says nothing about third-party safety audits; those are the ARGUS, WYVERN and IS-BAO tiers. 2 Verify current status with the issuing body.
- Part 91
The FAA's general operating and flight rules, which govern private and corporate flying that is not for hire. A company flying its own aircraft for its own people operates under Part 91; it may not sell seats or charter the aircraft to the public without a Part 135 certificate holder operating it. Management companies often run an owner's aircraft under Part 91 for the owner and Part 135 for charter. Verify current status with the issuing body.
- Part 145
The FAA rules for repair stations: an MRO certificated under Part 145 holds ratings that define what it may inspect, repair and return to service — the airframe, powerplant, propeller, radio, instrument and accessory classes, plus limited and specialized service ratings — each with limitations. The certificate number and ratings are checkable facts an engine can quote; "full-service MRO" is not. EASA issues its own Part-145 approvals. Verify current status with the issuing body.
- Part 295
The US Department of Transportation rules for air charter brokers, in force since 2019, which require a broker to disclose that it is acting as an agent or an indirect air carrier, to name the operator on request, and to make specified disclosures before payment. Stated plainly on a broker's own site, Part 295 compliance is a fact engines repeat for "how to book" questions. Verify current status with the issuing body.
- AOC
Air Operator Certificate: the ICAO term for the approval a national authority grants an operator to fly commercially, used in the UK, EU, the Gulf states and most of the world where the United States says "Part 135 certificate". An entity block for a non-US operator states the issuing authority and the AOC number; it does not claim a US certificate. Verify current status with the issuing body.
- ARGUS ratings (Gold, Gold Plus, Platinum)
Third-party safety ratings for charter operators from ARGUS International, in ascending tiers: Gold, Gold Plus and Platinum, with the higher tiers involving an on-site audit. Buyers and brokers use them, together with ARGUS TripCHEQ trip reports, to vet an operator beyond its certificate. State exactly the tier you hold and the audit date; a lapsed rating stated as current is an accuracy flag. 32 Verify current status with the issuing body.
- WYVERN (Registered, Wingman, Wingman Certified) and PASS
Third-party safety standards for charter operators from WYVERN, in ascending tiers described as Registered, Wingman and Wingman Certified, with a PASS (pilot and aircraft safety survey) report available per trip. Like ARGUS tiers, they are verifiable statements an engine reaches for when a prompt includes "safe"; state the exact tier and date. 32 Verify current status with the issuing body.
- IS-BAO Stage 1–3
The International Standard for Business Aircraft Operations, an audited safety-management standard administered by the International Business Aviation Council, registered in three stages: Stage 1 establishes the safety management system, Stage 2 shows it is being applied, Stage 3 shows it is embedded and sustained. Operators state the stage and registration date; the stage is not a rating of aircraft or crews. 3 Verify current status with the issuing body.
- IS-BAH
The International Standard for Business Aircraft Handling, the ground-handling counterpart of IS-BAO for FBOs and handling agents, also registered in stages. For an FBO, the stage held and the registration date belong on the location fact page alongside hours, fuel, hangar capacity and customs arrangements, because "which FBO" prompts are answered from exactly those facts. Verify current status with the issuing body.
- NATA Safety 1st
The National Air Transportation Association's safety training and professional-development program for FBO, ground-handling and line-service staff. An FBO that states its Safety 1st participation, with the date, gives crews and dispatchers a checkable fact; it is a training program, not an audit standard, and should not be described as a rating. Verify current status with the issuing body.
- SMS
Safety management system: the structured set of policies, risk assessment, assurance and promotion activities an operator, FBO or MRO runs to manage safety, and the basis of IS-BAO and IS-BAH registration. Corporate travel policies increasingly ask whether a charter operator has an SMS and an emergency response plan; a dated SMS statement on the safety page is the answer engines can quote. Verify current status with the issuing body.
- ICAO code
The four-letter airport identifier used in flight planning — KTEB for Teterboro, EGGW for London Luton, OMDW for Dubai Al Maktoum — as distinct from the three-letter IATA code passengers see. Crews and dispatchers ask by ICAO code, so an FBO or operator page that states both codes is matched for both forms of the question.
- OAI-SearchBot, GPTBot, ClaudeBot, PerplexityBot, Google-Extended
Answer-engine crawlers: the user agents that fetch pages for AI search and answers (OAI-SearchBot, Claude-SearchBot, PerplexityBot, Googlebot, Bingbot), for user-triggered fetches (ChatGPT-User, Claude-User, Perplexity-User) and for training (GPTBot, ClaudeBot, CCBot), with Google-Extended and Applebot-Extended as training control tokens. Blocking a training bot does not remove a site from ChatGPT search; blocking a search bot does. Our robots.txt allows all of them. 4