AI visibility in aviation: how ChatGPT, Claude and Perplexity decide which companies to recommend

Last week two companies signed up to The Flying Engineer and told us how they found us. One found us by asking ChatGPT. The other found us through Claude. Neither had searched Google.

That matches what we see in our own numbers. Traffic from AI assistants has spiked, and over the last two months it has overtaken the traffic we get from Google search.

For aviation companies this changes one practical question: when a buyer asks an assistant "who can charter a jet from Dallas to Aspen?" or "where can I buy a serviceable PW127 starter generator?", is your company in the answer?

How AI assistants find companies

ChatGPT, Claude, Perplexity, Microsoft Copilot and Google's AI Overviews work in two ways.

1. What the model already knows. Large language models are trained on public web pages. If your company was described clearly and often on the public web, the model may know you.
2. What it looks up now. For questions about companies, prices and availability, most assistants now run a live web search, read the top pages and write an answer from them, usually with links to the sources.

The second route is the one you can influence quickly. The assistant reads a handful of pages and repeats what they say. If those pages are directories, comparison articles and company profiles, the companies listed there are the ones that get named.

What the assistants look for

From the pages that get cited for aviation questions, a few patterns are consistent.

• Plain facts in text. Base airport, aircraft types, certificates, services, countries served. An assistant can quote "Part 135 operator based at Dallas Love Field (DAL) with a Citation XLS and a Challenger 350". It cannot quote a slider of photos or a tagline.
• The same facts in several places. If your website, your directory listings and your LinkedIn page agree on what you do and where, the assistant treats it as reliable. Conflicting addresses or old fleet lists make it cautious.
• Independent sources. A company that only describes itself is weaker than one that also appears in directories, industry news and association member lists.
• Pages that answer the actual question. "Private jet charter in Dallas" pages get cited for Dallas questions. A generic "Our services" page rarely does.
• Pages the assistant is allowed to read. Some websites block AI crawlers by accident, through a security plugin or a copied robots.txt rule. If GPTBot, ClaudeBot or PerplexityBot are blocked, those assistants can't read your site at all.

Practical steps for aviation companies

None of this needs a new website. Most of it is an afternoon of work.

1. Ask the assistants about yourself. Ask ChatGPT, Claude and Perplexity the questions your buyers ask: "best aircraft maintenance in [your city]", "who sells [your product] in Europe", "private jet charter from [your airport]". Note who gets named and which pages are cited. Those pages are where you need to be.

2. State the facts on your own site. One page per main service and location, with the details written out: airports, aircraft, certificates (Part 135, EASA Part-145, AS9120, FAA repair station number), response times, contact. Add your company details as structured data (schema.org Organization or LocalBusiness) so the facts are machine-readable.

3. Check your robots.txt. Open yourdomain.com/robots.txt and make sure it doesn't block GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot or Google-Extended, unless you've decided to.

4. Be listed where the assistants look. Aviation directories, association member lists (NBAA, EBAA, ASA, ARSA and national associations), airport FBO and tenant pages, and trade press. Keep the details identical everywhere.

5. Keep it current. Assistants favour pages with recent dates. An old fleet list or a closed base on a directory profile works against you. Update your listings when something changes.

6. Get mentioned in articles. Comparisons and "best of" articles are cited heavily. Being included, with accurate facts, matters more than any single page on your own site.

What we do at The Flying Engineer

We built the new version of The Flying Engineer with this in mind. Every company page states what the company does, where it is based and how to contact it, in plain text. Company pages carry structured data. Our robots.txt welcomes the main AI assistants by name, and we publish an llms.txt file, a plain-text guide that tells AI assistants how the site is organised.

Listed companies can add their own description, photos, logo and website link, and paid plans add placement and AI-visibility work on the listing. The details are on our plans page.

A note on ownership: I'm Radu Balas, founder of The Flying Engineer. I also run an agency, radubalas.com, which does SEO and AI visibility work for aviation companies. The steps above work whether or not you use either.

Quick checklist

• Ask three assistants the questions your buyers ask, and note who gets named.
• One page per service and location, with facts written out.
• Organization or LocalBusiness structured data on your site.
• robots.txt allows the AI crawlers you want.
• Identical company details on your site, directories and LinkedIn.
• Listed in the aviation directories and association lists your buyers use.
• Listings updated whenever fleet, bases or services change.

If your company is already on The Flying Engineer, claim your listing and check the details are right. If not, you can add it here.