CLARK AVIATION MARKETING

SEO + GEO for flight schools

Does AI recommend your flight school?

When a student asks where to learn to fly, most schools never come up. Run the check we ran 7,650 times in our research: live, against Claude, every answer archived.

Example from our researchDallas, Texas · Part 61 · 30 answers
  1. Verified school A26/30
  2. Verified school B22/30
  3. Verified school C13/30
  4. A verified Dallas school0/30

Invisible.

Named in 0 of 30 answers, while Verified school A was named in 26.

2 verified Part 61 schools in Dallas, names withheld · 7 more schools were named 1–9 times · archived transcript af1bbd70f456

A quick verification step appears once you start typing.

We email you a link to confirm your address first; once you confirm, we send the results and may follow up about them. Your school's name is matched against our verified dataset on our servers; it is never sent to the AI. How we handle your data.

Original research

Texas–Florida study, Sept 2026 (Claude Sonnet 4.6 from memory)

71% of FAA-certified flight schools in Texas and 79% in Florida were never recommended by AI.

We asked an AI model where to learn to fly 7,650 times and checked every answer against the FAA’s list of certificated flight schools. Most schools were never named once for their own city. We measured it; this is the data.

Every verified flight school in 11 states, and how often AI names them

Preliminary · Southeast study, 2 of 3 rounds · Oct 2026 · Claude Opus 5.5 with live web search. These numbers update after round 3. This newer study gave the AI live web search; the Sept 2026 study above asked it to answer from memory. Different conditions, so the two sets of numbers measure different things; they do not contradict each other.

How AI finds flight schools: answering from memory (Texas–Florida study, Sept 2026), the AI never named 71% of FAA-certificated Texas flight schools. With live web search (Southeast study, preliminary), it never named 12%. But across all 11 states, 28% of verified flight schools weren't named once in 20 answers about their own market. What the AI finds on the web shapes who gets recommended.

Type a city or a school name. Add 61 or 141 to see one school type.

  • Where the schools are
  • Part 141
  • Part 61

Preliminary · Southeast study, 2 of 3 rounds · Oct 2026 · Claude Opus 5.5 with live web search. These numbers update after round 3. Hover or tap a state for its numbers.

State numbers

Listed: FAA-certificated schools and schools with dated evidence of operating, as of Oct 1, 2026; schools we could not yet verify are counted, never named. University and college programs are listed and labelled; the AI study does not include them. AI numbers are per market (a city and a school type) and are shown only where at least 3 schools are in the market; we never publish a school’s own result.

State numbers and the directory by state
Preliminary · Southeast study, 2 of 3 rounds · Oct 2026 · Claude Opus 5.5 with live web search. These numbers update after round 3.
StateFAA-certificated never namedAll verified schools never namedLocations listedLocations awaiting verification
Alabama16.7% · 1 of 624.1% · 13 of 5447+19
Arkansasnot shown (fewer than 3)25% · 7 of 2826+8
Florida7.9% · 6 of 7625% · 46 of 184199+31
Georgia15.4% · 2 of 1331% · 26 of 8473+29
Louisiana0% · 0 of 410.3% · 3 of 2925+13
Mississippinot shown (fewer than 3)17.9% · 5 of 2824+12
North Carolina5.9% · 1 of 1720.5% · 15 of 7372+24
Oklahoma0% · 0 of 327.9% · 12 of 4343+11
South Carolina25% · 1 of 425% · 15 of 6040+25
Tennessee0% · 0 of 733.9% · 19 of 5657+11
Texas11.6% · 5 of 4333% · 108 of 327297+101

Fewer than 3 FAA-certificated schools: not shown (AR, MS), as for markets, because the share would be a school’s own result. LA, OK, SC: few FAA-certificated schools, so read the count, not the percentage.

It also recommends schools that no longer exist.

  • Aviator College · Fort Pierce, Florida · closed April 2026

    named 980 times in the Florida answers, and 159 times in Texas, where it never operated

  • Redbird Skyport · San Marcos, Texas · closed September 2019

    named 535 times in the Texas answers

Of the school names it wrote, 59% in Texas and 33% in Florida matched no verified school; an estimated 47% and 44% of those had no evidence of existing at all.

“No evidence” means three web searches found nothing. It is not proof a business never existed.

How we measured it

  1. 10 realistic student questions per city and school type, each asked 3 times: 30 query-runs per combo.
  2. Asked unaided: no list of schools in the prompt. The model answers from what it knows.
  3. Every school name in every answer was matched against a population verified from the FAA's Part 141 certificate list and dated web evidence.
  4. A school counts as never recommended when it is not named in any of the 30 query-runs for its own city and school type.
  5. Every response is archived verbatim with a SHA-256 hash; the methodology and results are documented for peer review.
How this was measured

Methodology v2.2, Sept 2026. Model claude-sonnet-4-6, temperature 1.0, ungrounded condition (no school list in the prompt). Grounded figures are the manipulation check. 252 city-and-school-type combos × 30 query-runs (10 queries × 3 samples) = 7,560 query-runs. Unit: brand within a state (campuses collapse). Population: primary flight training providers (aircraft-based certificate and rating training); simulator and type-rating centers excluded; university and college programs outside the measured scope. Never recommended: a brand never named in any of the 30 query-runs for the city-and-school-type combos of its campuses. Map colors use the FAA Part 141 certificate tier, the only tier comparable across states.

Texas · 153 city × school-type combos · 4,590 query-runs
Never recommended
FAA Part 141 certificate30 of 42 · 71%
FAA + dated web citation194 of 215 · 90%
All included253 of 281 · 90%
Mean citation rate: ungrounded · grounded · grounding multiple
FAA Part 141 certificate5.3% · 46.1% · 8.8×
FAA + dated web citation1.9% · 45.0% · 23.7×
All included2.0% · 45.0% · 22.1×
Florida · 99 city × school-type combos · 2,970 query-runs
Never recommended
FAA Part 141 certificate60 of 76 · 79%
FAA + dated web citation98 of 117 · 84%
All included112 of 132 · 85%
Mean citation rate: ungrounded · grounded · grounding multiple
FAA Part 141 certificate6.3% · 51.0% · 8.1×
FAA + dated web citation4.8% · 50.3% · 10.5×
All included4.3% · 49.8% · 11.6×

Grounded runs had the school list in the prompt. Grounded figures are the manipulation check.

This is our own research. Every answer is archived verbatim with a SHA-256 hash, and the method and every result are fully documented.

Methodology v2.2 · research commit 3f64b16 · 252 city-and-school-type combos, 7,560 query-runs.

01/The Problem

A student in Dallas asks for the best Part 141 schools within an hour of her. Is yours a name the AI knows?

She does not open ten tabs the way she would have in 2019. She asks, reads a short answer, and writes down two or three schools to call. Whether your school is one of them is decided before she ever reaches your website.

What she sees varies: sometimes a shortlist, sometimes a cited article, sometimes a local map pack. The pattern underneath does not vary. A few schools get named again and again, most are never named at all, and the ones that are named are not always the ones still flying.

GEO
Generative Engine Optimization: making your school something AI systems can find, read and describe accurately, the way SEO once did for the ten blue links.
About halfof U.S. adults report using AI chatbots, roughly one in four every day.5,119 U.S. adults surveyed 17–23 February 2026; 33% reported using them in 2024.Pew Research Center, June 2026 ↗
51%of 14- to 29-year-olds use AI daily (22%) or weekly (29%) — the age band most of your students sit in.1,572 respondents aged 14–29, surveyed 24 February–4 March 2026.Gallup, with the Walton Family Foundation and GSV Ventures, April 2026 ↗

We measured how often that happens to real flight schools. The next section is what we found.

02/Services

Two services. Here is exactly what we do to your site.

AI visibility (GEO)

Make your school legible to the systems students now ask first: the right pages, the right facts, stated where an AI can read them.

  • A page for each place you train, naming the airport and the training you do there
  • Pages that answer what students actually ask: cost per rating, fleet and availability, instructor experience, financing, time to checkride
  • Honest comparison pages for the schools you lose students to
  • Structured data (schema.org) so your programs, location, fleet and hours are machine-readable, not just visible
  • One consistent name, address, phone and hours everywhere they appear online
  • Citations that count in aviation: FAA listings, airport and association directories, flight-training publications
  • A re-run of the same measurement, on your school, every quarter

Technical SEO

The groundwork underneath: if a crawler cannot read your site quickly and completely, neither can an AI that learns from it.

  • A crawl of your whole site, with the findings ranked by what actually costs you students
  • Page speed and Core Web Vitals on the pages prospects land on, especially on phones
  • Titles, headings and metadata that say what the page is, written once at the template level so one change carries everywhere
  • Google Business Profile and local listings for every base you operate from
  • Fixes for the quiet killers: broken redirects, duplicate pages, missing alt text, orphaned training pages

03/Who We Serve

We work with Part 141 and Part 61 flight schools.

The prospect

Flight schools compete for a specific kind of prospect: serious about becoming a pilot, evaluating multiple programs, willing to spend $15,000 to $75,000+ on training. That prospect researches more carefully than the average buyer.

How they compare schools

They compare schools across forums, Reddit, YouTube, and increasingly through AI. They evaluate based on detail — fleet size, instructor credentials, completion rates, financing options, and increasingly, whether the school shows up when they ask an AI for recommendations.

Why we know them

We understand this buyer because the founder is one: a flight student working through his own ratings, who has sat on the prospect's side of the search. We know what students look for because we look for it ourselves.

04/Process

Measure, plan, build, measure again.

  1. Measure (week 1)

    We run the check you can run above, at full size: your school against every verified school in your market, across the questions students actually ask, repeated so a single lucky answer cannot flatter anyone. You get the numbers and the archived answers behind them.

  2. Plan (week 2)

    We turn the run into a ranked list: the pages that do not exist yet, the facts an AI cannot currently read, the listings that disagree with each other. Ordered by what most likely limits how often you are named, not by what is easiest to bill.

  3. Build (months 1–3)

    We write the pages, add the structured data, fix the technical faults and correct your business information everywhere it appears. Template-level changes first, so one fix carries across every page of your site.

  4. Re-measure (ongoing)

    We run the same measurement again each quarter and show you the movement: which questions name you now, which competitors still hold the answers, what changed since the last run. Same method every time, so the comparison is real. AI recommendations change from week to week, so each report states what was measured, not a promise.

05/About

A flight student who measured the problem.

Carson ClarkFounder

Embry-Riddle Aeronautical University

Carson founded Clark Aviation Marketing while studying at Embry-Riddle Aeronautical University and working toward his own flight ratings. He saw this problem from the student's side of the search first, which is why the company starts from measurement rather than opinion.

So we measured it: 7,650 answers, every one archived. That is the company in one line: aviation knowledge from inside the training world, joined to GEO work that is checked rather than claimed. Every statement we make about your visibility comes from a run we can show you.

Based between Texas and Florida, working with flight schools in Texas and Florida today, and expanding to serve schools nationwide. Our published results cover Texas and Florida, which is what the map above shows. The research now spans eleven states: 50,000+ AI answers when complete. Founder-led and small by design.

Carson Clark on LinkedIn

06/Book an audit

Book a 20-minute Flight School Visibility Audit.

In twenty minutes we show you where your school appeared in our measurement, who was recommended instead, and the fixes we would try first.

  • Your school measured across the questions students ask, in your city and school type
  • The schools named in your place, and how often
  • The pages, facts and listings most likely holding you back

Prefer to write first? contact@clarkaviationmarketing.com · +1 (830) 431-6487

A quick verification step appears once you start typing.

We use what you send to prepare and schedule your audit, and to follow up about it. Nothing else, and we never sell it. How we handle your data.