Your airline flies the routes. The load factors on KL to Kota Kinabalu and Penang to Singapore are healthy, the loyalty programme is active and the brand ranks for its own name on Google. Yet when a regional sales director asks ChatGPT which airline to use for weekly trips to Kuching, or a travel manager asks for the best carrier from KLIA to Jakarta for staff, your name is missing or described from a three-year-old blog post. That gap in airline brand ChatGPT recommendations is now a commercial problem, not a curiosity.

This article explains why Malaysian carriers fall out of AI answers, what travellers and corporate buyers in Malaysia actually ask, how to audit your position in a week, and which fixes give ChatGPT accurate facts to quote. It is written for commercial, marketing and corporate sales leaders at established airlines and airline groups.

Why airline brands go missing from ChatGPT recommendations

ChatGPT names an airline when it can find clear, current text that links the carrier to the route, cabin and traveller need in the question. Airlines often hold their best facts inside booking engines, apps and fare rules that a crawler cannot read, so other sites end up telling your story.

ChatGPT answers from two places. The first is what the model learned during training, which is fixed at a point in time and blurry on route launches, new aircraft or changed baggage rules. The second is live web search, where ChatGPT reads current pages and shows the sources it used.

OpenAI's crawler documentation says OAI-SearchBot "is used to surface websites in search results in ChatGPT's search features", and that "sites that are opted out of OAI-SearchBot will not be shown in ChatGPT search answers." GPTBot is a separate crawler for model training, and the two are controlled independently. Airline sites run heavy bot protection to stop fare scraping, and that protection can block AI search crawlers without marketing ever being told.

Why OTAs and blogs win the citation

Look at what a traveller's question needs: which carriers fly the route, which cabin suits a business trip, how much baggage is included, what happens if the flight is delayed. A travel blog comparison or an OTA guide answers that in one readable page. Your own site may answer it only after a search, a fare selection and three clicks into fare conditions.

Google's guidance on its AI features makes the same point for Search. It recommends "making sure that important content is available in textual form" and says there are no special requirements beyond normal indexing. ChatGPT is a different product, but the lesson carries across: facts that appear only after a booking query are hard for any AI system to cite.

Why leisure framing hides corporate strengths

Most public writing about Malaysian airlines is about cheap fares, holiday routes and seat sales. Corporate strengths such as frequency on trunk domestic routes, flexible fare families, lounge access, corporate programmes and disruption support are often described only in sales decks and contracts. If those facts never appear in public text, ChatGPT has nothing to say about them when a business buyer asks.

What buyers in Malaysia actually ask AI

Travellers and travel managers ask AI about airlines by route, flexibility, reliability and price, in English, Bahasa Malaysia and Mandarin. The prompts below are realistic patterns to test, grouped by buying stage.

Exploring options

  • "Which airlines fly direct from Kuala Lumpur to Kuching, and which is best for business travellers?"
  • "Full-service or low-cost for a two-day work trip KL to Bangkok?"
  • "Syarikat penerbangan mana yang terbaik untuk perjalanan kerja ke Sabah?"

Comparing carriers

  • "Compare checked baggage and change fees for KL to Singapore on the main Malaysian airlines"
  • "Which Malaysian airline has the best corporate travel programme for SMEs and large companies?"
  • "从吉隆坡飞槟城，哪家航空公司比较准时？" (Which airline is more punctual from KL to Penang?)

Checking risk before booking

  • "What are my rights in Malaysia if my flight is delayed by five hours?"
  • "Does [airline] refund to the original card or only give a credit shell?"

That last group matters. The Star reported that under MAVCOM's enhanced Malaysian Aviation Consumer Protection Code, effective 1 September 2024, passengers can get a refund if a flight is delayed five hours or more, refunds to the original payment method must be offered first, and cancelled flights must be removed from booking platforms promptly. Travellers ask AI about these rights, and whichever page explains them clearly is the one that gets quoted, often alongside your brand name.

How to check where you stand

You can build a baseline in a week with a spreadsheet, a few colleagues from commercial and digital, and a fixed prompt list. The aim is a record you can repeat each month, not a screenshot for one meeting. Our measurement method uses the same approach at a larger scale.

Step 1: Confirm OpenAI can crawl your public pages

Check robots.txt on your main domain and any separate holidays, cargo or loyalty domains for rules on OAI-SearchBot and GPTBot. Then ask your web or security team whether the CDN, WAF or anti-scraping service challenges these crawlers. You can keep booking endpoints protected while allowing route, policy and corporate pages to be read.

Step 2: Build a fixed prompt set from your network

Write 30 to 50 prompts. Cover your top routes by revenue, your domestic East Malaysia sectors, key regional routes from KLIA and Penang, and the corporate questions your sales team hears. Add prompts in Bahasa Malaysia and Mandarin, and at least five that name your airline directly so you can check accuracy.

Step 3: Run each prompt several times

Use a logged-out browser or fresh account and run each prompt three times. Answers vary between runs and users, so one result proves little. Note whether ChatGPT searched the web, because that tells you whether live pages shaped the answer.

Step 4: Log what each answer says

Record five things per run: whether your airline is named, whether the facts are right (routes, baggage, fare families, lounge access, alliance or codeshare partners), which rivals appear, how the tone compares, and which sources are cited. The cited sources column shows which pages ChatGPT trusts for your routes.

Step 5: Trace each wrong fact to a page

For every error, find where it came from. A wrong baggage allowance often traces back to an old blog post or a fare rule your own site still shows on a forgotten page. Fix your page first, then ask the third party to update.

How to fix it

Make your own public pages the clearest source of route, product and policy facts, then make sure trusted third parties repeat them. These steps are specific to how airlines in Malaysia publish and sell.

Publish route pages that answer questions

Create a permanent HTML page for each key route, separate from the booking widget. In the first paragraph, state the airports, typical weekly frequency, aircraft type, cabins offered, flight time and any codeshare partners. Add a short section on who the route suits, for example day-return business trips or onward connections. Update the page when schedules change and show a visible "last updated" date.

Write fare families and baggage in plain text

Put fare family names, included baggage, change and refund conditions and seat selection rules on a readable page, not only inside the booking flow or a PDF of conditions of carriage. Use a comparison table:

Fact travellers ask What to state in text
Checked baggage Allowance per fare family and route type
Changes Whether allowed, fee basis, cut-off time
Refunds Conditions, method of refund, typical timing
Delays and cancellations Your options in line with the MAVCOM code
Corporate benefits Programme name, who qualifies, what is included

State your passenger rights commitments clearly

Write a plain page on what passengers can expect when a flight is delayed, cancelled or rescheduled, using your own policies and the requirements of the Malaysian Aviation Consumer Protection Code. Have legal review it, keep it current, and link it from route and fare pages. A clear, accurate page here protects the brand when travellers ask AI about bad experiences, because the alternative source is usually a complaint thread.

Give corporate travel its own public pages

If your corporate programme is described only in sales decks, AI cannot recommend it. Publish a corporate travel page covering eligibility, benefits, flexibility, dedicated support channels and how travel management companies book it. Add short pages for priority sectors you serve, such as oil and gas crew movements to Miri or Labuan, or frequent Klang Valley to Johor and Penang business travel, as long as every claim is factual.

Add Airline structured data

Schema.org has an Airline type, a kind of Organization, with an iataCode property alongside standard fields such as legalName, url, logo, parentOrganization and sameAs. Mark up the corporate site so the airline, its IATA code, its parent group and its official profiles are stated consistently in machine-readable form. Structured data must match the visible page and does not guarantee anything, but it reinforces the same entity facts. Our technical AI visibility work often starts here.

Keep names consistent and earn credible coverage

Use one form of the airline name, group name and loyalty programme name across the website, apps, press releases, LinkedIn and Bursa or group filings. Then look at the sources ChatGPT cited in Step 4. Aviation and business press, airport route announcements and tourism board pages are worth more than another seat sale release. When you announce a route, put the airline name, both airports and the start date in the first sentence so every article that reuses it carries the facts AI needs. Our content service covers this kind of factual writing in English, Bahasa Malaysia and Chinese.

What to measure

Track four things monthly from the same prompt set: mention rate, accuracy, citation share and sentiment against regional peers. Read them as trends over several months, because individual answers shift between runs and users.

  • Mention rate: share of prompt runs naming your airline, split by route, language and leisure versus corporate prompts.
  • Accuracy: share of mentions with correct routes, baggage, fare rules and policy facts. Fix wrong refund or baggage facts first, since they cause complaints.
  • Citation share: how often your own pages are among cited sources compared with OTAs, blogs and rival airlines.
  • Sentiment and framing: whether answers describe you as reliable, flexible or budget, and whether that matches your positioning.

Report with the method stated: prompts, runs per prompt, dates and engines. Check again after schedule changes, disruption events and major route launches, since those are when stale facts do the most damage.

Common mistakes

  • Blocking AI search crawlers with anti-scraping rules that were only meant to protect fare data.
  • Keeping baggage and fare rules inside the booking flow where no crawler or AI tool can read them.
  • Writing only leisure content and leaving corporate strengths in private sales decks.
  • Letting old route and policy pages stay live after schedules or fare families change.
  • Making superlative claims such as "most punctual airline in Asia" without a named, current source.
  • Judging visibility from one screenshot instead of a repeatable prompt panel.

How Shubi helps

Shubi is an AI visibility company based in Malaysia. We run the prompt panels, crawler checks and source analysis described here across routes, languages and AI engines, then help airlines publish route, policy and corporate facts that AI tools can read and quote. Our AI Visibility Intelligence service is the usual starting point, and we do not promise rankings or guaranteed mentions. To talk it through, Message Shubi on WhatsApp.