How Airlines Actually Make Money
The global airline industry moves nearly five billion people a year and keeps somewhere around three cents of every dollar it takes. Understanding where the money actually comes from explains almost every irritating thing about buying a plane ticket.
Warren Buffett once described the airline industry as a business that had, in aggregate, destroyed more capital than almost any other over the twentieth century. He was not wrong. Airlines are capital-intensive, labour-intensive, fuel-exposed, cyclical, heavily regulated, and sell a product that expires worthless the moment the door closes. And yet some of them are extremely profitable. The difference is almost never the ticket.
The margin problem
IATA’s industry-wide figures give the clearest picture of how thin this business is. Across the whole global industry, in a good year, net margin runs in the low single digits.
| Metric | Order of magnitude | Comment |
|---|---|---|
| Global industry revenue | ≈$1.0 trillion/year | Crossed the trillion-dollar mark for the first time in the mid-2020s |
| Net profit margin | ≈3% | For comparison: consumer software runs 20–30%, supermarkets 2–4% |
| Net profit per passenger | ≈$6–7 | Roughly the price of an airport coffee, per person carried |
| Passengers carried | ≈5 billion/year | Now comfortably above pre-2020 levels |
| Load factor | ≈83% | The share of available seats actually sold — historically high |
A margin of three percent means a 3% fuel price move, a 3% demand shortfall or a 3% currency swing can erase the entire profit. This is the whole explanation for airline behaviour: an industry operating on that margin will optimise relentlessly, and will monetise anything it can legitimately charge for separately.
Where the money goes out
| Cost | Share of total | Notes |
|---|---|---|
| Fuel | ≈20–30% | The most volatile line by far. Many airlines hedge; hedging turns a variable cost into a bet, which has both saved and ruined carriers. |
| Labour | ≈20–30% | Pilots, cabin crew, engineers, ground staff. Largely fixed, heavily unionised, and rising in most markets. |
| Aircraft ownership and lease | ≈10–15% | Whether owned, financed or leased, the metal has to be paid for whether it flies or not. |
| Maintenance | ≈10% | Driven by cycles and flight hours, not calendar time. Engine overhauls are the big-ticket items. |
| Airport and en-route charges | ≈10–15% | Landing fees, terminal charges, air navigation charges. Largely non-negotiable. |
| Distribution and sales | ≈3–6% | Booking fees, agency commissions, GDS charges, payment processing. The line airlines have spent two decades attacking. |
| Everything else | Remainder | Catering, insurance, IT, ground handling, overhead. |
Revenue management: why the price keeps changing
Airlines do not have a price. They have a set of fare buckets, each with a limited number of seats, and an algorithm deciding when to open and close them based on how the flight is filling relative to a forecast.
A simplified single flight
| Bucket | Fare | Seats | Conditions |
|---|---|---|---|
| T | €79 | 12 | Non-refundable, no changes, no bag, no seat selection |
| V | €119 | 20 | Non-refundable, change for a fee, cabin bag only |
| Q | €169 | 30 | Change for a fee, one checked bag |
| M | €249 | 40 | Changeable, bag, seat selection, full mileage earning |
| Y | €480 | Unlimited | Fully flexible, fully refundable |
Two consequences follow. First, the price did not increase; the cheap inventory ran out. Second, fares can also fall close to departure if the algorithm concludes the flight will go out with empty seats — which is why last-minute bargains genuinely exist, just unpredictably.
Why business travellers subsidise everyone else
The flexible fare at the bottom of that table costs six times the cheapest one for the same seat and the same flight. The difference is entirely in the conditions, and the reason those conditions exist is fencing: airlines need a way to charge more to customers who value flexibility and less to customers who value price, without letting the second group buy the first group’s ticket. Advance-purchase requirements, Saturday-night stay rules and non-refundability are all fences. They are not arbitrary cruelty; they are the mechanism that lets a €79 fare exist at all.
Ancillary revenue: the thing that actually turned the industry around
In 2006 Ryanair began charging separately for checked bags. Within a decade, unbundling had spread across most of the industry, and ancillary revenue had become the difference between profit and loss for a large number of carriers.
| Ancillary source | What it is |
|---|---|
| Checked baggage | The original and still one of the largest. Charged per bag, per direction, and cheaper online than at the airport — deliberately. |
| Seat selection | Pure margin. The seat exists regardless; the airline is selling the right to choose it. |
| Priority boarding and fast track | Selling access to a queue. |
| Onboard sales | Food, drink, duty free. Modest revenue, high visibility. |
| Change and cancellation fees | Substantially reduced in the US market after 2020, but far from gone globally. |
| Commissions | Hotels, car hire, insurance, transfers sold at the point of booking. Very high margin because the airline carries no operational risk. |
| Cargo in the hold | On a passenger long-haul flight, belly cargo can be a meaningful share of the flight's revenue — and was the only thing keeping many wide-bodies flying in 2020. |
| Selling miles | See below. This is the big one. |
The scale is easy to underestimate. Industry-wide ancillary revenue is measured in the low hundreds of billions of dollars annually. For the most aggressive low-cost carriers, non-ticket revenue is somewhere around a third to nearly half of total revenue — meaning the fare is close to a loss leader and the business is really in the extras.
| Model | Ancillary share of revenue | Strategy |
|---|---|---|
| Ultra-low-cost (Ryanair, Wizz, Spirit, Volaris) | ≈30–45% | Lowest possible headline fare, monetise everything after |
| Low-cost (easyJet, Southwest, JetBlue) | ≈20–30% | Some inclusions retained as differentiation |
| Full-service network (Lufthansa, Delta, Singapore) | ≈10–20% | Fare families and upsell rather than pure unbundling; loyalty is the bigger lever |
| Gulf carriers | Lower | Compete on inclusion and service rather than unbundling |
The loyalty programme is the real business
This is the part that surprises people most. For several of the largest airlines in the world, the frequent flyer programme is worth more than the airline that owns it.
The mechanism: a bank issues a co-branded credit card, cardholders earn miles on everyday spending, and the bank buys those miles from the airline in cash. The airline is manufacturing a currency and selling it wholesale. It costs almost nothing to produce, is recognised as revenue on sale, and a meaningful portion is never redeemed at all.
- Delta’s American Express partnership has been publicly described by the airline as generating several billion dollars a year, with the airline having stated a multi-year ambition of around $10 billion annually from the relationship.
- During the 2020 crisis, several airlines raised emergency financing by pledging their loyalty programmes as collateral — United against MileagePlus, American against AAdvantage, Delta against SkyMiles. In multiple cases the valuation placed on the programme in those financings exceeded the entire market capitalisation of the airline.
- Air Canada bought its own loyalty programme back from a third party for over $2 billion, having previously spun it off — a fairly direct statement about which asset it considered strategic.
And distribution — the cost nobody sees
One line in the cost table deserves its own section, because it explains a great deal about the current shape of the industry. Selling a ticket costs money: the legacy GDS systems that connect airlines to travel agencies charge a booking fee per segment, agencies take commission, and payment processing takes a further cut. Across the industry this has historically run to several billion dollars a year in GDS fees alone.
That is why airlines have spent fifteen years pushing direct channels, why some have added surcharges on GDS-sourced bookings, and why IATA developed the NDC standard — a modern, API-based way for airlines to distribute their own content, with their own pricing, their own bundles and their own upsell, without paying a legacy intermediary per segment.
For the airline, NDC is a cost story and a merchandising story at the same time: cheaper distribution, and the ability to sell the ancillaries that now carry the margin. For a travel platform, it is a connectivity problem — dozens of airlines, each with its own implementation, quirks and certification process. Which is the problem Norba exists to solve.
Putting it together
A rough anatomy of a €200 short-haul economy ticket on a European low-cost carrier — illustrative, not any specific airline’s accounts:
| Line | Approximate |
|---|---|
| Base fare paid | €120 |
| Bag, seat, priority | €55 |
| Taxes and airport charges collected on behalf of others | €25 |
| — Airline revenue from this passenger | ≈€175 |
| Fuel | ≈€45 |
| Crew, ownership, maintenance, ground handling, airport charges | ≈€115 |
| Distribution and payments | ≈€8 |
| — Retained | ≈€7 |
Read that table again and most airline behaviour becomes legible. Why the bag costs extra: because without it there is no profit. Why the seat map is monetised: because it is free to produce. Why the airline wants you booking on its own app: because the distribution line is real money. Why the credit card offer appears three times during checkout: because the loyalty programme is the most profitable thing the airline owns. And why the flight is full: because at 83% load factor and three percent margin, empty seats are not survivable.
Norba is the modern bridge between travel apps and airlines. We connect booking platforms to dozens of airlines through a single, simple interface — so you can search, book, and manage flights without the complexity. Learn more at norba.io.