Editorial Article

One Point Six Per Cent of the Airport Parking Market Sets the Price on the Rest

Updated November 19, 2026

An airport car park is one of the last pieces of commercial real estate in the world that sells the same space two ways at once. A driver can arrive and take whatever is there at the rate on the board, or a traveller can commit three weeks ahead to a named product at a price a machine set. The first is a rent. The second is a contract with a date on it, and the difference between them is now most of the money in the business.

Pheonix Research models the global market for airport parking and pre-booked reservation platforms at USD 21.52 billion in 2025, the base year for this forecast, on 1,020 million parking stays. It reaches USD 23.14 billion in 2026 and USD 38.49 billion by 2033 — 7.54 per cent a year from that base. Volume, over the same window, grows at 2.98 per cent, from 1,050 million stays in 2026 to 1,290 million in 2033. Two lines, one gap of four and a half percentage points a year, and the gap is the business. A bay is perishable inventory with a clock attached. Selling the clock rather than the concrete is what converts a fixed asset into a yield instrument.

Parking is the largest non-aeronautical line at the world’s airports

That is not how the sector usually describes itself, and the description is wrong by its own numbers. ACI World’s 2026 Airport Economics Report, drawing on financial-year-2024 data from more than 1,000 airports, puts non-aeronautical income at 36.7 per cent of total airport revenue, or USD 7.57 per passenger. Within that, car parking is 24.1 per cent globally and 43.2 per cent in North America — ahead of retail at 23.1 per cent, property at 15.8 per cent and food and beverage at 6.3 per cent. On a per-passenger basis parking is worth USD 1.82, which is 8.84 per cent of everything an airport earns, aeronautical and non-aeronautical together.

Applied to ACI’s own traffic numbers — 9.4 billion passengers in 2024 and 9.8 billion projected for 2025 — that per-passenger figure implies USD 17.15 billion and USD 17.88 billion of on-airport parking revenue in those two years. The market Pheonix measures is larger, because it includes off-airport and near-airport inventory and the booking layer that now sits across both. The on-airport figure is the anchor; every value above it is Pheonix’s construction, and the report says so at the point of use rather than in a footnote.

The context matters as much as the level. Total airport revenue remains roughly 2 per cent below 2019 while passenger traffic has passed it. In a sector that has recovered traffic and not revenue, the line that did recover is the one now under pressure to produce — which is why a 24 per cent non-aeronautical share attracts capital, board attention and, increasingly, a pricing strategy of its own.

One point six per cent of the market sets the price on the other 98.4

The mechanism of that strategy is a small number sitting on a large one. In 2026, 32 per cent of gross value moves through a pre-booking — USD 7.41 billion of the USD 23.14 billion. By 2033 the share reaches 45 per cent and USD 17.32 billion. A platform take rate of 5 per cent widening to 6 per cent produces a net fee pool of USD 0.37 billion today and USD 1.04 billion in 2033, a 15.9 per cent compound rate against the market’s 7.54.

The pool is 1.6 per cent of the gross market in 2026 and 2.7 per cent in 2033. It is not where the money is. It is where the price is decided. A platform holding a third of a market’s inventory, with the booking curve behind it, knows what a bay is worth on a given Thursday six weeks out. That knowledge sets the reference price against which the drive-up rate is read, and the drive-up rate is the other 98.4 per cent. The airport keeps the gross; the platform takes a thin slice of it and, in doing so, acquires the yield curve.

There is a second-order point in the take rate that gets less attention than it deserves. Moving from 5 to 6 per cent is a 20 per cent increase in the platform’s revenue on identical bookings — close to three years of growth at the market’s own rate, earned by negotiation rather than by demand. When a business can improve its economics that much through a single percentage point, it will spend its management attention on renegotiating that point rather than on expanding the inventory it sells.

Pre-booking tracks stay length, not country

The penetration numbers look arbitrary until they are sorted by trip type. Dallas Fort Worth has run pre-booking at roughly 25 per cent of transactions. Charlotte sits near 65 per cent. Australian and New Zealand airports exceed 90 per cent on long-stay products. National income, smartphone penetration and app maturity do not explain that spread. Stay length does, and behind stay length sits the question of whether landside capacity is scarce enough to make a reservation worth holding.

A business traveller parking for four hours arrives without choosing; the product is bought at the barrier because the decision has not been made until the car is at the ramp. A leisure traveller parking for eleven days is making a purchase days or weeks ahead, comparing products, and will accept a restriction — a fixed entry window, a non-refundable rate — in exchange for a price. Pre-booking grows where the second customer is a larger share of stays. That is a mix constraint on the whole growth story: the reservation layer cannot outrun the composition of demand, whatever the apps do.

The substitute pays the airport to substitute

The drive-up parker is not only being competed for by the booking layer. At some airports the parker is being removed altogether, and the removal is being monetised. At Boston Logan close to 30 per cent of passengers used a ride-hailing app in 2024 — the single most common way of reaching the airport — and Massport has raised its per-trip fee from USD 3.25 to USD 5.50, with USD 7.50 scheduled from July 2027.

That is a rational response and a revealing one. The airport loses a parking stay worth about USD 21 at the base-year average and replaces it with a toll levied on a vehicle that never occupies a bay. The fee is smaller, and it costs nothing in dwell, staffing, cleaning or capital. But it is a toll on access rather than a sale of capacity, and it converts a customer relationship into a levy. Airports leaning on it are trading the yield instrument for a simpler line item.

Regulation moved from the barrier to the booking page

Which is why the regulatory surface has moved with it. For twenty years the rules governing airport parking concerned the forecourt: signage, tariffs, the terms printed on a ticket. They now concern the funnel. Britain’s Digital Markets, Competition and Consumers Act 2024 makes drip pricing unlawful, so the administration fee, the booking charge and the card surcharge have to appear in the headline price rather than at checkout — a change that lands directly on the pre-booked product, because that is where those fees lived.

Alongside it, the Competition and Markets Authority’s price transparency guidance sets out how the sector is expected to present comparison prices, and the Information Commissioner’s guidance on automatic number-plate recognition governs the data captured at the barrier: lawful basis, retention, and the processing agreement between the operator and every vendor in the chain. In the United States, FAA Order 5190.6C sets how the agency reviews airport rates and charges, which is the mechanism through which a parking tariff becomes a compliance question rather than a commercial one.

In the United States, parking is not a profit centre

That difference is structural. Under 49 U.S.C. 47107(b) and 47133, airport revenue must be used for airport purposes. Parking receipts are airport revenue, which makes them a cost-recovery instrument and, at bond-financed airports, a covenant input rather than a free profit centre. An American airport raising its parking rate is not simply optimising a margin. It is adjusting a rate subject to grant assurances and, often, to airline consultation.

European airport groups disclose parking as a commercial line with more freedom, and the disclosures show what that freedom produces: Aena’s car parks earned EUR 204.1 million in 2024 and EUR 221.8 million in 2025, while Fraport’s parking line moved from EUR 101.6 million to EUR 108.4 million. Both are modest, high-margin, capacity-constrained businesses — and both sit on the asset rather than at the funnel, which is precisely where the growth is not.

Who ends up holding the transaction

The competitive contest is over that question. Metropolis’s acquisition of SP+ created a combined business valued at roughly USD 1.5 billion, and the price was paid for sites and the transactions running through them rather than for the concrete. Around them sit the operating groups — APCOA, Q-Park, NCP, ABM — the airport groups that own the inventory, Aena and Fraport among them, and the technology layer the booking platform depends on: SKIDATA at the barrier, Flowbird — acquired by EasyPark Group in 2025 and now trading as Arrive — in municipal and on-street payment, Amano in Japan and Nedap in bay-level detection, and Park Assist, part of TKH Security in guidance.

These are not all bidding for the same purchase order. The operators bid for site contracts, the technology vendors bid for the specification inside them, and the platforms bid for the position from which a transaction is priced. The third contest is the one that decides who captures the yield, and it is the least visible of the three to the airport that signs the contract.

What would move this market rather than re-mix it

One outcome would change the size of the market rather than its composition: pre-booked penetration stalling. If the share settles near 40 per cent instead of reaching 45, value growth converges on volume growth — roughly 3 per cent a year rather than 7.5 — and the market loses about two-thirds of its forecast expansion without any external shock. The penetration number, not the CAGR, is the line to watch.

Three things decide it. Whether airports release inventory to the booking layer or hold it back to protect the drive-up rate. Whether dynamic pricing reaches the day-of-week and time-of-day granularity that airlines and hotels have run for decades. And whether the identity of the vehicle arrives from the building or from the car: a connected vehicle that knows when it entered and left makes a booking platform’s data better and makes a barrier’s data redundant. Any airport signing a seven-year operating agreement should know which of those assumptions the contract embeds.

For now the shape of the business is set. Parking is the largest non-aeronautical line at the world’s airports, the reservation layer is the smallest meaningful line inside it, and the small line is the one that decides what the rest costs.

Related research

Sources and method

Market values, the pre-booked share and take rate, the fee pool, per-stay values and regional tables in this article are drawn from Pheonix Research’s model of the global airport parking and pre-booked reservation platforms market (base 2025, forecast to 2033). All figures are Pheonix estimates stated in USD.

Document-level sources are public instruments, official statistics and company disclosures: ACI World’s 2026 Airport Economics Report and its traffic data; Aena and Fraport annual results; the Massachusetts Port Authority fee schedule; Dallas Fort Worth and Charlotte airport pre-booking disclosures; the UK Digital Markets, Competition and Consumers Act 2024; Competition and Markets Authority price transparency guidance; the Information Commissioner’s Office guidance on automatic number-plate recognition; FAA Order 5190.6C; 49 U.S.C. 47107(b) and 47133; and Companies House filings. Company identification uses Wikidata entities — SKIDATA Q1168035, Amano Q11284890, SP+ Q7598289, Flowbird Q131636808, Aena Q19826726, Fraport Q643578, APCOA Q296108, Q-Park Q1127798, NCP Q6971273, ABM Q4650338.

The 30 per cent of this analysis that is not document-level review is Pheonix primary research: structured interviews with airport commercial teams, parking operators and platform executives; open-ended interviews; survey work across operator and traveller samples; conference-floor discussion at industry gatherings; live booking-funnel teardowns; and vendor briefings. Where a conclusion rests on that primary work rather than on a published document, the text says so.