Duty free grew to a record in 2025. Conversion is slipping

European airport duty free ended 2025 with a record: €10.13 billion in value sales. Behind that headline the efficiency numbers moved the other way. Passenger traffic grew faster than sales, and average spend per passenger fell slightly. The record is real, and it is mostly a traffic story.

In brief

Value sales rose 5.5% in 2025, but passengers rose 5.8%; spend per passenger fell 0.3%.
The DFWC monitor shows footfall and purchase rates declining between 2023 and 2025.
Growth is traffic-led, and the conversion funnel is not measured in daily operations.

The key pointEuropean duty free grew in 2025 because more people flew, not because the stores converted better. Footfall, purchase rate and spend per passenger are flat or falling, and the funnel is not measured in daily operations.

The record, and the split underneath

The ETRC Business Performance Index for full year 2025 reported €10.13 billion in value sales, up 5.5% on 2024 and the highest figure in the sector's history. Unit sales rose 4.5%. Passenger numbers grew faster, at +5.8%, so average spend per passenger fell to €10.55, a decline of 0.3%.

The index is a panel of participating retailers, with passenger data supplied by ForwardKeys; it is not the whole market. It is nonetheless the most consistent measure the sector publishes. The pattern was present through most of the year, not only at year-end: the Q3 2025 Index reported spend per passenger up 0.4%, and the full-year figure sits slightly below zero. The trend is weak, not dramatic. That is exactly why it is easy to miss.

The funnel is slipping while traffic grows

At market level, value sales are the product of three rates: the share of passengers who enter a store, the share of those who buy, and the basket they take. The ETRC figure shows only the product of those rates. The industry's traveller monitors show each rate slipping.

The DFWC Full Year 2025 Global Shopping Monitor, compiled by m1nd-set from more than 24,500 international travellers, found that the share of passengers visiting duty free fell from 44% in 2023 to 40% in 2025, and that the purchase rate declined from 28% to 26%. Average spend stayed flat at about $130. International departures meanwhile rose 7% in 2025 to 2.28 billion.

Read together, the two datasets point in the same direction: the market grew because the base of travellers grew, while the funnel that converts them did not improve. In the same monitor, the share of travellers who said they bought elsewhere rose from 10% to 14%. The two sources measure different things — ETRC covers a European retailer panel, DFWC a global traveller survey — so the exact numbers are not directly comparable. The direction is what both show.

Why the funnel is hard to see

Revenue per passenger is a compound: it bundles footfall, conversion, basket and price into one figure. When that number is flat or falling slowly, management cannot tell whether the funnel held or whether passenger growth compensated for a weaker funnel. The industry therefore tends to describe itself with a headline that cannot show its own problem.

The links of the chain also live in different systems. Passenger mix sits with the airport and the airlines, transactions with the retailer's point of sale, stock with merchandising and replenishment, and price perception in market comparisons that no store system captures. Building one funnel per location, daypart and nationality segment is a data-joining problem, not a survey problem.

The ACI Asia-Pacific travel retail study reached the same conclusion from the operator side. It found that passenger recovery has not translated evenly into commercial performance, that airports with similar traffic recovery show notably different commercial outcomes, and that digital investments have delivered mixed success particularly in data integration and conversion. Airports with similar traffic are not similar businesses. The difference sits in the funnel.

Where the demand is being lost

The monitor data and the China Trading Desk Q2 2026 survey of Chinese outbound travellers point to where the funnel leaks. Among the Chinese travellers surveyed, 62.6% could not find the brands or products they intended to buy. A further 55.3% were discouraged by prices, and 39.2% said they lacked time. Price comparison is now routine: 98% compare duty free prices against e-commerce sites at least sometimes, while only 8.3% say they do most of their shopping at the airport.

Three different losses appear here. Not finding the product is an assortment and availability failure — a product-data and stock problem. Price comparison is a value-communication problem that a single store does not control on its own. Limited time is an operational problem of routing, staffing and checkout. Each needs a different fix and a different measurement.

One counterweight sits in the same data. Staff interaction keeps rising — 51% of travellers engaged with store staff in Q1 2026 and 56% in Q2 2026, according to the DFWC KPI Monitor, and about three quarters say the interaction positively influenced the purchase. It is the strongest conversion lever in the sector. It is also expensive and hard to scale, which is precisely why it needs measurement before it can be expanded.

What changes for operators

This is not a call to buy more tools. It changes three practical things.

Report the funnel, not just revenue per passenger. For each location, daypart and nationality segment, track eligible traffic, store entry, transaction, basket and margin. The first estimate will be rough — footfall often has to come from counters, queues or sampled conversion rates — but a rough funnel beats one blended number, because it says where to intervene.
Treat "not found" as a demand signal. When a high-spend segment cannot find what it came for, the response sits in product data, assortment and availability — the same reconciliation work between stock, sales and catalogue that FMCG teams already do with retailers.
Attribute changes to the funnel. Any assortment, media, price or staffing change should be measured against footfall, conversion and basket at the same location and time. Otherwise the sector keeps debating experience versus value without evidence about which one moved sales.

The underlying shift is a reporting change: instead of asking whether sales grew, ask which of the four rates changed, and whether it is the rate the business can act on. Traffic is largely outside commercial control. Footfall, conversion and basket are not.

What this analysis does not claim

Three limits matter. First, the declines are small and survey-based; they describe direction, not magnitude, and the market is at record highs. Second, the sources cover different scopes — a European retailer panel, a global traveller survey, an Asia-Pacific airport study and a Chinese traveller survey — so they agree on direction without being comparable in size. Third, traffic-led growth is not inherently bad. More travellers is a real gain.

The uncomfortable part is what the quiet numbers imply for a traffic slowdown. In Q2 2026 the DFWC monitor recorded global international departures at 99% of the prior year, pulled down by the Middle East conflict. When traffic stalls, a sector whose growth came from traffic has no conversion reserve to fall back on. The operators best placed to manage that moment are not necessarily the ones with the biggest stores. They are the ones who can already read the funnel.

Frequently asked questions about duty free conversion

Did European duty free really reach a record in 2025?

Yes. The ETRC Index reported €10.13 billion in value sales, up 5.5% and the highest on record. The growth was driven mainly by passenger volume: traffic grew 5.8% while spend per passenger fell 0.3%.

Is conversion actually getting worse?

The DFWC monitor shows the share of passengers visiting duty free falling from 44% in 2023 to 40% in 2025, and the purchase rate from 28% to 26%. These are survey-based, small changes. The reliable part is the direction, not the exact size.

Why does the headline number not show the conversion problem?

Because revenue per passenger bundles footfall, conversion, basket and price into one figure. Passenger growth can hide a weaker funnel, which is why the funnel itself has to be reported separately.