Frank Bosteels

The through-line

Frank Bosteels talks about travel technology the way someone talks about a tool that has already been absorbed: interesting, useful, not existentially threatening. "Wij zijn niet bang van AI. Wij denken dat hij een mooie opportuniteit biedt om onze job nog beter te doen." [1] The industry, in his telling, has been here before with the internet, which it eventually understood "geen bedrij is maar een opportuniteit en een hulpmiddel" [2]. What sits underneath every one of his positions is a conviction that demand for travel is not negotiable: "Want uiteindelijk reizen, daar gaan de mensen nooit van afzien. Dat is een elementaire behoefte. Dat is gelijk als eten en drinken." [1][2] Everything else, the pricing logic, the shape of the distribution channel, the destination mix, is a variable arranged around that constant.

The second recurring move is deflationary. Where others see AI, he sees data, and he punctures the hype with the same image every time: "een AI zonder data is eigenlijk een Porsche die in de garage staat zonder brandstof" [1][2]. Where others see the death of the travel agent, he sees a format change. Where others see boycotts and collapse, he sees a wave that will reverse when prices fall far enough.

On AI, and why data is the real story

The current capability boundary is clear to him: AI agents can already execute a simple, fully specified booking, a fixed airline, a known hotel chain, fixed dates, but complex trips remain out of reach [1][2]. Two problems stop it short of the interesting cases. The first is liability, which does not transfer to the machine: when an AI booking goes wrong, the traveler carries it in full [1][2]. The second is judgment, since agents cannot reliably tell an online scam from a legitimate offer [2].

The underreported story, in his view, is not the model but the fuel. Revenue management systems run on historical booking data and event calendars, and the industry consistently underestimates how central that data is relative to the AI conversation happening on top of it [1][2]. He locates the relevant technology close to home, noting Lighthouse in Gent and adding that "everything happens in Gent these days when it comes to technology" [1][2].

There is one place where he draws a line. AI-generated destination imagery in travel marketing is misleading advertising, and while promotional photography has always been retouched, generating places that do not exist is a different category of problem [2]. The consequence is concrete: people have walked into travel agencies carrying AI images of a Cyclades-style town, asking to book a trip there [1][2].

On why booking early became cheaper

The pricing logic of travel was inverted by the low-cost carriers, Ryanair foremost among them [1]. Where the old model discounted unsold last-minute inventory, the new one gives the cheapest seats to the earliest bookers and raises prices as the aircraft fills. The reasoning is unsentimental: "iemand die nu boekt, die moet wel reizen, die heeft geen keuze en he is gonna pay whatever it takes" [1][2]. That logic did not stay with the low-cost airlines. Legacy carriers and hotels adopted it wholesale through revenue management, which is precisely why the data question matters [1][2].

The counterweight is the traveler's own opportunism: "er is niks zo opportunistisch als een reiziger. Die voelt die ruikt ergens als er een goede deal te scoren is" [2]. That instinct explains an unusual summer. Normally almost everything is sold out by this point in the season, but higher post-pandemic prices, geopolitical uncertainty and good weather at home have pushed Belgians to postpone and wait for last-minute deals [1][2].

On the reinvention of the travel agent

He rejects the decline narrative on the numbers: "Voor elke reisagent die stopt zijn er nog altijd meer nieuwe die beginnen" [1]. The new entrants look nothing like the old ones. They are hyper-specialized, built around a single destination such as Tanzania, or around horse-riding holidays, or around a narrow demographic, and what they sell is expertise and a complete experience rather than a transaction [1][2].

His analogy for the shift is the corner grocery store: neighborhood grocers disappeared and then came back as franchise convenience formats. The concept adapted to a new reality instead of dying [1].

On how people choose where to go

Destination has stopped being the first decision, at least for younger travelers. The activity comes first, wellness, cooking, diving, horse riding, digital detox, and the place follows from it; collecting countries off a bucket list is finished [1][2]. Climate is redrawing the map alongside that. Summer demand is moving to what used to be off-season territory, the Alps, Scandinavia and Scotland, as people escape 45 degree heat, while ski resorts struggle in winter for lack of snow [1].

His own rule for value is contrarian and simple. Asked where you should travel, he answers: "naar een plaats waar niemand naartoe wil. Ik bedoel de plaats die het meest gegeerd is op het moment is natuurlijk het duurste, want iedereen wil daar naartoe." [1][2] Panama and Kythnos sit in that under-the-radar category [2].

On summer long-haul, and flying better

Long-haul in summer is usually a bad trade. Southeast Asia is in monsoon, southern Africa goes dark by five in the afternoon, Tokyo sits on the same latitude as Marseille and hits over 40 degrees, and Florida is humid [1][2]. His conclusion is that everything you actually need is within a two to three hour flight radius of Belgium [1][2]. When you do fly long-haul, the aircraft matters: modern types like the Airbus A350 cut jet lag through adaptive cabin lighting matched to time zones and better cabin pressure regulation, and the physical difference on arrival is noticeable [1][2].

On boycotts, waves and the American market

Bookings from Belgium to the United States are down roughly 30% since March on the back of Trump-related boycotts [1][2]. He is careful about what that actually means for the destination: Europeans are a small share of US tourism, with Canadians and South Americans mattering far more [1][2]. And he expects the mechanism to reverse itself. Vacancy produces price drops, a favorable dollar sharpens them, and the opportunistic traveler responds. Tourism moves in waves [1][2].

Takeaways

• AI agents can complete a simple booking with a fixed airline, hotel chain and dates, but complex trips are out of reach and the traveler retains full liability if it goes wrong [1][2].
• The real story in travel tech is data, not AI: "een AI zonder data is eigenlijk een Porsche die in de garage staat zonder brandstof" [1][2].
• Book early. Low-cost carriers flipped the logic so the first bookers pay least, on the reasoning that a last-minute booker "is gonna pay whatever it takes", and airlines and hotels now all price this way [1][2].
• Travel agents are not disappearing but respecializing, around single destinations, single activities or narrow audiences, the way corner grocers came back as franchise convenience formats [1][2].
• Choose the activity first and the destination second, and for value go "naar een plaats waar niemand naartoe wil", because the most coveted place is by definition the most expensive [1][2].
• Skip long-haul in summer: monsoon in Southeast Asia, 5pm darkness in southern Africa, 40°C in Tokyo. Everything you need is within a two to three hour flight [1][2].
• On long-haul, the aircraft type matters: the A350's adaptive lighting and cabin pressure regulation measurably reduce jet lag [1][2].
• Treat AI-generated destination imagery as misleading advertising; travelers have already asked to book trips to Cycladic towns that do not exist [1][2].