Brussels Turns Small Platform Deals Into Entrenchment Cases

Wire copy treated Luxembourg as a blessing of platform-entrenchment theory: buy an adjacent funnel, keep a tiny increment, still get blocked. The residual is narrower. Booking already sold Etraveli flights under contract. The Court still treated ownership of that funnel as a change in market structure, not a slightly thicker affiliate agreement.

The Court kept a veto the share math could not carry

On 9 September the General Court dismissed Booking Holdings’ challenge to the Commission’s 2023 prohibition of its €1.63 billion purchase of Etraveli Group. Reuters reported the headline: Brussels said the deal would have built a travel ecosystem rivals could not match, and Luxembourg agreed.

The press release is sharper. Judges accepted that the Commission’s hotel-OTA share increment was error-ridden and could shrink to a few tenths of a percent, and that Brussels had not shown hotels would pull inventory from rivals or pay higher commissions. That would have ended a conventional merger case. It did not end this one.

A concentration can lock in a weak competitive structure even when it barely moves market share.

The judgment in Case T-1139/23 treats reverse leveraging as a lawful theory of harm: Booking need not be dominant in flights to use flights to deepen hotel dominance. Cross-selling a room after a ticket is not “competition on the merits” under merger control simply because customers are not forced.

The partnership still exists. Ownership was the prohibited step.

The question the coverage skips is the one Booking litigated. Since 2019 it has sold Etraveli flights under its own brand. The UK Competition and Markets Authority, on the same deal, recorded what the contract withholds: Booking does not set flight prices, faces a cost disadvantage versus owning the stack, lives under a metasearch cap that protects Etraveli’s own sites, and sits in a term-limited affiliate deal Etraveli can exit. Internal papers treated the partnership as a bridge. “Project Bahamas” was meant to replace it.

If the harm is that flights are an important customer-acquisition channel, why is buying the channel worse than renting it?

The Court refused to let that counterfactual decide the case. It rejected Booking’s claim that the Commission had assumed a “zero-flights” world. Brussels accepted that Booking would probably keep selling flights. The judges then said the dispute was immaterial: under either counterfactual the hotel-share increment was still marginal. The prohibition does not require a large extra dollop of cross-sell versus the affiliate contract.

It rests on structure. The hotel-OTA market already runs on strong network effects and a wide gap — the Commission put Booking’s EEA hotel-OTA share in the 60–70% range. A small extra scale can still make rivals’ expansion harder. The increment would also arrive through one of the few hotel-customer channels Booking does not already dominate. Owning Etraveli would make it the leader in hotels and flights, a “connected trip” ecosystem that flight-led OTAs such as eDreams, Lastminute, TUI, and Trip.com would struggle to copy.

The contract can keep putting tickets on Booking.com. It cannot make Etraveli Booking’s preferred, fully controlled flight machine, and it is not a lasting change in market structure. The UK file is blunt: a term-limited partnership is not a merger.

That is a different instrument from the Digital Markets Act’s gateway test. A week earlier, in Opera’s failed challenge over Edge, the same court held that size thresholds do not make a service an “important gateway” if usage is small and control is shared. Merger control did the inverse: it blocked a below-threshold deal the parties referred to Brussels under Article 4(5), because an already-important hotel gateway would absorb one of the remaining ways to reach customers. Google’s search rewrite under the DMA clock is the conduct statute. This case is the structural one.

What still mis-prices the next small deal

Equity still treats this as a Booking-specific embarrassment. The mis-price is the next dominant platform’s “small” adjacency — a payments bolt-on or logistics feed that already looks like a partnership. If the target is one of the last independent funnels into a networked core, Luxembourg has said tenths of a percent are not a defense.

If Booking appeals, the Court of Justice will have to say whether consolidating an already weak structure can be a significant impediment even when the deal does not make today’s competition measurably worse. Watch that question. The partnership was never the forbidden object. Control of the last channel was.