Bedbanks on Open Rails

A bedbank buys or contracts hotel rooms at a net rate — a price with no commission attached, commonly 20–30% below the public rate — and resells them to tour operators, travel agents, airlines, corporate desks, and other platforms, keeping the markup. The hotel gets volume and reach it could not source itself; the bedbank gets inventory it can package with flights and transfers or place in markets where the hotel has no presence.

Contracts come in two shapes. A static contract fixes net rates by season and room type for a year, grants an allotment — a number of rooms held for the wholesaler each night — and sets release dates, after which unsold allotment reverts to the hotel (14 to 30 days before arrival is typical), with stop-sale rights for the hotel and on-request handling once the allotment is exhausted. A dynamic contract derives the net rate from the live public rate through channel-manager connectivity, so the wholesaler's price follows the hotel's yielding. Static contracts give the wholesaler a fixed cost to package against; dynamic contracts give the hotel control but move the packaging risk to the bedbank.

Everything in that paragraph is prose in a PDF, enforced by an account manager and an annual review. The bedbank's own resale contracts with its buyers are prose too, and the hotel never sees them.

Rate leakage is what happens when a net rate meant for a package surfaces, unbundled, on a consumer-facing site. A bedbank sells to a distributor, who sells to an affiliate, who lists the room on a metasearch or an OTA at the net rate plus a thin markup — below the hotel's own public rate and below the rate the hotel shows the major OTAs. The hotel's parity obligations to those OTAs are breached by a listing it did not create, its direct channel is undercut by its own inventory, and a guest who found the room at $150 arrives at a desk quoting $189.

The damage is threefold. Revenue: the hotel sold the room at net and then watched it displace a full-rate booking. Relationship: the OTA market manager, seeing the hotel undercut, applies pressure or ranking penalties the hotel cannot answer. Trust: the guest concludes the public rate is fiction. Hotels respond by cutting wholesalers, imposing package-only clauses, and auditing consumer sites for their own net rates — all after the fact, all by hand, all at a cost that scales with the number of intermediaries in the chain. The Dispute Resolution Working Group's taxonomy treats the resulting rate discrepancies as billing disputes; wholesale leakage is one of their quieter causes.

The root cause is structural, not moral. A net rate is a number with conditions attached — package only, these markets, this markup ceiling, this release schedule — and the number travels while the conditions stay in the contract. Every hand the room passes through sees the price and none of the terms.

It would be easy to read the above as a case against wholesale. It is not. Bedbanks serve three functions that direct distribution, including AI-agent distribution, does not replace.

• Packaging. A room that ships with a flight, a transfer, and a tour is a product the hotel cannot assemble; the tour operator's margin is earned on the bundle, and the net rate is the price of admission to it.
• Group and event demand. Festivals, conventions, sports, and incentive travel arrive as blocks. Somebody has to take the risk of holding 500 room-nights against demand that may not materialize; the wholesaler who pre-purchases absorbs that risk and is paid for it.
• Unreachable markets. A Memphis independent has no sales presence in Seoul, São Paulo, or Riyadh. A bedbank with buyers in those markets does, and the source-market agencies that sell there buy from the bedbank, not from the hotel.

Scale is the other reason. The working groups page notes that Hotelbeds alone contracts with over 300,000 properties; WebBeds, RateHawk, TBO, Bonotel, W2M, and MIKI Travel each aggregate inventory that no property could reach on its own. That aggregation is not going away, and ADAPT does not propose that it should. The proposal is that intermediation happen transparently, on open rails, with terms that machines can read.

Programmable settlement means that the conditions attached to a price are encoded with the transaction and enforced by protocol, not by trust in an intermediary. Applied to wholesale, four things change.

Machine-enforceable resale terms. The net rate contract becomes a structured object: permitted resale channels (package, group, B2B; consumer OTA yes or no), a maximum markup, geographic and source-market restrictions, and blackout dates. Every resale of the room carries the block's identifier and its terms. A consumer-facing listing that presents a net-derived rate without the packaging condition is not a rumor to investigate; it is a resale event that does not match its terms, visible to the property when it happens. Enforcement is whatever the terms say it is — settlement withheld, the block closed to that buyer, a penalty applied — with contested cases routed to a localized arbiter rather than to a lawyer's letter.

Encoded release schedules. Release dates stop being a calendar reminder. An allotment with an encoded release returns unsold rooms to the property automatically on the stated day, and the returned rooms are immediately available to every other channel, including AI agents. Partial releases, staged releases, and event-specific schedules are parameters, not negotiations.

Destination contributions per booking. The manifesto and the working groups describe Tourism Development Fund zones in which a destination's rule-set — local advisor commissions of 1–2%, arbiter fees, DMO contributions — attaches to each booking. A wholesale block can carry a per-room-night contribution to the DMO or fund, settled with the room rather than collected through a separate assessment. For a destination marketing organization this turns a wholesaler's event block from an invisible transaction into a funded one: similar in intent to the tourism improvement district assessments many US cities already levy, but settled at booking time.

Settlement on the property's terms. Today the bedbank pays the hotel on the bedbank's cycle, often weeks after checkout, through virtual cards that have to be reconciled by hand. With terms encoded, the property chooses: settlement on booking, on release, or on check-in; deposits and cancellation penalties execute themselves. The reconciliation work the home page describes for OTA statements disappears for the same reason — the terms that priced the room-night are attached to it.

There is a second reason bedbanks belong on open rails, and it is about demand rather than compliance. AI agents are becoming a discovery channel: the site's figures put roughly 140,000 hotels within reach of Perplexity and Selfbook today, and The March 2026 Inflection traced hotels going live inside ChatGPT and Google's agentic surfaces. Agents query structured inventory through protocols like the Model Context Protocol. They do not browse a bedbank's B2B portal.

A wholesale block that is encoded is also a wholesale block that is discoverable. When an agent assembling a festival weekend for a family asks for four nights across the drive market, an event block with rooms still inside its release window can answer that query alongside the properties' direct inventory — at the block's terms, with its packaging conditions intact, and with the resale recorded. The wholesaler gains a channel that did not exist; the property gains a sale it would otherwise have waited for the release date to recapture; the agent gains inventory it could not otherwise see. The working groups page states the alternative plainly: inventory that stays outside the protocol becomes invisible to the fastest-growing booking channel in the industry.

The working groups page sketches a local case, and it is worth working through as an illustrative scenario. CMA Fest runs each June in Nashville; Memphis sits three hours west, inside its drive market. A Memphis-based event wholesaler — a local operator with relationships across a dozen independents — pre-purchases 500 room-nights for festival week across 10 properties. Today that is ten contracts, ten release dates on a spreadsheet, ten settlements, and a resale channel the properties cannot see. On protocol it is one block with encoded terms.

Walk the terms. The wholesaler holds 500 room-nights at a $189 net rate and may resell them in packages and to groups, at up to 25% markup, but not on consumer OTAs; a resale that breaks that rule is detectable when it happens. Fourteen days before arrival, whatever is unsold auto-releases to the ten properties, which can then sell those nights direct or through agents at festival rates. Each room-night that settles carries a $2.00 contribution to the Memphis Tourism Development Fund zone. Settlement happens on release or on check-in, at each property's election. And the block is discoverable: an agent planning the weekend can see it beside the properties' own inventory. Illustration only, with round numbers: 500 room-nights at $189 is $94,500 of net revenue committed to ten independents before the festival, $1,000 to the destination fund, and no leakage audit afterward.

ADAPT's position toward bedbanks is the one it takes toward every intermediary that remains: intermediation is welcome; opacity is not. The working groups page offers bedbanks four things — enforceable resale terms, leakage prevention, event block management, and AI-agent distribution — and invites Hotelbeds, WebBeds, RateHawk, W2M, Bonotel, TBO, MIKI Travel, and local event wholesalers to the table. The register lists this article as a proposed joint session with wholesale industry stakeholders.

Three concrete proposals follow from this analysis. First, a block-contract object — the fields in the listing above — specified within the rate and inventory protocol work that MCP as the New Distribution Rail describes, so that a property's protocol-native PMS can publish a block and a bedbank's system can accept it without bespoke integration. Second, a resale-event record, so that every downstream sale of a block's room-night is attributable to a buyer and comparable with the terms. Third, a wholesale dispute category within the DR-WG framework, so that leakage and release disputes have an arbiter, a timeline, and an evidence standard.

The local event wholesaler is the place to start, because the incentives are already aligned: a Memphis operator buying regional rooms for a Tennessee event, contributing to a Memphis destination fund, settled through a protocol whose reference implementation runs on 200+ units in Downtown Memphis. If the model works at that scale — and the pilot exists to find out — the same object works for a bedbank with 300,000 properties. That is the sense in which bedbanks belong on open rails: not disintermediated, but legible.