What is a bedbank? How do agencies work with bedbanks?
A bedbank is an intermediary that aggregates the rooms of many hotels through contracts or connections and sells them at wholesale prices to agencies, tour operators and other travel companies. Agencies usually connect to bedbanks through an XML or JSON API; they receive the hotel list as static data and query prices and availability live on every search.
- Author
- hotelmapping.org Editorial Team
- Last updated:
- 8 min read
Bedbank in brief
A bedbank is an intermediary that sells hotel rooms at wholesale. The name comes from the words “bed” and “bank”: it pools the room inventory of many hotels in a “bank of beds” and sells it to agencies, tour operators, online travel sites, corporate travel companies and other intermediaries. In the industry the term “wholesaler” is also used in the same sense.
The bedbank’s core value is scale. It is not practical for a single agency to sign individual contracts with thousands of hotels worldwide. A bedbank sets up these contracts and connections once, and the agency reaches this inventory through a single integration.
Companies such as Hotelbeds, WebBeds, RateHawk, TBO.com and Paximum are named in this group. These suppliers are on hotelmapping.org’s list of ready-made connections; how each one’s data is mapped is explained on its own integration page.
What sets a bedbank apart from other players?
Hotel distribution has several players that look alike but do different jobs. The table below offers a general framework; in real life a company can take on more than one role.
| Player | What does it sell? | Who does it sell to? | Relationship with the hotel |
|---|---|---|---|
| Bedbank / wholesaler | Mainly hotel rooms | Agencies, tour operators, OTAs, corporate travel companies (TMCs) | Direct contract or connection via other suppliers |
| Tour operator | Packages of hotel, flight and transfer | End customers and agencies | Usually a direct contract, often with an allotment |
| OTA (online travel agency) | Hotels and other travel products | End customers | Via a bedbank, channel manager or direct contract |
| Channel manager | Does not sell; it is distribution software | Hotels | Distributes the hotel’s rates and availability to many channels |
What sets the bedbank apart is that it is an intermediary on both the supplier and the customer side. For an agency, a bedbank is a supplier; for a hotel, it is a sales channel.
How does a bedbank work?
A bedbank’s inventory comes in through two main routes:
- Direct contract: The bedbank signs an agreement with the hotel that sets prices and terms for a season. The agreement covers room types, board types, child policy, cancellation terms and sometimes a set number of reserved rooms (allotment).
- Indirect connection: The bedbank pulls inventory from another wholesaler’s, a channel manager’s or a hotel chain’s system and offers it to its own customers.
A few concepts come up often in contracted inventory:
- Allotment: The number of rooms the hotel reserves for the bedbank for specific dates.
- Release period: The number of days before arrival at which unsold reserved rooms go back to the hotel.
- Stop sale: The hotel halting sales for specific dates. The bedbank passes this information to agencies; if it is not passed on, an agency may sell a room that can no longer be sold.
Bedbanks usually give agencies a net rate. The agency adds its own margin on top of this rate and sells to the end customer. The payment model varies by bedbank and agency: methods such as credit limits, prepayment or virtual cards are used. These terms are set separately in each commercial agreement.
How does an agency connect to a bedbank?
An agency can work with a bedbank in two ways: booking manually through the bedbank’s web interface, or connecting to it through an API in its own system. For multi-supplier agencies, the second is what really matters.
An API integration usually goes through these steps:
- Commercial agreement: Price terms, the payment model and the markets to be accessed are determined.
- Test environment: The bedbank gives the agency test access. The agency develops search, booking and cancellation flows here.
- Certification: Many bedbanks check the agency’s flow before go-live; the correct request sequence, cancellation rules and error handling are reviewed.
- Live access: The agency starts selling with real prices and availability.
Static and dynamic data
A bedbank integration consists of two different data flows. This distinction is also the subject of the article what is hotel XML integration.
- Dynamic data: Price, availability, cancellation terms. Queried live on every search and can change within minutes.
- Static data: Hotel name, address, coordinates, hotel description, images, room list, facilities and destination codes. It changes rarely; it is usually received as a bulk file or through a separate content API.
A typical live flow looks like this: The user picks dates and a destination → the agency sends an availability query to the bedbank → the bedbank returns prices → the user selects a room → the agency re-validates the price → the booking is created → it is canceled if needed. The details of names and steps vary from bedbank to bedbank.
Static data: the invisible burden of integration
Although the live flow looks technically more complex, in the long run it is often static data that demands ongoing effort. That is because each bedbank sends hotels with its own supplier code, its own spelling of the name and its own destination structure.
For the same hotel, records from three bedbanks might look like this:
| Field | Bedbank A | Bedbank B | Bedbank C |
|---|---|---|---|
| Hotel code | KMR-20418 |
771302 |
TR-AYT-KP09 |
| Hotel name | Kemer Pine Garden | KEMER PINE GARDEN HOTEL | Pine Garden Kemer - All Inclusive |
| Destination | Kemer | Antalya | Antalya / Kemer / Göynük |
| Board | AI | All Inclusive | Her Şey Dahil |
| Room | DBL SV | Double Room Sea View | Standart Oda Deniz Manzaralı |
These three records describe the same hotel, but no field is identical. If the agency does not match them, the same hotel appears three times in the list, price comparison breaks and the user cannot tell which card is the right one. We cover the impact of this on conversion and cancellations in the duplicate hotel cost article.
The problem is not limited to the hotel level. Room names (what is room mapping) and board types (board types guide) are also written differently from one bedbank to another. When a user selects the “All Inclusive” filter, an offer that one bedbank sends as “AI” and one that another sends as “All Inclusive” must fall under the same filter.
Multiple bedbanks: one hotel, one ID
What an agency working with several bedbanks needs is a single ID for each hotel. Supplier codes are linked to this ID; the hotel is shown to the user once, while the price query goes to each bedbank with its own code.
Hotel mapping does this work. After normalization and scoring, the three records in the example above are gathered under a single HM ID. Name similarity, location, phone and web address combine to produce a match score; if the score is 0.97 or above, the match is accepted automatically, records between 0.80 and 0.97 go to an expert for review, and below 0.80 a new HM ID is opened.
The response below belongs to a GET /v1/hotels/HM-1061590 request in API v1; the hotel name and codes are fictional:
{
"hm_id": "HM-1061590",
"name": "Kemer Pine Garden",
"destination_id": "TR-07-KEMER-GOYNUK",
"destination": ["TR", "Antalya", "Kemer", "Göynük"],
"geo": { "lat": 36.6512, "lng": 30.5604 },
"supplier_codes": {
"bedbank_a": "KMR-20418",
"bedbank_b": "771302",
"bedbank_c": "TR-AYT-KP09"
},
"external_ids": { "client": "OTL-2231" },
"status": "active"
}
For rooms, room mapping, and for board types, board type mapping work on the same logic: “DBL SV”, “Double Room Sea View” and “Standart Oda Deniz Manzaralı” are linked to the same room group; “AI”, “All Inclusive” and “Her Şey Dahil” are linked to the same standard board code.
Many agencies already have a hotel ID: their own internal codes or the IDs of a mapping service they used before. We store the hotel IDs of your existing mapping provider or your own system as cross-references; in the response above, OTL-2231 in the external_ids field is the agency’s own hotel code. That way, adding a new bedbank does not reset the existing mapping.
What to ask when choosing and connecting to a bedbank
Before starting to work with a new bedbank, asking about the data side as well as the commercial terms will make your job easier:
- How is static data provided? A bulk file or an API? How often is it updated? Does a full file or only changes (delta dump) arrive?
- Are hotel codes permanent? What happens to the code when a hotel closes, changes its name or merges with another hotel?
- What shared identifiers exist? Does the static data include a cross-reference field (e.g. a chain code) that makes it easier to match the record with other systems?
- How do room and board names arrive? Free text or a coded list?
- How is the destination structure? City only, or at district and resort-area level?
- How reliable are the coordinates? Is there a source and a verification process for the coordinates?
The answers to these questions largely determine how much mapping work will come up after the integration. For the bedbanks’ own data needs, see the bedbank solutions page.
Summary
A bedbank is an intermediary that sells hotel rooms at wholesale and gives agencies a broad inventory through a single integration. The agency connects to the bedbank via API; it queries prices and availability live and receives the hotel list as static data. Working with a single bedbank is relatively simple. When a second and third bedbank are added, the real work begins in static data: recognizing the same hotel, the same room and the same board type across different spellings.
You can find the whole picture of this work in the what is hotel mapping guide. If you would like to map your own bedbank files, you can request a tailored quote.
Frequently asked questions
What is the difference between a bedbank and a tour operator?
A tour operator packages services such as hotels, flights and transfers and sells them to end customers or agencies. A bedbank mainly sells hotel rooms only, at wholesale, and its customers are usually travel companies. Some companies play both roles.
Can a small agency work with a bedbank?
Generally yes. Many bedbanks let agencies book through a web interface; for an API connection, technical infrastructure and in some cases a minimum volume or a certification process may be required. Terms differ at each bedbank.
Why does the same hotel appear several times in the list when working with multiple bedbanks?
Each bedbank sends the hotel with its own code, its own spelling of the name and its own coordinates. If these records are not matched, the system takes them for different hotels. Hotel mapping solves the problem by merging these records under a single ID.
Why are board and room names from bedbanks not standardized?
Each bedbank enters hotel contracts into its own system with its own rules. The same board type can arrive as "AI" in one place and as "All Inclusive" or "Ultra Her Şey Dahil" in another. That is why room and board type mapping is done separately.
hotelmapping.org Editorial Team
DIJI.TECH hotel data team
The DIJI.TECH team working on hotel mapping, supplier integration and travel software.
Related products
Hotel mapping solutions
A persistent HM ID for every property; supplier codes and cross-IDs in a single record.
Room mapping solutions
Breaks room names into type, class, view, bed and capacity; groups the same room under one room ID.
- Only with us
Board mapping solutions
Converts spellings like AI, UAI, Soft AI and HB+ into 10 standard board codes. AI and UAI offers stay apart.
Related articles
- What is hotel mapping? How it works: a complete guide What is hotel mapping? How supplier records merge into one ID, the 9-step process, score thresholds, room and board mapping, and how to measure quality.
- What is hotel XML integration? Static vs. dynamic data What is hotel XML integration and how does it work? Static vs. dynamic hotel data, the typical search-to-booking flow, and sample XML/JSON records.
- What duplicate hotel listings cost an agency: conversion, cancellations and complaints What do duplicate hotel listings cost an agency in conversion, cancellations and support? Calculation framework, variables and a hypothetical example.
- Accuracy or coverage in mapping? What 99.99% really means How accuracy, coverage and human review rates affect each other in hotel mapping, and the questions to ask when a provider claims 99.99% accuracy.