Preventing ADMs: how booking platforms stop agency debit memos before they happen

An Agency Debit Memo (ADM) is a charge an airline issues to a travel agency when it decides a booking or ticket broke its rules. In BSP markets, ADMs arrive through the airline’s settlement channel and are deducted from the agency’s remittance unless they are disputed in time. For a high-volume agency or B2B platform, ADMs are not an occasional nuisance — they are a running cost, and most of them trace back to something software could have caught.

Why ADMs exist

Airlines audit agency activity after the fact. Their revenue-integrity teams and tools look for patterns that cost them money or inventory: seats held and never ticketed, the same passenger booked several times, fares sold without honouring their rules, refunds that returned more than the ticket allowed. When they find one, they bill the agency through the BSP or the equivalent settlement process.

The ADM itself is usually small compared with the ticket, but it carries an administrative fee, it lands weeks after the original booking, and someone has to investigate every one. Each airline publishes its own ADM policy, so the rules an agency must follow differ by carrier.

The common causes

Most ADMs fall into a handful of groups:

• Booking violations — duplicate bookings for the same passenger, repeated cancel-and-rebook cycles to hold inventory (often called churning), fictitious or test names, and passive or inactive segments left in a PNR.
• Unticketed and no-show segments — seats held past the ticketing deadline or not cancelled before departure.
• Fare misapplication — a fare sold outside its rules: wrong booking class, minimum stay not met, advance-purchase conditions ignored, or an incorrect point of sale.
• Refund and exchange errors — penalties not deducted, non-refundable taxes refunded, or exchanges priced incorrectly.
• Commission and tax errors — wrong commission rates or tax codes on the ticket.

Almost every item on that list is a rule that can be expressed in code.

Prevention belongs in the platform

Manual training and checklists help, but in an automated booking platform the volume is too high and the bookings are created by software anyway. The controls belong in the booking flow and in background jobs.

Detect duplicates before booking

Before creating a reservation, check for existing active bookings with the same passenger name, date of birth (when available), carrier and travel dates. A close match should either block the booking or require an explicit override with a reason. For B2B platforms where many agents share an account, run the check across the whole account, not just the current user.

Housekeep unticketed bookings

Every reservation that has not been ticketed needs an owner and a deadline. A scheduled job should find bookings approaching their ticketing time limit, warn the customer or agent, and cancel the segments if nothing happens. This is covered in more detail in ticketing time limits and automated PNR housekeeping.

Never leave inactive segments behind

Schedule changes and failed bookings often leave segments in states such as unconfirmed, waitlisted or cancelled by the airline. These should be removed from the PNR promptly. A sync job that retrieves active PNRs and flags non-confirmed status codes is one of the cheapest ADM controls an agency can build.

Validate the fare you sold

Store the offer exactly as priced — fare basis, booking class, fare rules, taxes and the pricing response — with the order. At ticketing time, re-price and compare. If the stored fare and the current pricing disagree on anything that matters, stop and route the case to a queue rather than issuing a ticket that will be audited later.

Calculate refunds from the rules, not by hand

Refunds and exchanges are a common source of ADMs because the calculation involves penalties, used segments and tax refundability. Use the GDS or airline refund and exchange quotations where they exist, keep the quote with the order, and require a second check for any manual override.

Block test and placeholder data in production

Test bookings created in a live environment are a classic ADM trigger. Enforce environment separation for supplier credentials, and validate passenger names against obvious placeholders before any live booking call.

Make every booking explainable

When an ADM arrives, the agency has a limited window to accept or dispute it. Disputes succeed only with evidence. A booking platform should be able to show, for any ticket:

• the offer and fare rules at the time of sale,
• every change to the booking, with timestamps and who made it,
• the supplier requests and responses involved, with personal data masked,
• the refund or exchange quotation used.

This is the same audit trail that makes idempotent booking flows debuggable, and it pays for itself the first time a disputed ADM is reversed.

Treat ADMs as a feedback loop

Import ADMs into the platform and link each one to the order it concerns. Then categorise them by cause, airline and channel. The pattern usually points to a specific gap: a sub-agent repeatedly creating duplicates, a fare type the pricing logic mishandles, or a refund path that bypasses validation. Fix the cause in code, and the same category of ADM should stop appearing.

For B2B and consolidator platforms, the same data decides who pays. If a downstream agent caused the ADM, the platform needs the booking history to pass the charge on fairly and the reporting to show agents their own ADM exposure.

Summary

ADMs are what airlines charge for booking and ticketing behaviour that breaks their rules. The causes are well known and mostly mechanical: duplicates, unticketed and inactive segments, fares sold outside their rules and miscalculated refunds. A booking platform that checks for duplicates, housekeeps PNRs, validates fares at ticketing, calculates refunds from the rules and keeps a complete audit trail will prevent most ADMs and win more of the disputes that remain.