
By Forecepts Team
4 June 2026

By Forecepts Team
4 June 2026
A corporate booking does not begin and end at the moment a traveller clicks "confirm". From the first search to the final reconciled payment in the TMC's account, a single booking passes through roughly seven distinct stages — and at each one, something specific happens, something specific can go wrong, and the TMC faces a specific decision. Most discussions of travel technology describe these stages from thirty thousand feet: a booking "flows through the system". This article does the opposite. It walks the whole route at ground level, one stage at a time, showing the mechanism, the failure mode, and where the TMC adds or loses value at each step.
This is a map of the whole journey. Each stage links to a deeper article on that specific topic, so you can read the overview here and dive into any single stage when you need the detail. The throughline is simple: a booking is only as profitable as the weakest stage it passes through, because a failure at any stage costs money or trust downstream.

The journey starts with a search. What the traveller sees as a tidy list of flight and hotel options is the visible surface of a content layer underneath: inventory aggregated from one or more GDSes, NDC content from airlines, low-cost carrier content via an aggregator, and direct hotel APIs. The completeness of that content determines whether the traveller sees the genuinely best option — or only part of the picture. This is where the travel booking engine does its work, drawing on GDS and NDC content sources.
Where it breaks:
As the traveller selects an option, the system checks it against the corporate client's travel policy in real time: fare class, hotel rate cap, advance-purchase window, preferred suppliers. An in-policy booking is marked compliant; an out-of-policy booking is flagged before it proceeds. The critical word is "before" — policy enforcement only works at the point of booking, not after the trip. How this enforcement actually works, and why post-trip flagging is reporting rather than enforcement, is covered in Why Travel Policy Compliance Fails And How to Fix It.
Where it breaks:
A booking that needs sign-off routes to an approver. Mature programmes use a hybrid model: in-policy bookings are passively approved and flow straight through, while out-of-policy bookings are actively routed to a human. Multi-level chains, delegation, and clear approval status all live here. The depth of this stage — and why approval design is really a corporate booking tool decision — is covered in Corporate Booking Tool: A TMC's Guide for 2026.
Where it breaks:
Once approved, the booking is ticketed: the reservation becomes a confirmed ticket against a PNR (for GDS bookings) or an Order Reference (for NDC bookings). Every fare carries a ticketing deadline, and auto-ticketing controls decide when the system issues directly versus when an agent must intervene. This is also where the GDS-versus-NDC distinction starts to matter operationally, because NDC tickets are serviced differently — see NDC Fares: An Operational Decision Framework for TMCs.
Where it breaks:
With the ticket issued, the booking becomes a financial transaction. The PNR or Order Reference is converted into billable items — fare, taxes, fees, service charges, markup — and an invoice is generated. This is the first stage handled by the mid-back office rather than the booking front-end, and the handoff between them is one of the most common points of data loss. The mechanics of this conversion are covered in What Is a Mid-Back Office System for Travel Agencies?.
Where it breaks:
Every ticket issued has to be reconciled against BSP or ARC sales reports, and refunds, voids, and ADMs (Agency Debit Memos) tracked against the original transaction. Multi-GDS agencies multiply this work because each GDS settles separately. This is where a weak back office costs real money, and the reconciliation mechanics — including how mixed GDS and NDC bookings complicate the picture — are detailed in What Is a Mid-Back Office System for Travel Agencies?.
Where it breaks:
Finally, structured records flow into the agency's accounting system and management dashboards: sales by route, by client, by supplier, and profitability per booking. This is what turns a year of transactions into commercial intelligence — and it is only possible if the previous six stages produced clean structured data. The reporting and accounting-integration layer is covered in the mid-back office article, and the link between captured data and realised savings is explored in Business Travel Savings: How TMCs Help Corporate Clients Cut Costs.
Where it breaks:
Read the seven stages together and a pattern emerges: the risk is rarely within a stage — it is at the handoffs between them. A booking engine that does not pass clean data to the approval layer; an approval that does not feed the ticketing queue; a ticket that does not flow into invoicing; an invoice that does not reconcile automatically. Each handoff that requires a manual export and re-import is a point where data is lost, delayed, or corrupted. The cumulative cost of these disconnects — not the cost of any single stage — is what separates a TMC that scales profitably from one whose margins erode as volume grows.
This is the core argument of Digital Transformation in the Travel Industry: What It Means for TMCs.
A booking is only as profitable as the weakest stage it passes through. A perfect search means nothing if approval stalls past the ticketing deadline. Flawless ticketing means nothing if reconciliation is too slow to dispute an ADM. The stages are a chain, and the chain is only as strong as its handoffs.
Each stage carries an Asia-Pacific complication that systems built for single-market programmes tend to miss:
• Search — regional LCCs (AirAsia, Scoot, Cebu Pacific, VietJet) carry a large share of point-to-point travel and need aggregator integration to appear at all.
• Policy and approval — multi-timezone approval chains shrink the effective window before a ticketing deadline; an approver in another country may not see a request until the deadline is near.
• Ticketing — dual GDS setups (common in the region) mean tickets issue across more than one settlement system.
• Invoicing — country-specific tax and e-invoicing rules (Singapore GST, Malaysia SST and e-invoicing, Indonesian VAT, Indian GST) each require correct invoice formatting.
• Settlement — multiple IATA BSP countries, multiple currencies, and local payment methods each settle and refund differently.
• Reporting — multi-branch, multi-currency operations need clean separation in reporting without losing the consolidated view.
Forecepts builds the stages as connected modules: SWIFT IBE handles search and content (stages 1 and the front end of 4); SWIFT CBT handles policy check, approval, and corporate booking (stages 2-3); and SWIFT Mid-Back Office handles invoicing, settlement, reconciliation, and reporting (stages 5-7). Because the modules share data, the handoffs that break disconnected stacks are automated. Other approaches exist; the point is that the stages must connect, however a TMC assembles them.

Forecepts builds travel technology for TMCs and travel agencies across Asia-Pacific — IBE, Corporate Booking Tool, and Mid-Back Office as connected modules that pass data cleanly from search through to settlement. Talk to the team to see how the stages connect for your operation.
Frequently Asked Questions
At a practical level, seven: search and content, policy check, approval, ticketing, invoicing, settlement and reconciliation, and reporting. Some programmes collapse or combine stages — passive approval, for example, makes the approval stage invisible for in-policy bookings — but every booking touches all seven functions even when some are automated to the point of being unnoticeable.
The handoffs between stages, more than any single stage. The most common operational failures are a booking engine that does not pass clean data to the back office, and an approval step that stalls a booking past its ticketing deadline. Both are failures of connection rather than failures of any one system in isolation.
Not necessarily a single product, but the stages do need to be connected so data flows without manual re-entry. Some TMCs run a connected suite; others integrate best-of-breed tools. What matters is that the handoffs are automated — a collection of excellent but disconnected tools usually performs worse than a well-connected integrated stack, because the cost lives in the gaps between tools.