NDC handled roughly 24% of indirect airline ticket sales in Q1 2026, up from 11% in 2023, and IATA projects around 65% of indirect bookings will be NDC-powered by 2026. At the same time, OTA share of NDC transactions has actually dropped from 85% to 77% year over year as traditional leisure agencies invest more directly in NDC connectivity themselves. Travel distribution is shifting fast, and that makes the off-the-shelf-vs-custom decision for your booking engine more consequential than it used to be. This guide breaks down the real costs and trade-offs.
Before you start: know your position
- [ ] Product type — standard flight/hotel booking, multi-day tour packaging, or a mix
- [ ] Number and type of suppliers you need to connect to (GDS, bedbanks, direct airline/hotel APIs)
- [ ] How differentiated your UX or personalization needs to be as a competitive moat
- [ ] Internal engineering capacity to own supplier integrations and their ongoing changes
- [ ] Time-to-market pressure — are you testing a new market/segment or scaling a proven one?
The cost picture
Off-the-shelf / white-label booking engine
- [ ] SaaS tiers starting around $99/month, up to one-time licensing models
- [ ] GDS/bedbank connectivity often bundled as a one-time fee rather than ongoing cost in some vendor models
- [ ] Time to launch: typically 2-8 weeks
Custom-built travel platform
- [ ] Flight + hotel MVP with one GDS and one bedbank: $25K-55K
- [ ] Full multi-source B2B/B2C portal with white-label, markup engine, and ledger/accounting: $60K-150K+
- [ ] Enterprise-grade platform (Amadeus/Sabre/Travelport plus automated ticketing, refunds, reissue, CRM, analytics): $190K-300K+
- [ ] Per-supplier GDS/NDC integration: $5K-20K each standalone; multi-supplier GDS API integration work alone can run $35K-90K+
- [ ] NDC integration cost varies significantly per airline — there's no standard API, so each carrier's implementation is a separate engineering effort
When off-the-shelf wins
- [ ] Standard flight and hotel booking flows without unusual packaging logic
- [ ] Fast time-to-market matters more than deep customization
- [ ] Limited internal development team
- [ ] Testing a new market or segment before committing significant capital
When custom is justified
- [ ] Proprietary itinerary or packaging logic — multi-day tours, dynamic bundling that off-the-shelf engines can't represent
- [ ] Aggregating many disparate suppliers under custom business rules
- [ ] Differentiated UX as a core part of your competitive positioning
- [ ] AI-driven personalization or pricing that needs deep control over your own data, not a vendor's shared roadmap
- [ ] Hitting a ceiling on customization or growth inside a white-label vendor's constraints
Hidden costs and integration challenges to pressure-test
- [ ] Payment gateway costs go well beyond the headline rate — factor in setup fees, per-transaction fees, currency-conversion fees, refund/chargeback fees, and cross-border processing fees, all of which are commonly underestimated in initial scoping
- [ ] Multi-currency support and fraud rules add ongoing engineering overhead beyond the initial integration, not a one-time cost
- [ ] Real-time inventory sync across GDS and bedbanks (Hotelbeds, RateHawk, direct suppliers) requires continuous reconciliation logic — this is an ongoing operational commitment, not a build-once feature
- [ ] Post-integration costs that are frequently missed in initial budgets: markup engine, cancellation/refund/reissue workflows, ledger/accounting sync, admin panel, and ongoing support
- [ ] As of early 2025, only 13% of online booking tools supported self-service changes for NDC bookings — confirm whether your vendor or your custom build genuinely handles NDC servicing, not just NDC search and book
Red flags to check before you commit either way
- [ ] A white-label vendor who can't explain their NDC roadmap per airline, given how uneven carrier-by-carrier implementation still is
- [ ] A custom build estimate that treats GDS/NDC integration as a one-time line item rather than an ongoing maintenance commitment as carrier APIs evolve
- [ ] No clear plan for real-time inventory reconciliation across multiple suppliers
- [ ] Payment integration scoping that only accounts for the base transaction fee, not currency conversion, chargebacks, and cross-border processing
How to use this guide
Map your product complexity and supplier count against the buy/build scenarios above, then price out the full integration picture — not just the headline licensing or development cost. If NDC servicing capability or proprietary packaging logic is central to your business model, treat that as a strong signal toward custom or a hybrid build.