title: “Flight Supply Strategy — Complete Findings & Decision Framework” category: Supply Strategy status: Draft created: 2026-02-25 related:
← Section Index · Master Index
Flight Supply Strategy — Complete Findings & Decision Framework
Executive Summary
This document consolidates all findings from our discussions regarding flight supply strategies:
- OTA Model — redirect-based, capital-light, fastest time-to-market
- NDC / Direct Airline Integration (DAI) — higher margins, higher complexity
- Hybrid Model — phased approach balancing speed, risk, and margin growth
It includes operational, technical, commercial, and strategic insights to support a board-level decision.
[!IMPORTANT] This is not a technical preference decision. It is a capital allocation, risk management, and long-term positioning decision.
1. How Meta-Search Platforms Work
Meta-search platforms perform five core functions:
- Aggregate supply from multiple providers (OTAs, airlines, aggregators)
- Normalize and deduplicate itineraries
- Rank and display comparable flight options
- Redirect users (or optionally handle checkout)
- Track click attribution for revenue
[!NOTE] Key realities:
- Not all airlines are queried directly — supply comes through intermediaries
- APIs are called in parallel with timeouts and caching
- Prices can change due to offer expiration or inventory changes — “price mismatch” is inherent to the model
2. OTA Model
What It Is

You integrate multiple Online Travel Agencies (OTAs). Users compare prices on your platform and are redirected to the OTA to book.
sequenceDiagram
actor User
participant Meta as Meta-Search Platform
participant OTA as OTA Supplier
User->>Meta: Search Flights
Meta->>OTA: Query Inventory (Parallel)
OTA-->>Meta: Return Offers
Meta->>User: Display Comparison
User->>Meta: Click Offer
Meta->>OTA: Redirect User to OTA Check-out
Note over OTA: OTA handles Payment, Ticketing & Support
User->>OTA: Complete Booking
Platform Responsibilities
| Responsibility | Description |
|---|---|
| Search UI | Display flight results with filters and sorting |
| Deduplication | Merge identical itineraries from multiple sources |
| Provider comparison | Show price differences across OTAs |
| Redirect tracking | Track click-throughs for CPC/CPA revenue |
Advantages
- ✅ Fast time-to-market — minimal infrastructure needed
- ✅ Low operational complexity — no payment or refund handling
- ✅ Capital-light — no working capital exposure or float risk
Disadvantages
- ❌ Lower margins — limited to CPC/CPA commission
- ❌ No booking ownership — users don’t have booking history in your system
- ❌ Partner dependency — reliant on OTA reliability and data accuracy
Refund & Booking History
- Managed entirely by the OTA
- Users do NOT have booking history in your system
- Support burden falls on the OTA, not you
3. NDC / Direct Airline Integration (DAI)
What It Is

Integration with airline retail APIs using the NDC (New Distribution Capability) standard — an IATA protocol that enables airlines to sell directly through third-party platforms.
sequenceDiagram
actor User
participant Meta as Meta-Search Platform
participant NDC as Airline NDC API
User->>Meta: Search Flights
Meta->>NDC: AirShoppingReq
NDC-->>Meta: AirShoppingRS (Offers)
Meta->>User: Display Results
User->>Meta: Select Offer
Meta->>NDC: OfferPriceReq (Confirm Price)
NDC-->>Meta: OfferPriceRS
User->>Meta: Checkout & Pay
Note over Meta: Meta handles PCI/Payment
Meta->>NDC: OrderCreateReq (Issue Ticket)
NDC-->>Meta: OrderViewRS (Ticket Issued)
Note over Meta: Meta handles Refunds & Support
NDC Lifecycle
flowchart LR
A[AirShopping] --> B[OfferPrice]
B --> C[OrderCreate]
C --> D[OrderChange]
C --> E[OrderCancel]
| Stage | Purpose |
|---|---|
| AirShopping | Browse available offers |
| OfferPrice | Confirm pricing for a specific offer |
| OrderCreate | Complete the booking and issue ticket |
| OrderChange | Modify dates, passengers, or ancillaries |
| OrderCancel | Cancel the booking and process refund |
Advantages
- ✅ Higher margin potential — markup + commission
- ✅ Richer ancillary content — seat selection, baggage, meals
- ✅ Airline differentiation — access to exclusive fares and bundles
- ✅ Brand ownership — you own the customer relationship
Disadvantages
- ❌ High technical complexity — Offer → Order lifecycle is non-trivial
- ❌ Offer expiration — offers have TTLs (time-to-live) and can expire mid-checkout
- ❌ Servicing burden — refund/change processing falls on you
- ❌ Payment compliance — PCI-DSS certification required
- ❌ Airline variability — each airline implements NDC differently
4. Why Multiple Integrations Are Needed
Even with NDC integration:
| Challenge | Explanation |
|---|---|
| Coverage fragmentation | NDC is per-airline — no single source covers all carriers |
| Airline variability | Each airline implements NDC differently (schema, flow, features) |
| Commercial differences | Commission rates, incentives, and terms vary per airline |
| Resilience | Multi-supplier architecture provides redundancy and failover |
[!NOTE] Even with an aggregator like Verteil or Duffel, multi-sourcing may be required for full corridor coverage and negotiation leverage.
5. Hybrid Model (Recommended for Most Startups)

Structure
| Phase | Activity | Focus |
|---|---|---|
| Phase 1 | Multiple OTAs | Early revenue, optimize unit economics |
| Phase 2 | Add selective NDC/DAI | High-volume routes, margin improvement |
| Phase 3 | Optional limited checkout | Only after compliance & ops readiness |
Benefits
- ✅ Early revenue with reduced risk
- ✅ Gradual margin improvement as NDC is added
- ✅ Controlled complexity growth
- ✅ Preserves future optionality
Decision Gate
Proceed to NDC phase ONLY if:
Incremental Margin > Incremental Operational Cost + Risk Premium
Cross-reference: See Aggregated Reference § Hybrid Model for the full phased roadmap with Gantt chart.
6. Supplier Landscape Overview
| Provider | Type | Complexity | Startup Fit | Best For |
|---|---|---|---|---|
| Verteil | NDC Aggregator | Medium | Moderate | GCC corridor airlines (Emirates, Etihad, Qatar) |
| Duffel | Modern Airline API | Medium | High | Developer-first builds, broad global coverage |
| Travelfusion | Direct Connect + LCC | High | Moderate | LCC-heavy markets |
| TPConnects | NDC Aggregator | Medium | Moderate | GCC-focused aggregation |
| Amadeus | GDS + NDC | Very High | Low | Enterprise-scale operations |
| Sabre | GDS + NDC | Very High | Low | Enterprise-scale operations |
| Travelport | GDS + NDC | Very High | Low | Enterprise-scale operations |
| Kiwi | OTA | Low–Medium | High | Fast OTA model, virtual interlining |
Cross-reference: See Duffel vs Verteil for a detailed head-to-head comparison.
7. Strategic Trade-Off Matrix
| Dimension | OTA | Hybrid | NDC Heavy |
|---|---|---|---|
| Time to Market | ✅ High | ⚡ Medium | ❌ Low |
| Margin Potential | ❌ Low | ⚡ Medium | ✅ High |
| Risk | ✅ Low | ⚡ Medium | ❌ High |
| Operational Burden | ✅ Low | ⚡ Medium | ❌ High |
| Capital Requirement | ✅ Low | ⚡ Medium | ❌ High |
quadrantChart
title Supply Model Risk vs Margin Profile
x-axis Low Risk --> High Risk
y-axis Low Margin --> High Margin
quadrant-1 High Risk & Reward (NDC Heavy)
quadrant-2 Balanced Growth (Hybrid)
quadrant-3 Fast & Safe (OTA Only)
quadrant-4 Low Value (Avoid)
"OTA Only": [0.15, 0.2]
"NDC Heavy": [0.85, 0.85]
"Hybrid": [0.45, 0.65]
8. Weighted Decision Framework
The final decision should be weighted against these factors:
| Factor | Implication |
|---|---|
| Time-to-market urgency | Higher urgency → favor OTA |
| Capital availability | Less capital → favor OTA; more → consider NDC |
| Risk tolerance | Risk-averse → OTA; risk-tolerant → NDC |
| Margin ambition | Higher margin targets → NDC |
| Engineering maturity | Junior team → OTA; experienced → Hybrid/NDC |
| Operational readiness | No ops team → OTA; ops team → Hybrid/NDC |
Quick Decision Guide
| If you prioritize… | → Choose |
|---|---|
| Speed & safety | OTA |
| Margin & ownership | NDC Heavy |
| Balanced growth | Hybrid ✨ |
9. Board-Level Framing
[!CAUTION] This is not a technical choice. It is:
- A capital allocation decision
- A risk management decision
- A long-term positioning decision
The board must define risk tolerance, capital envelope, and 3-year positioning goals before selecting a path.
10. Recommended Evolution Pattern
For most startups, the recommended path is:
flowchart LR
A["1. Launch with OTA\n(Months 0–6)"] --> B["2. Optimize Unit Economics\n(Months 3–9)"]
B --> C["3. Add Selective NDC\n(Months 6–12)"]
C --> D["4. Consider Checkout\n(Month 12+)"]
- Launch with multiple OTA integrations
- Optimize unit economics (CTR, conversion, supplier quality)
- Add selective NDC on high-volume routes where margin uplift justifies complexity
- Consider on-platform checkout only after compliance and ops readiness are confirmed
Final Conclusion
The Hybrid strategy provides the best balance between:
- ⚡ Speed
- 🛡️ Risk
- 💰 Margin
- 🔄 Optionality
However, the final decision must align with:
- Geography — which corridors are highest priority
- Funding — available capital and burn rate
- Ambition — margin targets and market positioning goals
- Operational capacity — team size and maturity
Quick Flashcards
What is the primary trade-off of the OTA Model vs. the NDC Model?
OTA model offers fast time-to-market with lower operational risk but yields smaller margins (CPC). NDC offers high margin and customer ownership but requires deep technical investment and handles full checkout/support complexity.
Why is a Hybrid Strategy recommended for most startups?
It allows a startup to generate early revenue rapidly through standard OTAs, validating the product, before selectively deploying capital-intensive NDC integrations on high-volume routes.
Prepared as a consolidated strategic reference document.