Continuous Pricing: Beyond Filed Ladders

Contents

Executive Summary & Plain Definitions
Legacy vs Dynamic vs Continuous: Comparison Table
Why Move Beyond Filed Ladders
ATPCO Context & Constructs
Reference Architecture Layers
Conceptual Diagrams
Maturity Ladder & Transition Phases
Transition Roadmap (12–36 Months)
Operating in the Hybrid (Interim) World
Path to an RBD‑less / Attribute World
Revenue Accounting & Settlement Evolution
Interline & Alliance Implications
Example System Capability Categories
Data, Governance & Controls
Key KPIs & Observability
Risks & Mitigations
Plain-Language Glossary
Implementation Checklist
References & Disclaimers

1. Executive Summary & Plain Definitions

This article demystifies three pricing regimes as airlines modernize toward Offer & Order (IATA NDC + ONE Order) while integrating ATPCO fare filing infrastructure and emerging continuous pricing capabilities.

Plain Definitions (Simplified)

Legacy Filed Fare Ladder: Discrete price points (e.g. Y, B, M, H, Q, K, L, T) each with a published base fare, rules, and booking code (RBD). Systems choose the lowest open booking class whose rules fit.
Dynamic Pricing (Discrete Adjustments): Injecting controlled adjustments (discounts / mark-ups / virtual classes) around existing filed fares. Still anchored to the nearest filed base; produces a “denser” but finite ladder.
Continuous Pricing: Generating (or selecting) any price within an allowed continuum inside revenue guardrails, not limited to the original ladder steps. Return of price is contextual (channel, customer attributes, demand state) and tokenized as an Offer.
RBD-less (Attribute-Based) Retailing: Seat and product availability expressed through attributes & capacity tokens rather than the single-letter booking class driving price.

Core Shift: From selling pre-published “buckets” to calculating / composing contextual offers while preserving auditability for accounting, settlement, tax, and interline.

2. Legacy vs Dynamic vs Continuous: Comparison Table

Dimension	Legacy Filed Ladder	Dynamic (Discrete)	Continuous
Price Points	Dozens per O&D/date	Hundreds (filed + virtual/adjusted)	Theoretically many (bounded continuum)
Anchor	Filed base fare in ATPCO	Filed fare + rule-based adjustment	Model output (bid price / willingness curve)
Mechanism	Class availability + pricing engine	Dynamic class creation or discount codes	Offer generation service with guardrails
Control Levers	RBD protection levels	RBD + adjustment rules	Bid price / elasticity / product feature bundling
Interline Friendliness	High (common constructs)	Moderate (needs mapping)	Low–Moderate (requires token translation)
Channel Consistency Risk	Low	Medium	Higher (needs strict governance)
Revenue Optimization Potential	Baseline	Incremental uplift	Highest (granularity + personalization)
Complexity & Cost	Legacy sunk cost	Moderate	High (tech + process + governance)
Audit & Settlement Ease	Established	Needs augmentation	Requires Offer → Accounting mapping layer

3. Why Move Beyond Filed Ladders

Reduce distance between optimal bid price and nearest filed fare (less revenue left “between rungs”).
Enable customer-centric product & upsell packaging beyond single-letter classes.
Support personalized ancillaries (seat, bag, lounge) in one price expression.
Compete with LCC algorithmic repricing and OTA merchandising experimentation.
Prepare for attribute-based distribution (Offer & Order) where RBD is not the primary retail concept.

4. ATPCO Context & Constructs (Publicly Known Functions)

ATPCO provides the industry infrastructure for fare filing (base fares, rules categories, footnotes, routing, surcharges, taxes, branded fares, optional services, rich content). Dynamic and continuous pricing must respect or translate:

Base Fares & Rule Categories: Still define legal combinability, advance purchase, change/refund penalties.
Branded Fares (Product Catalog): Offer structure (Brand Y, Brand Flex, etc.). Dynamic pricing may adjust within brand value boundaries.
Optional Services / SSR/EMD Mapping: For ancillaries, continuous pricing can incorporate or discount bundles; still needs correct EMD and tax application.
Data Distribution (Fares, RBD Mappings): Downstream pricing engines ingest ATPCO feeds; continuous layer must track versions to retain audit trails.
Governance: Even in continuous pricing, compliance (e.g., display parity rules, tax calculation) relies on consistent reference to filed constructs.

Validation Step: Engage ATPCO solution / product experts to confirm permissible adjustment ranges and ensure no misuse of constructs (placeholder: “Confirm with ATPCO”).

5. Reference Architecture Layers

Demand Intelligence: Forecasts, elasticity curves, bid prices.
Control Layer: Legacy RM (nesting, O&D optimization) + dynamic price optimizer.
Offer Construction: Combines seat/brand/ancillary attributes; chooses or computes price.
Regulatory & Policy Guardrails: Taxes, surcharges, minimum / maximum price, parity constraints.
Offer Cache / Distribution: NDC API, direct channels, (optionally) controlled caching.
Order Management: Pure order record or hybrid ticket+order mapping.
Settlement & Accounting Adapter: Maps Offer Items to legacy coupon / EMD or order-based clearing.
Analytics & Governance: Price realization vs model; leakage & audit logs.

6. Conceptual Diagrams

Price Granularity Evolution ------------------------------------------------- Legacy Ladder: |Y| |B| |M| |H| |Q| Price Axis ($) -----^----^-----^------^-----^--- Lost Revenue Zones = gaps between filed points Dynamic (Discrete): |Y| |Y1| |B| |B1| |M| |M1| Smaller gaps, still step function Continuous: Smooth curve aligned to demand model ^ chosen contextual price (tokenized)

Hybrid Pricing Architecture (Simplified Flow) ----------------------------------------------------------------- Search Request -> Offer Engine -> Demand Model (real-time bid price) -> Policy Guardrails (min/max, parity) -> ATPCO Reference (rules, brand attributes) -> Dynamic Adjuster (±% around anchor) -> Price Continuum Resolver (select final) <- Offer (price + components + provenance token) Commit -> Validate token (versions, guardrails) -> Create Order (or Order + Ticket mapping) -> Revenue Accounting Adapter (coupon or virtual)

Maturity Ladder ----------------------------------------------------------------- Level 0: Static filed ladder only Level 1: Virtual buckets (pre-filed micro-fares) Level 2: Rule-based discounts/mark-ups (dynamic discrete) Level 3: Demand-model guided selection inside pre-defined band Level 4: Continuous price generation + real-time bundling Level 5: Attribute/RBD-less offers + order-native settlement

7. Maturity Ladder & Transition Phases

Each maturity level adds controlled complexity. Avoid skipping levels unless you have high data quality & governance capacity.

L0 → L1: Add sub-ladders (virtual buckets) to reduce price gaps without altering core accounting.
L1 → L2: Introduce real-time conditional adjustments (load factor, booking window) with strict audit logs.
L2 → L3: Incorporate bid price / elasticity signals to choose best available discrete point inside an allowed range.
L3 → L4: Enable interpolation/extrapolation to generate intermediate prices; introduce Offer tokenization referencing guardrail set.
L4 → L5: De-couple price from RBD; product attributes drive bundles; settlement moves to Order-centric with virtual coupon layer.

8. Transition Roadmap (Indicative 12–36 Months)

Months 0–3 Discovery: Inventory current fare constructs, identify average fare gap to bid price, data quality assessment (no-op changes for baseline).
Months 4–6 Foundation: Implement provenance tagging (controlVersion, pricingStrategyId). Add virtual buckets for top 20 O&Ds.
Months 7–12 Dynamic Discrete Rollout: Launch rule-governed adjustments (±X%) with monitoring of realized uplift vs control groups.
Months 13–18 Model Integration: Integrate bid price service; enforce guardrail corridor (min filed, max willingness). Introduce Offer tokens in NDC responses.
Months 19–24 Continuous Engine Pilot: Interpolate intermediate price; restrict distribution to selected channels (web, mobile). AB test churn effects.
Months 25–30 Attribute Expansion: Begin RBD decoupling: product features enumerated; create mapping service (RBD ↔ attribute set).
Months 31–36 Order-Native Settlement Pilot: Generate virtual coupons for accounting; partial RBD suppression in direct channel; refine interline fallback strategy.

Tip: Maintain a rollback lever at each phase (feature flag toggling dynamic layer off while preserving baseline filed fare availability).

9. Operating in the Hybrid (Interim) World

During transition you will run dual logic:

Dual Price Expressions: Offer includes (a) computed price, (b) nearest filed anchor + variance.
Fallback Path: If commit validation fails (stale model), system reverts to nearest permitted discrete price or re-quotes with user transparency.
Channel Segmentation: Start continuous pricing only on direct + NDC partners able to consume tokens; keep EDIFACT/GDS on improved discrete ladder.
Monitoring: Divergence windows between model recommended and realized sale tracked as lost optimization metric.

10. Path to an RBD‑less / Attribute World

Attribute Catalog Build: Define features (refundability, change flexibility, seat pitch, bag allowance, priority services).
Mapping Layer: RBD → AttributeSetId for legacy compatibility. Persist version.
Offer Construction API: Accept requested attributes; compute price independent of discrete class letter.
Capacity Tokens: Replace RBD open/close with attribute capacity pool (e.g., Flex Seats remaining).
Progressive De-emphasis: Hide RBD from consumer UX; maintain internally until interline fully supports attribute semantics.
Decommission RBD for Direct: Once auditing & settlement operate on attributes + Order Items reliably, remove RBD gating for direct channels first.

11. Revenue Accounting & Settlement Evolution

Today: Ticket coupons + EMDs feed revenue accounting and proration (interline). Challenge: Continuous prices may not map 1:1 to pre-filed base+adjust taxes if expressed as net + markup.

Interim Approach

Generate a Virtual Fare Basis referencing anchor filed fare + adjustment code.
Store Price Decomposition: base anchor, adjustment amount, ancillary components, tax tables version.
Issue standard e-ticket / EMD with auxiliary fields referencing OfferId for traceability.

Order-Native (Target)

Order Item lines itemize each priced component (seat bundle, ancillary, dynamic adjustment).
Revenue recognition uses Order ledger (no coupon) with standardized attribute taxonomy.
Settlement (BSP/ARC) still requires bridging: produce a clearing record mapping Order Items → pseudo coupon until industry clearing house accepts pure orders.

Accounting Controls: Reconstruct “what would filed fare have been” for auditing tax and surcharge compliance. Differences logged for audit team.

12. Interline & Alliance Implications

Problem: Partner still expects RBD + filed fare. Your continuous price must interoperate.

Near Term: Use anchor filed fare for prorate, allocate adjustment proportionally across sectors (distance or revenue share weighting).
Token Passing: Include OfferContextToken in interline request; if partner cannot consume it, revert to discrete anchor.
Proration: Maintain legacy proration engine; feed it synthetic coupon values consistent with sum of sector splits.
Future: Order-based interline settlement service (industry initiative) accepts full component price lines; no need to compress into RBD-based fare basis codes.

Mitigation: Flag risk of value leakage if partner forces rounding up/down to ladder steps—monitor proration variance vs intended continuous distribution.

13. Example System Capability Categories (Illustrative)

(Vendor names are examples of publicly known categories; validate fit & roadmap directly with providers.)

Fare Filing & Reference: ATPCO platforms (fare, rules, branded fares, optional services, rich content).
Revenue Management / Optimization Engines: Network O&D optimizers (e.g., PROS RM, Accelya, Amadeus, Sabre RM suites) producing bid prices / protection levels.
Dynamic / Continuous Price Engines: Modules layered on RM (e.g., PROS Dynamic Pricing, Accelya FLX, Amadeus Altéa Dynamic Pricing, Sabre solutions) or in-house microservices.
Offer Construction / NDC APIs: Airline-built or platform (Offer/Order management layer from PSS or specialized vendor).
Order Management & Accounting Adapters: ONE Order pilots, middleware mapping offers to ticket/EMD or direct order ledger.
Analytics & Experimentation: A/B frameworks, price realization dashboards, anomaly detection for guardrail breaches.

Disclaimer: Capability names are generic; confirm current feature support (continuous interpolation, tokenization, attribute mapping) with each vendor—roadmaps evolve.

14. Data, Governance & Controls

Provenance Envelope: Each Offer includes: pricingMethod (FILED|DYNAMIC|CONTINUOUS), anchorFareId, adjustmentType, modelVersion, guardrailId, taxTableVersion.
Guardrail Service: Central rules: minMargin, maxDiscount%, price parity (channel A vs B), tax compliance.
Explainability: Log feature influences (e.g., loadFactor=78→+4%, advancePurchase=21d→baseline).
Fairness / Compliance: Explicit exclusion lists to prevent unintended discrimination (avoid using protected attributes / proxies).
Model Lifecycle Mgmt: Version, canary release, rollback, drift detection (monitor realized vs predicted conversion uplift).

15. Key KPIs & Observability

Price Gap Reduction (avg absolute difference between optimal bid price and sold price).
Uplift vs Control (revenue per seat vs legacy ladder cohort).
Guardrail Breach Rate (# offers rejected at commit / total).
Offer->Order Reprice Rate (should trend downward with stability).
Interline Rounding Loss (difference between sold continuous price and interline accepted sum).
Model Drift (prediction error of conversion uplift vs actual).
Attribute Coverage (% of direct offers constructed without RBD dependency).

16. Risks & Mitigations

Risk	Description	Mitigation
Channel Disparity	Consumer confusion / regulatory scrutiny	Parity guardrails + reason codes
Model Overfitting	Unstable prices harming trust	Out-of-sample tests, dampening factor
Accounting Complexity	Delayed close & audit issues	Virtual fare basis mapping + reconciliation job
Interline Rejections	Continuous price not accepted	Fallback anchor price handshake
Tax Miscalculation	Edge discount breaks tax rules	Central tax engine mandatory invocation
Latency Inflation	More compute per request	Async pre-compute bands + cache safe corridor
Governance Drift	Multiple unaligned pricing algorithms	Single pricing policy registry

17. Plain-Language Glossary

Filed Fare: Official price with rules published to industry (via ATPCO).
Fare Ladder: Ordered list of discrete filed prices for the same journey.
RBD (Reservation Booking Designator): One-letter code controlling seat inventory and pricing bucket.
Dynamic Pricing (Discrete): Adjusting to more granular steps still tied to filed anchors.
Continuous Pricing: Selecting any price in a model-authorized range, not just a pre-filed step.
Bid Price: Minimum acceptable revenue for a seat given network demand forecasts.
Guardrail: Rule preventing a price from falling outside safe or compliant bounds.
Offer: A time-bound proposal with price + product attributes (before commitment).
Order: Record of customer entitlement replacing ticket + PNR concept.
RBD-less: Retail model not relying on booking class letters for price determination.
Virtual Coupon: Transitional artifact linking order items to legacy accounting.
Proration: Splitting revenue of a multi-carrier journey among participants.

18. Implementation Checklist

Baseline price gap analysis vs bid price.
Define guardrail corridor (min/max vs anchor fare).
Implement Offer provenance (pricingMethod, modelVersion).
Virtual fare basis mapping service (anchor + adjustment).
Token validation at commit (reprice if expired).
Monitoring dashboard: guardrail breaches, uplift, reprice rate.
Interline fallback logic (continuous → anchor negotiation).
Attribute catalog & RBD mapping version control.
Model experimentation harness & rollback switch.
Accounting reconstruction service (audit: filed vs sold).
Tax engine integration validated for discount scenarios.
Risk log with owner & mitigation status reviewed monthly.

19. References & Disclaimers

ATPCO public materials on fare filing, branded fares, optional services (consult official ATPCO documentation for authoritative definitions).
IATA NDC & ONE Order high-level industry publications.
Public conference presentations (AGIFORS, T2RL, PODS) discussing dynamic / continuous pricing concepts.
Vendor whitepapers (PROS, Accelya, Amadeus, Sabre) describing dynamic pricing approaches (verify current state directly).
Industry articles on attribute-based retailing and RBD de-emphasis.

Disclaimer: This article synthesizes publicly known concepts and general engineering patterns. It does not include proprietary algorithms, confidential ATPCO data, or vendor-specific implementation secrets. “Dynamic” and “Continuous” definitions vary; confirm alignment with ATPCO and internal legal/compliance before execution. Placeholder steps marked “confirm with ATPCO” require direct validation.

Trademarks belong to their respective owners. No endorsement implied.

Back to all blogs