Weighted Revenue Sharing Rules

• Weighted revenue sharing rules are formal mechanisms that allocate revenue based on weights reflecting contributions, performance metrics, and contextual criteria.
• They generalize equal and proportional rules by integrating domain-specific factors, ensuring fairness and efficiency through robust axiomatic frameworks.
• Practical implementations in AI, digital platforms, and transportation use algorithmic adjustments like thresholds and temporal decay to enhance reliability and reduce manipulation.

Weighted revenue sharing rules are formal mechanisms for allocating a pool of revenue among multiple parties, with each party’s share determined via a system of weights reflecting various forms of contribution, entitlement, or engagement. These systems have been foundational in industries as diverse as AI data markets, cloud computing, transportation, sports broadcasting, streaming, and digital platforms. Weighted rules generalize both equal and purely proportional allocations, embedding context-specific information—such as performance, usage, or exogenous criteria—into the division formula, and are often preferred for their flexibility, scalability, and ability to encode incentive-relevant distinctions between participants.

1. Formal Frameworks and Canonical Formulas

Weighted revenue sharing rules typically allocate a revenue pool $R$ among $N$ participants $i=1,\dots,N$ by assigning each a (possibly vector-valued) weight $w_i > 0$, and distributing shares as

$s_i = \frac{w_i}{\sum_{j=1}^N w_j} \cdot R.$

This proportional-weight rule forms the mathematical backbone of schemes ranging from IATA airline prorates and sports broadcast allocations to machine learning data attribution (Bergantiños et al., 11 Dec 2025, Bergantiños et al., 5 Jan 2026, Zhang, 2023). In contexts where weights arise endogenously from workflow, performance metrics, or cooperative game-theoretic values (e.g., marginal contribution, Shapley value, core allocations), the system coordinates with domain-specific models and axioms.

Weighted rules can be contrasted with unweighted (equal split) schemes and can further interpolate between these extremes using parametric families, as in convex combinations or hierarchy-aware generalizations (Bergantiños et al., 2023, Yu, 14 Jan 2026).

2. Axiomatic Characterization and Theoretical Guarantees

Weighted revenue sharing rules can be uniquely characterized by axioms reflecting additivity, independence, homogeneity, ratio preservation, and various neutrality or fairness properties. The literature establishes, for example, that:

• For airline ticket revenue sharing, rules satisfying additivity, null-player, independence of other airlines, and ratio preservation are precisely the weighted-flights rules parameterized by a system of positive leg weights $w_e$ (Bergantiños et al., 11 Dec 2025).
• In streaming and subscription platforms, homogeneity, additivity, and various forms of anonymity (stream-based vs user-based) distinguish pro-rata, user-centric, and weighted hybrid rules; certain axioms (e.g., fraud-proofness, strong Sybil-proofness) further restrict to specific forms (Bergantiños et al., 2023, Ghosh et al., 6 Nov 2025).
• In cooperative games, the Shapley value, weighted Shapley value, Ortmann proportional rule, and core-based allocations have theoretical underpinnings guaranteeing efficiency, fairness, and, under appropriate cost or revenue structures, Nash or Wardrop equilibrium existence (Cao et al., 2017, Gopalakrishnan et al., 2014).

Key theorems state, for example, that for any exogenously fixed weights, the weighted-flights rule is uniquely determined by additivity, null-player, and pairwise homogeneity (Bergantiños et al., 11 Dec 2025); or, for streaming platforms, that the ScaledUserProp rule is the unique (parameterized) user-additive, fraud- and bribery-proof revenue division mechanism given certain monotonicity constraints (Ghosh et al., 6 Nov 2025).

3. Application Domains and Domain-Specific Weight Construction

Weighted revenue sharing is pervasive across multiple verticals, with weight systems tailored to the particular economic and technical context:

• AI Data Attribution: Providers are assigned weights based on detected content similarity or classifier-based engagement to training texts—the sum of raw classifier probabilities or embedding cosine similarities over all production prompts—normalized to define revenue shares (Zhang, 2023).
• Airline Industry: IATA’s Standard Prorate Factors (SPFs) derived from Ticketed Point Mileage (TPM), adjusted by regional/route-specific multiplicative factors and rounding, govern weights in a manner justified axiomatically and operationalized through published manuals (Bergantiños et al., 11 Dec 2025).
• Sports Broadcasting: Weights can encode historical performance (e.g., points), ticket revenue, or audience market size, with each dimension corresponding to legislatively or contractually specified sub-pools (Bergantiños et al., 5 Jan 2026). Parametric hybrid rules interpolate between egalitarian (equal split) and performance-driven allocations.
• Digital Content Platforms: In music streaming, the pro-rata rule uses total stream counts, user-centric uses per-listener proportional weights, and α-hybrid rules define weights as convex combinations of these, enabling platforms to balance superstar and niche compensation (Bergantiños et al., 2023, Yu, 14 Jan 2026).
• Dynamic or Cooperative Game Settings: Weighted Shapley and marginal-contribution rules assign player-specific λ-weights, encoding priority or privilege, and serve to guarantee Nash equilibrium and potentiality in resource-sharing environments such as edge-cloud computing or facility location (Cao et al., 2017, Gopalakrishnan et al., 2014).
• Market Mechanisms: In two-sided or networked markets (e.g., ISP-WiFi, tree bargaining, Bertrand with affiliate fees), weighted shares arise from bargaining models or endogenous solution of fixed-point equations, adapting dynamically to context-specific outside options, cost asymmetries, or power balances (Susanto et al., 2017, Ghosh et al., 2013, Pabari et al., 11 Feb 2025).

4. Practical Adjustments and Algorithmic Implementation

Modern weighted revenue sharing schemes incorporate a range of practical adjustments to align incentives, ensure robustness, and manage scale:

• Thresholds: Providers with engagement scores below a minimal threshold are excluded or set to zero weight, with survivors’ weights renormalized (Zhang, 2023).
• Quality/Freshness Multipliers: Providers passing human review or meeting currency criteria may have scores multiplicatively boosted (Zhang, 2023).
• Temporal Decay: Engagement or performance scores can be time-discounted using exponential factors to prevent stale data from dominating (Zhang, 2023).
• Fraud-Resistance and Bribery-Proofness: Parametric hybrid rules (e.g., α-interpolation between pro-rata and user-centric, or ScaledUserProp caps) restore fraud-proofness under known limits of adversarial intervention and guarantee revenue per additional user is bounded (Ghosh et al., 6 Nov 2025, Yu, 14 Jan 2026).
• Computation: Aggregation is amenable to map-reduce paradigms; score normalization and payout are scalable to large $N$ by virtue of additivity and local computation, while fixed-point or bargaining equations in tree and network settings are solved via polynomial-time iterative methods (Ghosh et al., 2013, Zhang, 2023).

5. Comparative Properties, Fairness, and Strategic Implications

Weighted sharing rules are compared along dimensions including efficiency, group stability (core membership), fairness/inequality, incentive alignment, and robustness:

• Core Membership: Rules assigning shares along the winning coalition or highest-value path in cooperative games (e.g., egalitarian fixed point in tree trading networks) yield allocations in the core, guaranteeing stability against blocking (Ghosh et al., 2013, Cao et al., 2017).
• Fairness: Parameterized families allow for Lorenz-dominance ordering; increases in performance- or popularity-weighting shift distributions toward greater inequality, while compromise rules balance egalitarian and meritocratic effects (Bergantiños et al., 5 Jan 2026, Bergantiños et al., 2023).
• Robustness: Fraud- and bribery-proof rules (especially user-additive, capped, or proportionally attenuated forms) limit the payoff a coalition can extract from user-bot attacks or collusion, a property violated by global pro-rata formulas but achieved by ScaledUserProp and user-centric schemes (Ghosh et al., 6 Nov 2025, Yu, 14 Jan 2026).
• Efficiency and Social Welfare: Optimal settings of parameters (e.g., referral shares α in the shared-revenue Bertrand game) can expand the size of the feasible welfare region, permitting joint profit gains and improved consumer surplus compared to naive or monopoly allocations (Pabari et al., 11 Feb 2025).
• Sensitivity to Strategic Behavior: Proportional rules reward global popularity and may incentivize manipulation (e.g., fraudulent streams), while user-centric or capped rules mitigate perverse incentives but can cross-subsidize or flatten rewards for top performers (Bergantiños et al., 2023, Yu, 14 Jan 2026).

6. Illustrative Examples and Empirical Insights

A range of canonical and parameterized examples ground the theory:

• Airline Two-Leg Itinerary: For fares allocated between two airlines using respective SPF weights of 1299 and 4760, the split is approximately 21%:79%, highlighting the effect of weighted systems over simple equal division (Bergantiños et al., 11 Dec 2025).
• Streaming Platform with α-Hybrid Rule: For α=0.2, the revenue formula for an artist interpolates between global and user-centric weights; the larger α, the greater the payoff to highly streamed artists, but incentive compatibility with respect to fraud requires α be capped according to the detectability of fake streams (Bergantiños et al., 2023, Yu, 14 Jan 2026).
• Tree Trading Network (Egalitarian Fixed-Point): Computation proceeds by binary search along monotonic upward and downward curves in the edge share variable, converging to a polynomial-time solution unique for a given competitor value configuration (Ghosh et al., 2013).
• Subscription Platform (ScaledUserProp): Each user’s contribution to the artist pool is $\min(\gamma\,\lVert w_i\rVert_1,1)$, with $\gamma$ chosen so the sum of capped contributions matches total redistributable subscription fees, tightly controlling pay-per-stream dispersion and manipulation (Ghosh et al., 6 Nov 2025).

7. Open Challenges and Extensions

Current and future research extends weighted revenue sharing principles to settings with:

• Dynamic populations: Consistency and population monotonicity as organizations, providers, or creators enter and exit.
• Multi-criteria aggregation: Simultaneous weighting by orthogonal dimensions (e.g., performance, market size, guaranteed minimums).
• Nonlinear or networked interactions: Beyond linear proportionality to address complementarities, externalities, or nonlinear synergies in value generation.
• Algorithmic detection and transparency: Efficient algorithms for manipulation detection, explainable scoring, and third-party auditability in large-scale platforms (Zhang, 2023, Ghosh et al., 6 Nov 2025).
• Game-theoretic optimal design: Balancing incentive compatibility, efficiency, and fairness under strategic competition and asymmetric information.

Weighted revenue sharing thus serves as a unifying framework for principled allocation in complex, multi-agent environments, supporting both theoretical rigor and practical robustness across diverse economic and technological contexts.