Hierarchical SAW with AHP-Derived Weights for Mustahik Prioritization: Benchmarking and Validation in Zakat Institutions

Abstract

Manual mustahik (zakat beneficiary) selection in Indonesian Zakat Collection Units (UPZ) is prone to transcription errors, recency bias, and poor scalability — undermining the equitability and auditability required under BAZNAS institutional standards. This study proposes a web-based Decision Support System (DSS) integrating a hierarchical Simple Additive Weighting (SAW) model with Analytic Hierarchy Process (AHP)-derived criterion weights. A five-criterion, eleven-sub-criterion model was developed from BAZNAS assessment standards and validated by three domain experts, yielding a Consistency Ratio of CR = 0.033 (< 0.10). SAW rankings were benchmarked against TOPSIS and VIKOR using Spearman’s ρ and Kendall’s τ on a 15-candidate dataset, indicating strong preliminary agreement (SAW–TOPSIS: ρ = 0.943; SAW–VIKOR: ρ = 0.886), pending larger-dataset replication. Sensitivity analysis confirmed top-3 ranking stability under all ±10% and most ±20% weight perturbations. Performance benchmarking demonstrated linear scalability with sub-100 ms computation at N ≤ 1,000 candidates. Black Box testing across 63 scenarios achieved a 100% pass rate. The system was implemented using the Laravel MVC framework with a role-based three-tier approval workflow (Surveyor → Admin → UPZ Head). Primary limitations include a small expert sample (n = 3), a limited validation dataset (N = 15), and the absence of formal User Acceptance Testing, which constitutes the primary direction for future research.