Organisational Readiness for AI Contact Center Adoption in Indonesian Banking

Abstract

Artificial intelligence (AI) is transforming customer service operations in banking, with contact centers as a primary deployment. This study examines the causal structure of organisational readiness factors for AI contact center adoption in Indonesian banking and develops a scenario planning framework to guide implementation strategy across institutional contexts. The research employs a sequential multi-method design: ten in-depth interviews (IDI) across six functional domains (contact center operations, product management, regulatory and risk, technology and AI, compliance, and customer perspective); reflexive thematic analysis to identify and consolidate implementation factors; DEMATEL (Decision-Making Trial and Evaluation Laboratory) causal analysis by eight domain experts; MICMAC structural classification; and four-scenario strategic planning. Thematic analysis of interview transcripts consolidated 10 final factors using discriminant validity criteria. DEMATEL analysis (n = 8; normalisation parameter s = 33.875; significance threshold α = mean(T) + 1·SD = 0.8259) identified 19 significant causal relationships. The 10 factors were categorised within the Technology-Organisation-Environment (TOE) framework. DEMATEL causal analysis revealed that all five Cause factors (r−c > 0) belong to these dimensions: FIN, MAN, PEL, KOM, and NAS, while the Effect factors belong to these dimensions: AKU, KEA, DAT, INF, and REG. AI Response Accuracy is the most structurally central factor, confirming its role as the convergence point of all causal investment pathways. MICMAC structural analysis classified Financial Readiness, Management Commitment, and Agent Competency as Driving variables; Technology Infrastructure, AI Response Accuracy, and Data Security as Relay amplifiers; Data Availability and Regulatory Compliance as Output indicators; and HR Training and Customer Readiness as Autonomous variables. CLD analysis identified five reinforcing loops (R1–R5) and one balancing loop (B1). The scenario planning framework constructs a two-axis matrix: Organisational Readiness (X = FIN×0.55 + MAN×0.45, weights proportional to r−c magnitudes) and Ecosystem Support (Y = NAS). Four scenarios are mapped against KBMI institutional profiles: Scenario A (Optimal Transformation, KBMI IV), Scenario B (Independent Innovation, KBMI III), Scenario C (Ecosystem-Led Adoption, Digital banks), and Scenario D (Pre-Adoption, KBMI I–II conventional banks). The principal finding is that AI contact center implementation in Indonesian banking is fundamentally an organisational challenge that manifests as a technology challenge. Financial readiness and management commitment must precede technology selection; data quality and AI accuracy are outputs of organisational investment, not independent prerequisites. These findings invert the conventional technology-first implementation narrative and provide empirically grounded, KBMI-stratified guidance for banking practitioners, regulators, and future research.