OCBC Bank's establishment of a regional service analytics team in 2024 represents a deliberate shift from reactive to predictive customer service operations, with measurable results: the proportion of calls answered within 60 seconds rose from 52% in 2023 to 80% by 2025. The bank's approach centres on identifying recurring friction points—forgotten card activations, overseas spending blocks, international dialling charges—and intercepting customers before they contact the centre through proactive push notifications and SMS. This deflection strategy has materially reduced inbound call volume, whilst the introduction of an in-app calling feature for overseas customers addresses a specific barrier identified through data: 92% of customers with card issues abroad were avoiding contact due to IDD charges and verification friction. The operational logic is sound: prevent the problem from escalating to agent handling by removing the conditions that create the contact in the first place.
The implications for CX teams extend beyond call reduction metrics. OCBC's use of generative AI to analyse millions of call minutes for quality patterns and coaching opportunities signals a maturation in how analytics platforms can drive frontline performance at scale—a capability that raises questions about whether traditional quality assurance workflows, built around sampling and manual review, remain fit for purpose in organisations with sufficient data volume. The bank's commitment to following up on one- and two-star ratings and escalating systemic issues to product owners also demonstrates how analytics can function as a feedback loop that transcends the support function itself, though this requires organisational structures capable of acting on insights. For teams already managing complex omnichannel environments, the OCBC model suggests that the competitive advantage lies not in implementing AI tools generically, but in the specificity of the problems you choose to solve first—and whether your data infrastructure can identify those problems before customers do.
OCBC data analytics for customer service The Straits Times