SK Telecom's rollout of AI across T World stores represents a deliberate shift from customer satisfaction metrics toward operational intelligence and real-time consultation quality. The initiative spans three distinct user groups—customers, agency owners, and frontline staff—with tailored AI agents designed to address friction at each touchpoint. Customer-facing deployments include personalised store search (launching by 2027), consultation analysis that automatically summarises interactions, and pre-visit booking systems to reduce wait times. For agency owners, the AI agent diagnoses store performance and workforce efficiency; for T Crew, it functions as an embedded knowledge system during consultations, flagging weak areas and suggesting improvements in real time. The foundation for this rollout already exists: SKT has been collecting STT-based survey data since June and piloting secure consultation recording across 300 stores, establishing the data infrastructure needed to train these agents.
The implications for CX teams are substantial, particularly around the tension between automation and human accountability. SKT's approach—embedding AI agents within human-led consultations rather than replacing them—mirrors the agent augmentation model gaining traction in enterprise CX, but the consultation analysis function raises a critical question: as AI systems automatically summarise and diagnose consultation quality in real time, how do teams maintain coaching rigour and prevent the metrics themselves from becoming the target rather than proxies for genuine customer outcomes? The phased rollout also signals that large retailers are moving beyond chatbots and deflection metrics toward systems that measure consultation completeness and trust—a shift that should prompt CX leaders to audit whether their current metric stacks (CSAT, AHT, NPS) still capture what matters when AI is handling triage and frontline staff are handling complexity.
SKT's investment in agency owner and staff-facing agents is equally telling. By treating T Crew as a distinct user segment requiring real-time guidance and diagnostic feedback, SKT acknowledges that frontline performance is a CX lever that most platforms still underinvest in. This creates a secondary implication: as AI agents become standard infrastructure for customer-facing teams, the competitive advantage shifts to organisations that can integrate agent performance data with customer feedback loops—meaning CX teams will need tighter alignment with operations and workforce management than traditional support-only structures allow.
SKT to roll out AI across T World stores to upgrade consultations and operations 디지털투데이