The adoption-to-excellence gap in AI-powered customer experience has reached a critical inflection point. Nearly 70% of CX teams have deployed some form of AI, yet only 2% have achieved Centre of Excellence (CoE) maturity—a chasm that exposes the difference between implementation and operational mastery. This disparity reflects a familiar pattern in enterprise software: early adopters rush to deploy generative AI and agentic systems to capture competitive advantage, but few organisations have built the governance, measurement frameworks, and cross-functional alignment required to move beyond pilot-stage performance. For Zendesk administrators and support leads currently managing these deployments, the question becomes acute: are you optimising for quick wins or building the infrastructure that separates mature programmes from those destined to plateau?
The implications cut across three dimensions. First, the 68-percentage-point gap signals that most teams lack standardised approaches to AI governance, quality assurance, and continuous improvement—the hallmarks of CoE operations. Second, this creates opportunity for vendors and consultancies offering maturity frameworks, but it also suggests that teams investing in platforms like Salesforce Agentforce or competing agentic solutions may be acquiring powerful tools without the organisational scaffolding to extract their full value. Third, the concentration of CoE programmes likely skews toward larger enterprises with dedicated AI centres and budget for specialised talent, meaning mid-market and smaller teams face a steeper climb to operational excellence without external guidance or tooling that abstracts away complexity.
What separates the 2% from the rest is not technology choice but execution discipline: documented workflows, measurable KPIs tied to business outcomes, regular model retraining cycles, and cross-functional ownership between CX, data, and product teams. For teams currently in the 68%, the path forward requires honest assessment of whether your current stack and team structure can support the governance layer that CoE status demands. The risk is not that AI adoption will stall—it won't—but that organisations will accumulate technical debt and missed ROI opportunities by treating AI as a feature rather than a capability requiring systematic maturation.
AI Adoption in CX Nears 70%, Yet Only 2% of Programs Reach the Center of Excellence Standard, Forethought Report Finds Business Wire