Valoir's Rebecca Wettemann identifies a critical inflection point in enterprise AI adoption: the market has shifted from fear of missing out to fear of messing up. Organizations have moved beyond early-stage experimentation and now face mounting pressure to demonstrate measurable business outcomes from AI investments without eroding customer trust, inflating operational costs, or destabilizing existing workflows. This represents a fundamental change in how CX leaders evaluate AI initiatives. Where pilots once served as proof-of-concept exercises, they now face scrutiny as potential harbingers of expensive failures. The tension is acute because the same leaders championing AI adoption must simultaneously manage legitimate concerns about hallucinations, unpredictable pricing models, and the persistent gap between vendor promises and real-world performance—concerns reinforced by recent evidence that third-party generative AI tools are outperforming brand-built chatbots.
Data quality emerges as the primary technical barrier, yet it remains largely invisible in vendor marketing. Wettemann emphasises that AI cannot deliver reliable customer-facing outcomes without consistent, current, and correct data—a requirement that exposes the fragility of many existing CX stacks. For teams running multiple platforms (Zendesk, Freshdesk, Salesforce Service Cloud), this means AI ambition often collides with data integration reality. The implication is stark: organizations cannot simply bolt agentic AI onto legacy systems and expect results. Instead, Wettemann advocates breaking work into smaller, discrete skills where the business case is defensible, then scaling methodically. This approach directly challenges the vendor narrative of comprehensive, end-to-end AI solutions.
The workforce dimension adds urgency to this recalibration. As AI handles routine interactions, human agents shift toward supervision, refinement, and higher-value problem-solving—a transition that demands rethinking compensation structures, performance metrics, and onboarding programmes. CX leaders must ask themselves whether their current talent strategies and incentive systems are designed for this new reality, or whether they remain locked into metrics that reward volume over quality. For teams already managing agent burnout and retention challenges, this represents both opportunity and risk: AI can elevate work, but only if the organizational infrastructure supports it.
How can CX leaders scale AI without creating more risk than value? That is the central question in this CX Today interview with Rebecca Wettemann, Principal at Valoir, as she joins Rob Wilkinson to unpack how enterprise thinking around AI in customer experience is changing. The conversation shows ho