This discussion with Matt Price, CEO of Crescendo, opens with a recent forecast suggesting that the cost per AI-driven resolution could eventually exceed that of offshore human agents. Price acknowledges that while predictions remain uncertain, enterprises are beginning to evaluate the true cost of AI. This includes training, margin pressures, and operational realities that were previously overlooked.
A key theme is the emergence of hidden costs in AI deployments. Price explains that many organizations underestimate the additional staffing, workflow complexity, and customer experience challenges that can arise when AI is introduced. Rather than reducing workload as expected, AI can sometimes increase friction. This then leads to lower customer satisfaction and higher demand across alternative support channels. These factors highlight the gap between early AI projections and real-world outcomes.
The conversation also examines the transition from traditional seat-based pricing models to outcome-based approaches. Advances in measurement and observability now allow companies to track entire customer journeys across AI and human interactions. This enables more accurate assessment of whether issues are fully resolved and whether customers are satisfied. Price argues that pricing should reflect these end-to-end outcomes. The alternative is partial metrics such as AI containment rates, which can ignore unresolved issues downstream.