Cisco exec says adopting AI is like 'surgery without the drugs'

Integrating artificial intelligence into a vast global organization is no easy feat, and Cisco's chief customer experience officer, Liz Centoni, likened the process to undergoing 'surgery without the drugs.' This analogy reflects the discomfort and challenges faced when trying to transform workflows within a company employing tens of thousands across various locations worldwide. As a leading provider of networking, security, and collaboration technologies, Cisco's customer support division plays a crucial role in its operations. Centoni, who oversees approximately 20,000 employees in customer experience, is currently leading an initiative to evolve this division into an AI-native services organization. The journey has proven to be more complex than merely attaching AI to existing processes, a strategy that previously resulted in merely speeding up flawed systems. For instance, when Cisco initially implemented generative AI to create case summaries during handovers between support engineers, the outcome was less than satisfactory. Centoni noted that while this approach made transitions smoother, it ultimately 'annoyed our customers faster' as it did not resolve the core issue of directing inquiries to the appropriate expert from the outset. This realization prompted a significant workflow redesign, incorporating 'intelligent routing' to ensure cases are directed to the right expert initially. Cisco's customer experience division handles approximately 1.5 million support cases annually, and now nearly 88% are correctly routed on the first attempt. According to a Cisco spokesperson, success in customer service is now measured by the number of interactions that require only a single handoff or none at all. Centoni emphasized that the most effective areas for AI integration are those involving repeatable workflows that can be executed autonomously with over 90% accuracy. Recently, Cisco introduced Cisco IQ, a comprehensive digital interface designed to serve as a singular source of truth for customers, addressing recurring issues and enhancing overall experiences. This platform aids in identifying preventable outages and minimizes time spent on data interpretation, simultaneously reducing the volume of frustrating support calls. Ultimately, Centoni asserts that the true measure of success for any AI initiative is not just increased efficiency. Each project must demonstrate its capability to cease unproductive work, drive revenue growth, enhance profit margins, build customer trust, or support teams in developing future innovations.