Cisco's Painful AI Transition: A Surgical Transformation

Integrating artificial intelligence into a colossal entity like Cisco, a company valued at nearly $500 billion with over 80,000 employees worldwide, is no small feat. For Liz Centoni, Cisco's Chief Customer Experience Officer, this transformation is akin to 'surgery without the drugs', a painful but necessary overhaul.

Rethinking Old Workflows

As a provider of networking, security, and collaboration technology, Cisco's customer support is critical. With roughly 20,000 employees in the customer experience division, Centoni is spearheading the shift towards becoming an AI-native services organization. It's not enough to simply bolt AI onto existing processes, a strategy Cisco initially pursued only to discover it accelerated flawed workflows, particularly in customer support.

The early use of generative AI to create case summaries for support engineers inadvertently 'annoyed customers faster,' as handoffs became more efficient without resolving the underlying problem. The real goal wasn't the handoff, Centoni realized, but ensuring that cases reach the right engineer at the outset.

Intelligent Routing: A New Approach

This insight led to the development of 'intelligent routing,' which directs cases to the appropriate expert from the start. The results are promising, of the 1.5 million support cases Cisco handles annually, nearly 88% are now correctly routed on the first attempt. Cisco's current metric for customer service success focuses on minimizing the number of calls that require multiple handoffs.

So, why does this matter? In an era where customer satisfaction is key, effective AI integration can be a big deal for maintaining competitiveness. Yet, the process isn't as simple as adding AI into the mix. It demands a fundamental redesign of workflows.

Targeting Repeatable Workflows

Centoni highlights the significance of deploying AI in repeatable workflows, where tasks can be autonomously performed with over 90% accuracy. The recent launch of Cisco IQ, a digital interface for support and professional services, exemplifies this approach. It's designed to be a 'single source of truth' for customers, addressing recurring pain points, detecting preventable outages, and reducing time wasted on data interpretation.

Ultimately, the test for any AI initiative isn't just enhanced efficiency. As Centoni puts it, each project must demonstrate tangible benefits, whether revenue growth, margin expansion, increased customer trust, or enabling teams to innovate. In a company as vast as Cisco, this is no small challenge. Yet, if successful, it could redefine customer experience in the tech industry. The FDA doesn't care about your chain. It cares about your audit trail.