Customer service leaders have long been promised that AI is the silver bullet for all their (many) challenges: streamlining workflows, enhancing customer satisfaction, and, most importantly, reducing costs via lower call volumes. This enthusiasm has only intensified with the adoption of generative AI (genAI) in the enterprise.
To be clear: GenAI will transform the customer service landscape. Large language models have significantly expanded AI’s potential in customer service, and the future looks bright.
But — and it’s a big but — one pretty major oversight is holding back progress: Enterprises are not capturing the tacit knowledge that employees have but AI solutions don’t.
What Does This Gap Mean?
For years, contact center vendors have painted the vision of training AI to be as good as your “best” agents — you just need to train AI with your call transcripts. The problem with that approach is the transcript doesn’t capture many elements (and often, those elements aren’t captured anywhere). Tacit knowledge that AI systems don’t capture includes intuitive decisions that a human employee makes or context-dependent conditional knowledge — for example, undocumented exceptions to company policies. Without this knowledge, enterprises will struggle to automate anything beyond the simplest customer service inquiries.
What Can You Do?
Capturing this tacit knowledge must start with brands evolving to having AI — not humans — lead and own conversations from start to finish. Humans will then transition to supporting AI behind the scenes to problem-solve exceptions within AI-led interactions. We expect a new type of agent workspace to emerge — one that is specifically designed to capture tacit knowledge and begin putting it to work through AI models. Here’s a visual example of how things would change in this scenario:
If you think that this represents a big shift in how contact centers operate today, then you’re right; it certainly does. And that’s why we think the transition to fully AI-led customer service will take years to mature and will evolve across three phases.
Learn about these phases — and what enterprises should do right now to prepare for the future of AI-led customer service — in our report, Tacit Knowledge Will Power The AI-Led Contact Center. Forrester clients can also book a guidance session with me to understand what’s ahead and start mapping their next steps to developing an AI-led contact center.
