A stark reality check awaits contact centre leaders pursuing AI transformation: while 95% of enterprise AI initiatives fail to reach production, the underlying cause isn’t technological limitations – it’s the critical shortage of expertise needed to execute successfully.
This was the sobering message from Stuart Dorman, chief innovation officer at Sabio Group, who was addressing delegates at the recent UK National Contact Centre Conference – challenging the industry’s approach to AI adoption while revealing why voice interaction could become the dominant customer service channel within three years.
“Big tech wants you to believe AI deployment is simple – just plug in and transform your operation,” says Dorman. “The reality is starkly different. Anyone can build a demo bot in an afternoon, but scaling AI to handle thousands of voice interactions at high performance levels requires expertise that many organisations simply don’t possess.”
Dorman’s keynote exposed the uncomfortable truth behind why so many AI implementations fail. Successful deployment demands a rare combination of skills: user experience design, linguistics expertise, AI prompt engineering, data science capabilities, systems integration knowledge – and, in the contact centre – a deep understanding of operations.
“The cost of AI technology is plummeting – ChatGPT tokens have reduced 1 000-fold in just three years for example – but expertise to exploit this technology is difficult to find as demand outstrips supply,” Dorman says. “It’s becoming particularly harder in the contact centre where organisations are struggling to find the skilled resources required to bring AI to life within a CX transformation.”
This creates a paradox: while underlying AI model costs decrease 10-fold annually, the specialist knowledge required for implementation becomes increasingly expensive and scarce.
“Most AI deployments are internally facing ‘knowledge agents’ and these are great as a starting point, but they are only scratching the surface of what AI can do,” Dorman explains. “There are lots and lots of exciting potential use cases where AI has an external focus and this is where the true potential for AI technology lies. That’s something that we’re actively discussing with our customers at the minute.”
Contradicting predictions of text-based dominance, Dorman argues that voice will become the fastest-growing customer service channel over the next three years – driven by a fundamental human preference.
“That’s providing organisations stop trying to prevent customers from speaking to them by hiding phone numbers and forcing the use of poorly implemented chatbots,” he says.
“We speak at 150 words per minute, but type at only 40,” he told delegates. “As AI eliminates traditional channel boundaries, customers will naturally gravitate towards voice interaction. The future isn’t about replacing phone support – instead it’s about making every digital touchpoint conversational.”
The conference tackled a provocative proposition: can AI actually deliver superior customer experiences compared to human agents for specific interaction types? Early implementations suggest this isn’t just possible, but inevitable.
“What happens to your operation when you can deliver better experiences at 10 times lower cost?” Dorman asks. “This isn’t about doing more with less – it’s about unlocking entirely new service possibilities.”
For the estimated 30% of Fortune 500 companies planning single AI-enabled channels by 2028, the message is clear: success depends more on implementation expertise than technology selection.
“The organisations thriving in AI transformation aren’t necessarily those with the biggest budgets,” says Dorman. “They’re the ones who recognise that expert guidance through the complexity of real-world deployment is non-negotiable.”