We live in an age where someone was able to deploy a voice AI to book them a dentist appointment, only to discover that the dentist also had an automated voice AI receptionist, the two AIs subsequently had a conversation with one another for over two hours, possibly talked about the weather among other things, but crucially failed to get an appointment booked.

While this may be a glib example of AI’s increasing influence on our day-to-day lives, it also highlights that the world of customer contact as we know it is being thrown up in the air, with organisations of all types and size trying to evolve at breakneck speed and experiment in myriad ways, in an attempt to keep up.

The question for contact centre leaders is not whether to change, but how to change and why.

Whether AI is going to have an impact on our operations is not up for dispute, but understanding in what way and to what end is where the real complexity begins.

The Pace of Change

AT&T forecast that artificial natural language conversations were “just around the corner” in 1986. Gartner predicted in 2023 that 95% of customer service interactions would be powered by AI by 2025.

It’s 2026 and we are still some distance from realising either of these assessments. So, while the pace of change is real, we are also guided by acts such as that of Klarna, a company that in 2024 bet big on the AI revolution by upending its operation model and laying off all of its 700 customer service advisors in one fell swoop. 14 months later it had reversed the decision and rehired humans, acknowledging that it had been a failed experiment.

What does hold up is the principle that operational design matters more than technology. The most common failure in contact centre transformation is not selecting the wrong technologies and platforms – it is making technology fit old processes that have yet to be liberated to imagine what is possible.

The sequence that tends to work in is as follows – define it, design it, differentiate, disrupt. In that order. As with the Klarna example, too many organisations often skip straight to point four – ‘disrupt’.

However, the consequences of poor design are not abstract. According to recent research, bad customer service costs organisations around the world $3.7 trillion annually in lost sales. Yet companies consistently recognised for customer satisfaction delivered annualised returns of 17.7% between 2006 and 2024, compared to 10.5% for the S&P 500. The business case for getting this right is not a soft one.

Does the boardroom understand this? Have we articulated the importance of good operating model design in a language that holds their attention? Service failures are seldom treated as commercially critical. This is the tension that operational leaders navigate every day – making the case upward for investment in something that leadership tends to treat as a cost centre, while simultaneously managing a workforce increasingly anxious about what AI means for their roles.

Not Always About Headcount

This anxiety is not unfounded. Global organisations other than Klarna have laid off staff. Despite this, there are also major organisations that have shown that an operational model transformation isn’t merely about headcount. IKEA, for instance, used AI to handle nearly half of its customer contacts – and then retrained 8,500 contact centre workers as interior design advisers.

The technology and the human future of the contact centre are not necessarily in opposition, but they will only coexist well where organisations make deliberate choices about both.

So while we cannot necessarily predict the operating models of the future with absolute confidence, what we can do is build organisations designed to adapt – with clear principles, honest economics and the awareness to accept that the destination is one that continues to move. And that it is also one where, alongside humans, AIs may also be interacting with other AIs to try to deliver customer outcomes such as booking dentist appointments.

Hopefully with more success, in the future.

Themes examined in this article resulted from topic discussions and viewpoints provided at CCMA’s ReThink…Your Operating Model event in London, 10 March 2026, in partnership with TKG. 

Supported by