Thank you for Subscribing to Hospitality Business Review Weekly Brief

Beyond the bot: Scaling the Cx in the age of Ai


The Scaling Paradox
Anyone occupied with scaling a fastgrowing business is familiar with the tension between Growth on the one hand, and Quality on the other. Traditionally, scaling has meant a risk of diluted service, longer wait times and often outsourcing.
AI has come a long way to assist, not as a replacement for human touch, but with an infrastructure which allows humanity to scale.
That’s why in 2026, the goal of AI in CX isn’t just deflection. It’s proactive relevance.
The Evolution of the Tiered Model
There’s little doubt that AI has taken away lots of simple tasks which were considered level one troubleshooting for support teams, with zero latency. This has given countless opportunities for anyone working with customer support.
Additionally, AI surfaces data, sentiment and documentation, reducing the cognitive load for humans so they can focus on problem-solving or more complex tasks. In that sense, one might argue that Level two is the new level one in terms of complexity, and it raises the question of how the need for seniority and talent profiles will shift in our teams over time.
Early adopters are leading the game. Those who know how to use AI effectively in their role and integrate it for a seamless customer experience are defining what tomorrow’s CX will look like.
But seamless bot flows, co-pilots, summaries and perfect translation skills aren’t enough. AI offers an array of new possibilities where customers can be surprised and delighted.
Scaling with AI needs to be Ethical by Design, so that even at a massive scale, the system is set up to review, decide and take actions within the scope we want and within the values we have.
Predictive CX: Countering the Inbound Ticket
Agentic AI opens new doors to how businesses can turn the customer journey into a proactive, outbound and value-driven experience. This can take a significant load off the shoulders of Account and Customer Success Managers, and help anyone in post-sales teams prioritise accounts showing indication of bad health.
This is not only necessary but already a growing customer expectation - they expect us to spot issues and errors, notify them and ideally fix any problems before they have detected anything themselves. The good news is that it can absolutely be done.
By combining account specific data, based on user patterns and behaviours, operational information and insights and clever AI guidance and actions, any issues related to account health or fluctuating operational results, can be detected, fixed and notified to the customer - in that order.
And this level of personalisation is the new world. Customers get their own AI agent, entirely adapted to their needs and expectations, an AI agent who knows the customer better than any human will have the capacity to do, and proactively taking actions based on customer permissions and expectations. This is where the magic starts to happen.
Other flows may involve an Onboarding Specialist, Account Manager or a Support Agent earlier, to ensure a human takes part of the process. Not necessarily because the AI agent cannot handle everything, but because a human will sometimes be needed to add value wherever some sort of sentiment and empathy are required.
The Human Guardrail
Which brings us back to this: When we scale, we aren’t just scaling output, we are scaling decisions. If these decisions are made by an AI which isn’t built on integrity and fairness, then we aren’t scaling a business but rather scaling a liability.
This is why scaling with AI needs to be Ethical by Design, so that even at a massive scale, the system is set up to review, decide and take actions within the scope we want, and within the values we have.
Consider These Questions
As we scale, let’s ask ourselves some challenging questions to keep us on track: If we removed our AI tomorrow, would our customers miss the convenience, or would they simply be relieved to finally be able to speak to a person quicker again?
Are we using AI to actually solve root causes, or are we just getting faster at answering the same avoidable questions?
In five years, when every competitor has the same ‘perfect’ AI responses and handling, what will be the unique Human Signature that defines our brand experience?