The balance between hi-tech and hi-touch

Recent research from MIT revealed that as many as 95% of company-wide AI projects ultimately fail. An impressive figure, but perhaps not surprising. Many of these projects focus too much on the visible front end (such as a chatbot that directly interacts with customers) while neglecting the back end, the systems and processes that should feed the information.

The result: garbage in = garbage out. If the data you use to feed a chatbot or AI assistant is incomplete, outdated, or poorly structured, you’ll only get poor output in return. AI, in that case, is not the holy grail. At best, it’s just the tip of the iceberg.

Take, for example, the so-called copilots for customer service employees. They promise to pre-sort emails, chats, or phone calls and automatically generate reply suggestions. Handy, but only if the input is correct. Once again, bad data in means bad suggestions out.

Moreover, copilots are often not properly tuned to the context of an organization. A standard model can formulate general answers, but it misses the nuance that’s so important in customer interactions. Think of specific tone of voice, business rules, or exceptions that are critical for customers. Without proper training and integration, a copilot can actually slow things down instead of speeding them up.

A copilot can therefore be a valuable accelerator, but never a miracle cure that instantly takes your customer service to the next level. It’s a tool, not a replacement for processes, systems or human expertise.

Here’s another point: not everything labeled as “AI” is truly AI. Many customer contact solutions are nothing more than clever flows or step-by-step scripts. A chatbot that doesn’t recognize your address simply because you ask for a human agent isn’t intelligent, it’s just a script. The difference matters, because it determines whether a solution can truly learn and adapt or whether it only follows a fixed route.

We’re seeing that AI is becoming less of a hype and more of a commodity: a technology that naturally integrates into the way businesses operate. This shift requires a new approach. No longer experimenting just to “do something with AI,” but deliberately thinking: where does AI add value, how do you ensure data quality and how do you integrate technology with existing processes and teams?

The challenge is not choosing between technology or human contact, but finding the right balance. Hi-tech can speed up processes, eliminate repetitive tasks, and be available 24/7. It delivers scalability and efficiency, and helps customers quickly with simple or repetitive questions.

Hi-touch (the human side) is indispensable when complexity and emotion come into play. Calming an angry customer, offering reassurance, or probing deeper into a problem requires empathy and flexibility. Qualities AI does not (yet) possess. On top of that, customers often simply want the reassurance that a real person is listening.

The organizations that succeed today are not those that blindly apply AI everywhere, but those that deliberately divide roles: where can technology add value, where is the human indispensable, and how can the two work together as one team? Companies that master this will reap the benefits of both worlds: speed and scalability on the one hand, trust and humanity on the other.

AI in customer contact is maturing, but the path is not linear. At one moment, expectations are sky-high; the next, we see disappointment as projects fail. That’s natural: AI is not a plug-and-play solution you simply switch on. It requires conscious choices, a solid foundation of data and processes and above all a willingness to let technology and people work hand in hand.

The promise of AI lies in speed, scalability, and efficiency. But the real value only emerges when organizations combine that power with the human touch. Employees supported rather than overshadowed. Customers helped with empathy, not just with answers.

The coming years will reveal which organizations play this game well. They will discover that it’s not about “AI versus human,” but about AI with human. That balance is not only the key to successful projects but also to sustainable customer relationships.