Our latest thoughts on the issues and solutions in customer experience.

AI Is Scaling Your Worst Customer Experiences. Here’s Why

Written by Mark Gould | Apr 29, 2026 3:01:04 PM

 

Over the last few weeks, I have talked with several AI companies that can automate up to 80% of customer service interactions. They promise faster responses, lower costs and fewer people required. On paper, it sounds impressive.

But as I listened, one question kept coming to mind: What happens when you scale a poor customer experience?

Because that’s exactly what many retailers are about to do.

The promise of AI in customer experience

There is no doubt that AI will transform customer experience. It already is with chatbots handling first-line queries, voice AI replacing call centre interactions and automated messaging becoming the default.

The promise is clear: faster resolution with lower operational costs and consistent responses. For leadership teams under pressure to reduce costs, it is a compelling proposition. But there is a problem.

The uncomfortable truth

Most customer experience problems are not caused by technology, they are caused by poorly designed interactions, inconsistent colleague behaviour and a lack of commercial thinking at the frontline.

The experience is broken before the technology is applied, and when you introduce AI into that environment, you don’t fix the problem, you scale it.

A simple example from the shop floor

A customer returns to a store with an item they want a refund for and the colleague checks the policy and responds: “I’m sorry, we can’t refund that without a receipt.”

This is technically correct. The company policy has been followed and the interaction is complete. But from the customer’s perspective the issue isn’t resolved, the tone feels dismissive and the outcome feels unfair. So, the customer does what customers always do next and calls customer service.

Now the interaction moves to a contact centre or chatbot. The system is efficient, the call is answered quickly, and the AI agent follows the process. The same policy is explained: “I understand your frustration, but unfortunately without a receipt we can’t process a refund.”

This is technically correct, consistent and efficient, but the problem is not solved. With AI in place, this process has become faster, more consistent and more scalable with the same response delivered across chat, email and voice at speed and scale, with perfect consistency.

That’s the problem, because what is being scaled is not a great experience; it is a rigid one. The customer doesn’t feel listened to, helped or valued, they feel blocked. Now that experience is no longer occasional, it is consistent, predictable, and everywhere.

The difference between efficiency and experience

This is where many leaders get caught. AI delivers efficiency, but efficiency is not the same as experience. Efficiency reduces cost and experience drives revenue.

In retail, revenue growth always beats cost reduction over time. The risk is that organisations optimise for the wrong outcome. They become faster, cheaper, and easier to deal with, but not better.

There were multiple moments in this example where the experience could have been recovered. The colleague could have explored alternatives, a gesture of goodwill could have been considered, or a more flexible interpretation of policy might have been applied. Not every customer should get a refund, but every customer should feel listened to and that an effort was made to help them.

That’s the difference between policy enforcement and customer experience. AI didn’t create this experience, it simply made it faster, more consistent, and harder to escape.

What AI cannot do (yet)

AI is powerful, but it has limits. It can process information, generate responses and follow logic. But it struggles with judgement, context, commercial intuition and human warmth. These are precisely the elements that differentiate great experiences from average ones.

The real risk for retail leaders

The danger is not that AI fails, it is that it works exactly as designed. Because if what it is scaled is poor tone, irrelevant responses and missed opportunities then the outcome is predictable. A faster, cheaper, consistently mediocre experience at scale.

Before investing heavily in AI, leaders should ask a different set of questions:

    1. Are our current customer interactions effective?
    2. Do our colleagues know how to create value in each interaction?
    3. Are we clear on what “good” looks like at the point of contact?

Because technology will amplify whatever answers you have today, whether they are good or bad.

A better way forward

The opportunity is not to resist AI; it is to use it intelligently. The most effective organisations will fix the experience first, define the behaviours that drive value and then use AI to scale those behaviours. Not the other way around.

A closing thought

AI will not define the future of customer experience, it will expose it. The organisations that succeed will not be the ones that automate the fastest, they will be the ones that fix the experience first, then scale it.

In the end if you don’t design your customer experience deliberately, AI will scale it accidentally, and that is a very expensive mistake.

 

At RetailCX, we specialise in helping organisations harness the power of leadership and employee engagement to enhance customer experiences. Contact us to learn how we can support your journey toward a more innovative and customer-centric future.