There’s so much noise right now around AI in customer experience, with lots of focus on faster responses, lower costs and smaller teams.
On the surface, a lot of it does make sense. After all, if AI can handle more of the workload, then surely things will be way more efficient.
What we are seeing in reality is a lot more nuanced.
Some businesses are seeing real gains, but most businesses are running into issues they didn’t expect. Such as inconsistency, gaps in information and business processes that don’t really connect.
This raises a different question altogether…
Is the problem with AI or is it what the AI is working with?
AI doesn’t create structure. It depends on it
It may seem like an obvious statement to make, but AI isn’t magic. It can’t fix disconnected systems, incomplete data or workflows that are unclear.
For most businesses, their system setup is:
- Fragmented across multiple tools
- Missing key pieces of customer context
- Built on processes that don’t quite line up
Now, let’s be honest, AI can’t fix that. It can only operate within that framework.
And because it’s faster, it tends to expose those issues more quickly and more frequently.
What this looks like in practice
On paper, adding AI sounds straightforward. In most cases, businesses just want to be able to automate customer responses, pull up information faster and reduce the pressure of teams.
Underneath all of this, the same problems will still exist. An AI agent can only respond based on the data is can access. And if the data is incomplete, the response it provides will be too. If the systems aren’t connected, then the context will get lost. And if the workflows are unclear, decisions will become inconsistent.
Then instead of improving customer experience, you just end up with:
- Faster responses that aren’t always accurate
- Decisions made without the full picture
- Journeys that still feel disjointed
The experience doesn’t necessarily get better, it just happens more quickly.
Why some businesses are seeing results (and others aren’t)
As you can maybe now tell, the difference isn’t the AI itself, but what sits underneath it.
The businesses seeing real impact from AI have usually done the work first.
They’ve:
- Connected their systems
- Centralised customer data
- Simplified how work flows across teams
So, when AI is introduced, it has something solid to build on. Because strong foundations have already been laid, the AI enhances what’s already working.
For others, AI is added on top of a setup that’s already under strain. And instead of improving performance, it highlights where things aren’t working as they should.
AI should enhance people, not work around broken systems
There’s a tendency to look at AI as a way to reduce reliance on people. But in most cases, the real opportunity is to support them.
- To remove friction.
- To surface the right information at the right time.
- To make it easier to do good work consistently.
That only works when the underlying structure supports it. Otherwise, AI becomes another layer that teams have to work around.
Before adding AI, look at how things actually work
If there’s one thing worth doing before introducing AI, it’s this:
Take a step back and look at how your operation actually works day to day. Be radically honest and ask yourself the following questions…
- Where does information live?
- How many systems are involved in one interaction?
- Where does context get lost?
- Where do delays or inconsistencies happen?
This is what AI will inherit, then it will amplify it.
The real opportunity
AI has huge potential in customer experience, that’s unquestionable.
But the businesses that get the most from it won’t be the ones who move fastest.
They’ll be the ones who build properly.
Because AI doesn’t fix bad architecture. It exposes it.
If you want to understand how AI would actually perform in your environment, start by looking at the structure underneath it.
That’s usually where the real opportunity sits.
