Customer experience teams have spent years improving how they respond once a customer gets in touch. Response times are faster, routing is more accurate, and automation has taken pressure off frontline teams. What hasn’t changed to the same extent is how often customers need to get in touch in the first place.
Tackling demand before it reaches the contact centre
Take Openreach as an example. The company is in the middle of a nationwide fibre rollout, managing millions of customer appointments tied to broadband upgrades. These appointments depend on coordination between scheduling systems, field engineers, and network readiness. When something shifts in that chain, the customer is affected whether or not they have been told.
In many organisations, that situation leads to the same outcome. Customers get in touch to check whether everything is still on track, and those checks quickly build into a large share of inbound contact.
Openreach has changed how those moments are handled. Using AI, they can track what is happening across the operation and trigger updates when something moves. If an appointment needs confirming or is likely to change, the customer is contacted while there is still time to respond.
As a result, inbound contact has reduced by around a third, missed appointments have reduced, and repeat queries are less common. Customer satisfaction has improved alongside that, pointing to a clearer experience rather than simply faster responses.
Most demand is created by uncertainty, not complexity
The pattern behind this is familiar across industries. Customers often reach out because they do not have a clear view of what is happening. The issue itself is not always difficult to resolve, but the lack of information creates enough uncertainty for them to act on.
At scale, those interactions make up a significant share of overall demand. Many organisations end up spending a large portion of their time responding to queries that come up because something was not communicated early enough.
This shows up clearly in UK telecoms data. Ofcom’s complaints reporting consistently highlights issues such as service faults, delays, and poor communication as leading drivers of customer complaints, rather than complex technical issues.
The same dynamic appears in other sectors where services depend on coordination and timing. During recent airline disruption, one travel group reported a 75% increase in customer calls as passengers tried to understand what was happening with their journeys. The issue itself was already known, but without timely updates, people still needed to get in touch to make sense of it.
AI is moving into the operation, not just the interaction
A lot of the early work with AI in customer experience has focused on what happens once a conversation starts. It helps answer questions and supports agents, which has improved how those interactions are handled.
What is starting to change is where AI sits. Instead of being added at the point of contact, it is being connected to the systems that shape the experience, including scheduling, service status, and availability.
When those systems are connected, changes can be picked up and acted on as they happen. That might involve sending an update, adjusting an appointment, or confirming what happens next without waiting for the customer to ask. The interaction becomes a follow-on from that, rather than the starting point.
In most organisations, the information already exists when something shifts, but it often sits in one part of the operation and does not lead to a response elsewhere. When that link is in place, the response can deal with the situation directly rather than explain it afterwards. This can mean confirming a revised time or updating the next step without requiring the customer to get involved.
When customers do not need to check what is happening, demand becomes more stable. Fewer people are reacting to the same issue at the same time, which makes it easier to manage.
The nature of the work also changes. Less time is spent responding to repeat queries, which allows more focus on cases that do require attention.
Using AI in this way comes down to how well systems are connected. When a change is detected, there needs to be a clear path to act on it. If that link is missing, the customer ends up filling the gap by getting in touch. That is where a lot of the remaining friction sits.
Reducing demand, not just handling it
The Openreach rollout shows what happens when that gap is closed. The complexity of the operation is still there, but it is less visible to the customer.
For organisations looking at AI in customer experience, the shift is practical. It is not only about handling interactions more efficiently, but about reducing how often they are needed.
That depends on aligning the experience with what is actually happening behind the scenes, rather than explaining it after the fact.
