ServiceNow has released new research showing that UK customers collectively lose 445 million hours a year dealing with poor customer service, despite widespread investment in AI. The findings underline a growing disconnect between what organisations think AI is delivering and what customers actually experience.
The study, The CX Shift: Customer Expectations in the AI Era, surveyed more than 34,000 executives, service agents, and consumers globally, including 2,465 respondents in the UK. It found that the average UK consumer loses 9.7 hours a year navigating slow, fragmented service experiences, equivalent to more than a full working day.
That time is not just an inconvenience. It is a productivity tax that customers increasingly refuse to pay.
When Speed Improves but Satisfaction Does Not
Nearly half of UK consumers surveyed say AI has improved customer service. Speed, efficiency, and always-on availability have all seen measurable gains. More than half of respondents say AI has improved after-hours and 24/7 support.
But the same research reveals why those gains are not translating into stronger loyalty. 51% of UK consumers cite a lack of empathy as their top frustration, while 53% say they would switch to a competitor after just one poor or slow experience.
For Shakira Talbot, Group Vice President, CRM EMEA at ServiceNow, the issue is structural rather than technological:
“The root cause isn’t a lack of AI investment. Most CRM systems were built to record interactions, not resolve them.”
The result is faster responses that still fail to feel human, connected, or accountable.
Channel behaviour exposes this contradiction. While 82% of UK customers prefer phone support, 76% attempt self-service first. Almost half say current chatbots fail to understand their questions or concerns.
Customers are not rejecting AI. They are rejecting experiences where automation cannot escalate cleanly, share context, or hand over ownership.
Asked what customers really want, Talbot emphasized, “Customers want to feel heard and resolved, not just routed.”
That distinction matters. Routing is efficient. Resolution builds trust.
The Hidden Bottleneck Inside the Contact Centre
The research shows the pressure is not only external. UK service agents spend just 49% of their working week actually addressing customer issues. The rest disappears into administration, system-hopping, and information chasing.
Nearly eight in ten agents must log into three to five systems to resolve a single issue. More than a third cite inconsistent customer data as a major challenge.
From an execution standpoint, this fragmentation undermines both AI and human performance. AI tools trained on partial data cannot act with confidence. Agents without a single view of the customer cannot show empathy at speed.
One of the most striking findings is the perception gap between customers and leadership teams. While lack of empathy tops customer frustrations, only 20% of UK executives prioritise it.
Nearly half of customers are frustrated by being transferred between departments, yet fewer than four in ten executives see this as a major issue, Talbot warned:
“When AI and human agents operate in different systems with different views of the customer, empathy breaks down.”
This misalignment explains why AI investments often deliver technical success but emotional failure.
Why CRM Must Become a System of Action
At the core of the problem is architecture. Fewer than half of UK organisations have integrated data into a single source of truth. Less than one in four have enterprise-wide AI strategies that break down departmental silos.
The research argues that CRM must evolve from a system of record into a system of action. That means unifying front office and back office workflows so AI can move beyond answering questions to actually resolving issues.
Without that shift, AI remains an interface layer. With it, AI becomes an operational engine.
The findings suggest the next phase of CX transformation will not be defined by smarter bots or faster responses. It will be defined by organisations that redesign service around ownership, continuity, and empathy.
AI can scale speed. Only connected systems can scale trust.
For UK CX leaders, the question is no longer whether AI works. It is whether their operating model allows it to work for customers.
Join the conversation
Join our LinkedIn community (40,000+ members): https://www.linkedin.com/groups/1951190/
Get the weekly rundown: Subscribe to our newsletter: http://cxtoday.com/sign-up
