How quickly do UK customers expect a response?

82% expect a response within one hour. The more significant figure is that the average issue still takes three to four days to resolve, and 53% of consumers say they would switch to a competitor because of slow service.

Is a fast automated acknowledgement better than nothing?

Only if resolution follows. An instant reply sets an expectation of movement, so days of subsequent silence reads worse than a slower honest reply. If resolution takes three days, a factual update on day one is worth more than an acknowledgement in the first minute.

What should we measure instead of response time?

Time to resolution, first-contact resolution rate, handoffs per issue, and reopened issues. Response time is already solved in most businesses, so continuing to report it shows improvement while customer experience worsens.

Why do issues take days to resolve in a small business?

Almost always waiting rather than working: the issue reached the wrong person, the resolver lacked a fact held elsewhere, a decision was not requested until late, or nobody owned it. Routing, context assembly and explicit ownership address all four.

You Replied In Four Minutes And Fixed It In Four Days

82% of UK consumers expect a response from a business within one hour. The average customer issue takes three to four days to resolve, and 53% say they would switch to a competitor because of slow service. Almost half report experiencing poor customer service in the past year.

Those two numbers describe one problem. Replying fast is cheap and automatable. Resolving fast is neither. Most businesses have done the first and called it a customer service improvement.

Why a fast acknowledgement can make things worse

An instant reply starts a clock in the customer's head. "Thanks, we have received this and will look into it" sets an expectation of movement. Four days of silence after that reads as being managed rather than being helped, and it is more irritating than a slow reply would have been, because the business has demonstrated it can respond quickly when it wants to.

This is the same pattern as the trust problem with AI support. The failure is not the automation, it is automation applied to the visible, easy part of the job while the hard part stays exactly as slow as it was.

The metrics that actually matter

Metric	What it measures	Worth optimising
Time to first response	How fast you say hello	Already solved in most businesses
Time to resolution	How long the customer's problem lasts	Yes, this is the one
First-contact resolution rate	Share fixed without a handoff	Yes, the biggest lever
Number of handoffs per issue	How often the customer repeats themselves	Yes, inversely
Reopened issues	Whether "resolved" meant resolved	Yes, catches false progress

If you track only the first row, you will report an improving service while your customers experience a deteriorating one.

Where the days actually go

In small businesses the three to four days is rarely work. It is waiting:

The issue reached the wrong person and sat until someone noticed. This is the single largest cause and the most automatable.
The person who could resolve it lacked a fact held elsewhere: an order record, a delivery status, a previous conversation.
It needed a decision from someone who was not asked until day three.
Nobody owned it. Visible in any shared inbox where everyone can see a message and no one is responsible for it.

None of those are speed-of-typing problems, which is why faster replies do not touch them.

What to automate, specifically

Routing to the right person first time, based on what the issue is rather than who happens to be free. Every avoided handoff removes a wait and a repetition.
Assembling context before a human opens the ticket. Order history, previous contacts, account status. The resolver should not have to go and look.
Escalation on elapsed time, not on complaint. If an issue has not moved in 24 hours it should surface automatically. Most businesses discover a stalled ticket when the customer chases.
Status updates during the wait. If resolution genuinely takes three days, a factual update on day one is worth more than an instant acknowledgement on hour zero.
Explicit ownership. One name against every open issue, assigned automatically.

The part that should worry you most

Financially vulnerable consumers are markedly less likely to say customer service has improved, at 21% against 32% of non-vulnerable consumers, and more likely to say it has worsened, 31% against 25%.

That gap is not an accident of sampling. Customers under financial pressure tend to have the complicated problems: payment difficulties, disputed charges, cancellations, things that do not fit a standard path. Systems optimised for the easy majority route them into the slowest queue. With consumer arrears rising, that cohort is growing, and it overlaps heavily with the customers whose business you are most at risk of losing.

Where to start

Measure time to resolution for the last fifty issues, and count the handoffs. If the median is days and the handoff count is above one, the fix is routing and context, not response speed. Both are mechanical and neither requires anyone to type faster.