When business efficiency starts damaging customer trust

This is not a fresh definition of effort. What customer effort means after someone has already made contact is that work. This article is about the internal measure that can hide it. The wider pattern, in which several small misses add up while each internal number still looks fine, is how small customer failures add up.

A fast transfer can still be unfinished

Passing a customer to the right person can be the quickest route to an answer. It looks efficient on the first person's clock: the call was short, and it left their queue. For the customer it may be the start of another wait, another explanation, and no one who still has the first conversation.

Which?'s 2024 research on UK energy and broadband recorded that kind of experience directly. People described not being able to get through, calls being cut off, long waits, and being passed between departments. The research does not measure whether any department's handling time improved. The inference that a fast transfer can flatter an internal speed target, while the problem remains with the customer, is a reading of that evidence, not a figure Which? published. Why making customers chase you damages trust is where that handoff is followed from the customer's side, including the share who gave up. Those percentages do not need to be repeated here.

A closed record can leave the problem open

"Ticket closed" is a business record. It can mean the issue was solved. It can also mean the case was passed on, the customer went quiet, or the file was shut because a time limit ran out. There is no verified figure in this evidence for how often a closed case is still open for the customer. The risk is simpler than a percentage: if the thing being counted is the closure, an unresolved problem can look like success.

A failed cleanup can make the original problem worse. In experimental service research, Kenny Basso and Cristiane Pizzutti looked at a double deviation, a service failure followed by a recovery that also fails, and found that the failed recovery intensified the damage to trust. That work does not set a refund amount or a script. What a failed recovery does after the original problem is a separate question. The point for the metric is only that "closed" and "put right" are not the same event.

A harder exit can also mean fewer people join

Making it difficult to cancel can reduce the number of cancellations the business records this month. That is an attractive internal result. It can also be the reason someone never starts.

Citizens Advice published polling on 8 March 2024 from an Opinium survey of 3,000 UK adults, carried out between 26 January and 1 February 2024. Fifty-four percent said they had decided not to take out a subscription because they were worried it would be difficult to cancel. Those are self-reported answers about subscriptions, not a log of abandoned purchases, and they are not evidence about a trades firm's cancellation policy or a shop's returns desk. They do show a conflict inside one kind of business: a process that protects the current renewal count can put the next customer off. Why a customer who stays is not necessarily a customer who trusts you is the distinction between being retained and being trusted.

The same release estimated that people had spent £688 million on unused subscriptions in the previous year. That is consumer spending. Some of it is money the businesses kept, so it must not be described as revenue those firms lost. Among people who said they had accidentally taken out a subscription, 40 percent put it down to an auto-renewal they did not know about, 39 percent to a free trial they forgot to cancel, and 24 percent to thinking they were making a one-off purchase. Those reasons belong to accidental subscriptions. They are not the same finding as the 54 percent who said they stayed away.

Automation is a problem when it blocks a person

A confirmation, a reminder, or a straight answer to a routine question can remove effort for both sides. The difficulty is a process that treats "the customer did not reach a person" as a success, including when they still needed one.

On 22 September 2026, Citizens Advice published "Behind digital walls", about essential services such as energy, water, broadband, and parts of public support. The page argues that poor digital design, including looping chatbots and awkward forms, can leave people unable to get the help they need, and it calls for a practical route to a person by phone or in person, and for the ability to move between online and offline help. It does not publish the survey percentages that appeared in news reports, so those are not used here. It is also about essential services, where people may have little choice of provider. It does not describe a one-person firm's email acknowledgement, or a message that only confirms an opening time.

For a small business the useful line is narrower. Automation fits when it removes waiting on something a person does not need to judge. It works against trust when the customer has to defeat it before anyone who can finish the job will take the problem.

What to put next to the speed target

Keep the speed measure if it tells you something true. Put a second question beside it: did this customer have to contact us again about the same thing, because the last contact did not move the work? A falling handling time, a high share of conversations that never reach a person, or a falling cancellation count does not answer that. The customer who had to come back, or who never joined because leaving looked too hard, is the missing half of the number.

If those two pictures of the same week do not match, how I help is a way to look at the stretch after someone gets in touch, not only at how quickly the first reply went out.

References and further reading

1. Which? (26 September 2024), energy and broadband customer service. UK evidence that customers were passed between departments, could not get through, and faced long waits. No handling-time figure from companies is reported there.
2. Basso and Pizzutti (2016), Trust Recovery Following a Double Deviation, Journal of Service Research. Experimental finding that a failed recovery can intensify the original trust damage. Not a compensation formula.
3. Citizens Advice (8 March 2024), polling on subscription traps. Opinium survey of 3,000 UK adults. The 54 percent finding is self-reported avoidance of subscriptions. The £688 million is estimated spending on unused subscriptions, not revenue businesses lost.
4. Citizens Advice (22 September 2026), Behind digital walls. Essential services. Supports the call for a route to a person and for movement between online and offline help. News-report percentages about chatbots are not on this page and are not used.