MELLORCATalk to Mellorca
Moments

Human Agency

When "The System Won't Let Me" Becomes the Answer

A sentence that ends customer conversations and quietly drains the person who has to say it. What lies behind it and how to tell a real control from a stale rule.

Mellorca Moments2026-10-10

A customer phones to ask why a follow-up call never came. The person who answers is sympathetic and knows exactly what ought to happen next. She says the words many of us have heard: "I'm sorry, the system won't let me." The customer sighs and hangs up. She sighs too. Nothing has been resolved, and nobody in the conversation made the decision that produced the result.

This piece explores that sentence and what it does to the people on both sides of it, with a focus on customer follow-up. It is a hypothesis, built from general research on working conditions and on how systems are designed to communicate. It is not a study of any particular South African business, and the example is illustrative.

A sentence that ends the conversation

"The system won't let me" works as an answer because it closes the matter. It is blameless, final and impersonal. For the employee, it is a way of surviving a situation where they cannot help. For the customer, it is a wall with no handle.

Behind the sentence is a gap between what a person is trusted to do and what the tool allows. In some businesses that gap is deliberate. A refund limit, a locked field or a mandatory status can protect against fraud, error or inconsistency. In others, it is accidental: a rule configured years ago for a situation that no longer exists, or a limitation that no one has ever thought to question.

The World Health Organization's fact sheet on mental health at work names lack of control over job design or workload, and an unclear role, among the risks to mental health at work. The UK Health and Safety Executive's control standard describes the aim that people feel able to influence how they carry out their work, and to use their abilities and initiative. A person whose standard reply is "the system won't let me" has been placed, by design or by neglect, in the opposite position.

An illustrative example

Imagine a hypothetical South African furniture retailer with an after-sales team of six. When a customer reports a damaged item, the system requires a delivery reference to open a follow-up case. Customers who bought in-store and collected the item themselves have no delivery reference. The system will not accept a blank field, and the team lead is the only person who can override it, which she can do only during office hours.

A customer calls at half past four on a Friday. The agent can see the purchase, the photograph and the customer's patience wearing thin. She cannot open the case. She says the system will not let her, promises to ask the team lead, and writes a note on a sticky label. On Monday, the note is under a keyboard. The customer calls again. The second agent says the same sentence in a gentler voice.

Both agents are competent, and nobody wrote the rule with the intention of upsetting customers. It was an ordinary design decision that left no room for a perfectly ordinary case. The cost is borne by the customer, who waits, and by the agent, who is repeatedly cast as the face of a decision she did not make. This links to a theme in the service agent who apologises for a system they cannot fix, and to the wider question of ownership in why the person closest to the problem can't fix it.

Where the dignity goes

There is more at stake than efficiency. When people are asked to carry out a customer-facing role without any ability to resolve what customers bring, they are asked to absorb frustration they cannot act on. The self-determination research summarised at selfdeterminationtheory.org describes autonomy and competence as basic psychological needs, and reports that when needs are thwarted, people are more likely to experience ill-being and poorer functioning. An agent who cannot help is denied the experience of competence in the very task she was hired for.

It would be too strong to claim that every employee who says this sentence feels diminished. Some find it a relief to have a clear boundary. The claim is narrower: where the boundary is unexplained, unreasonable or immovable, it tends to cost people their sense of being able to do the job well.

Customers, for their part, are entitled to fair dealing. The Consumer Protection Act, which the government summary describes as aiming to promote a fair marketplace and improve standards of consumer information, is not a customer-service manual. But a business that routinely answers legitimate complaints with a refusal it cannot explain is working against the spirit of that purpose.

What messages do help

When a system must refuse something, the way it refuses matters. Nielsen Norman Group's guidance on error messages advises stating the specific problem rather than a vague "an error occurred", suggesting a way forward, keeping the user's input so they do not start again, and keeping the tone neutral without blaming the user. These principles apply as well to a human agent explaining a limit as to a screen.

  • Say why. "The system needs a delivery reference to open a case" is better than "the system won't let me", because it points at something that can be changed.
  • Say what happens next. The customer should leave the conversation knowing who will act, by when and how they will hear.
  • Offer a route. If the agent cannot decide, there should be a clear escalation route, as set out in how to design better escalation paths for business systems.

Changing the answer rather than the phrasing

Better wording only goes so far. The more durable improvement is to find the rules that produce the sentence and ask whether they should exist.

  1. Collect the refusals. For a few weeks, ask the team to note every time the tool prevented a sensible action. The list is usually shorter and more specific than expected.
  2. Sort them. Some limits protect something real, such as a financial control. Some are historical accidents. Some are gaps in the process rather than the software.
  3. Move the decision to the right person. A refund limit might stay, but be raised for experienced staff, or have a fast approval route. This is the same pattern as in why approval processes become bottlenecks.
  4. Fix the configuration. Software should follow the workflow, not the other way round, a principle discussed in why software configuration should follow workflow design.

The trade-offs

  • Controls exist for reasons. Loosening a limit can open the door to errors or abuse. Some of the rules that frustrate agents are the ones that protect the business and its customers.
  • Discretion is uneven. Giving agents more latitude can mean different customers get different outcomes, which is the kind of inconsistency standardisation was meant to remove.
  • Responsibility shifts. An agent who can decide can also be blamed for a poor decision. Some staff prefer the cover of the system. Autonomy and accountability arrive together.
  • Training and support are needed. People need to know the boundaries and have somewhere to go when a decision is hard.

Interpreting the pattern

The Mellorca reading is that "the system won't let me" is sometimes an honest description of a necessary control and sometimes a symptom of a decision that nobody has revisited. The difference is worth finding out. Where it is the latter, the cost falls on a customer who is not helped and an employee who is not trusted, and the repair is usually modest: a clearer route, a sensible override or a field that accepts a case it should have anticipated.

Sources and further reading

Related Mellorca reading

Article summaryA sentence that ends customer conversations and quietly drains the person who has to say it. What lies behind it and how to tell a real control from a stale rule.
Explore our solutionsStart a conversation