When a Gesture Becomes Policy: Governing Discretionary Remedies in Service Operations

Every service operation runs a second economy that never appears in the standard work instructions. It consists of the small credit applied to a bill, the fee waived without a policy reference, the replacement sent without return, the voucher offered at the end of a difficult call. These actions are usually described as goodwill. They are more accurately understood as discretionary remedies: decisions taken at the edge of policy to protect the customer relationship and keep the service moving.

Treating them as informal kindness creates two problems. First, the organisation cannot see the cost or the pattern. Second, the judgement that makes discretion useful is left ungoverned, which is a risk when fairness expectations are under scrutiny. For UK operations leads, the task is not to eliminate discretion. It is to make it visible, testable and accountable without turning every small decision into a committee.

Discretion is a control, not a courtesy

A control is any mechanism that keeps a process within acceptable bounds. Discretionary remedies do exactly that. They absorb the variation that rigid rules cannot anticipate: a vulnerable customer, a failed handover, a supplier delay, a communication that landed badly. In that sense, a goodwill payment is a pressure valve.

But pressure valves need inspection. If the same valve opens every day at the same point, it is not relieving an exceptional condition; it is compensating for a design fault. That distinction is the heart of any serious audit. The question is not simply how much was given, but where, why and in what pattern.

Map where judgement is exercised

Start by locating the decision points. In most service models, discretion sits in at least three places: the frontline adviser, the team leader or supervisor, and the complaints or resolution team. Each has a different authority limit, a different record-keeping habit and a different view of what counts as fair.

Build a simple inventory. For each channel and journey step, record who can approve what, up to which value, and on what grounds. Do not rely on the policy manual alone. Sample recent cases and compare the written authority with the actual decision. The gap between the two is often the most useful finding.

A workable audit sample might cover 30 to 50 cases across four categories: fee waivers, service credits, replacement or rework, and gesture payments. For each, capture the reason given, the value band, the approver, the channel, the customer segment and whether the issue was a repeat. Anonymise the data; the purpose is pattern detection, not performance surveillance.

Read the pattern as diagnostic evidence

The most valuable output of the audit is not a total spend figure. It is a map of friction. A cluster of credits around a particular billing step points to a defect in that step. Repeated waivers after a supplier handover point to an interface problem. Gesture payments that rise whenever a specific channel is busy suggest that the channel is failing under load.

Present the findings as a rate, not just a sum. Goodwill per hundred transactions by journey step is more revealing than a monthly total. It normalises for volume and makes comparison possible. Pair it with complaint reasons and repeat-contact rates. Where those measures move together, the discretionary remedy is probably masking a structural issue rather than resolving an individual one.

This is where the operating model lens matters. A remedy that appears generous at the customer level can be a symptom of poor design upstream. The audit should therefore ask a second question after every finding: what would have to change in the process, the information or the handover so that this remedy was less necessary?

Consumer Duty and the fairness test

For FCA-regulated firms, discretionary remedies sit close to the Consumer Duty. The Duty requires firms to act to deliver good outcomes and to pay due regard to the needs of customers, including those in vulnerable circumstances. It does not require identical treatment in every case. Fairness is not uniformity; it is justified consistency.

That distinction needs to be explicit in governance. A customer who receives a smaller remedy than another in similar circumstances may still have been treated fairly if the difference rests on relevant factors and is recorded. A customer who receives a remedy because they complained loudly is a different matter. The risk is not discretion itself but unexplained variation.

The audit should therefore test three things. Are decisions documented with a reason that a second reviewer could understand? Are authority limits applied consistently across channels? Are outcomes monitored for patterns that might indicate bias or disadvantage to particular groups, including customers in vulnerable circumstances? These are evidence questions, not just policy questions.

Govern the judgement, not the gesture

The temptation after an audit is to tighten rules until discretion disappears. That usually backfires. Frontline teams respond by finding workarounds, or by escalating every marginal case, which slows the service and pushes the judgement further away from the customer.

A better approach is to govern the conditions in which judgement is exercised. Set clear principles: remedies should be proportionate, explainable and recorded. Define value bands with named approvers. Provide a short reason taxonomy so that data can be aggregated without forcing advisers into a script. Require a second look only where the value or the pattern warrants it.

Then create a review rhythm. A monthly meeting with operations, finance, risk and complaints can examine the goodwill rate, the top three reason codes and any outliers. The meeting should have authority to commission a process fix, not merely to note the spend. That is how a discretionary remedy moves from informal kindness to an operating control with feedback.

A practical starting point

Begin with a 30-day sample and a single question: where is discretion doing work that the process should be doing? Report the answer by journey step, not by individual. Share the findings with the teams who exercise the judgement, because they usually know why the pressure valve is opening. Then agree one process change and one governance change.

The goodwill ledger is not a ledger of generosity. It is a record of where the service model meets reality. Read it carefully and it will tell you more about your operation than many of the formal metrics that sit above it.