The Proof Behind the Promise: Auditing Consumer Duty Evidence Across Service Operations

Consumer Duty has been in force since July 2023, and by 2026 the question for many UK firms in scope is no longer whether a policy position exists. It is whether the operating model can produce credible proof that good outcomes are being delivered. Board packs often show volumes, handling times and closure rates. Those measures describe activity. They do not, on their own, show that a customer understood the product, received the right support, avoided foreseeable harm or obtained a fair result. The evidence trail is the operating model artefact that connects the promise to the proof.

Begin with the outcome claim, not the process map

A useful audit starts with the claims the firm makes about customer outcomes. These may appear in the annual Consumer Duty board report, the outcomes monitoring plan or a business unit's self-assessment. Select three or four claims that matter for service operations, such as: customers in financial difficulty reach the right support route without repeating their circumstances; vulnerable customers receive adjustments that are recorded and honoured; or complaint outcomes are consistent across channels. For each claim, ask a simple question: what evidence would convince a sceptical reviewer that this was true for a representative sample of customers? That standard shifts the focus from whether a process exists to whether the record can support the assertion.

The audit should then select two or three customer journeys that carry the claim. A journey is preferable to a department because it crosses boundaries, and boundaries are where evidence most often disappears. Trace the journey backwards from the outcome, not forwards from the procedure. Start with a closed case, a resolved complaint or a completed remediation, and ask what record exists at each stage that would allow an independent reviewer to reconstruct what happened and why.

Test the trail at four seams

The first seam is intake. Evidence here includes the channel used, the customer's stated need, any vulnerability indicator and the commitments made at first contact. A common weakness is that intake records are designed for routing rather than for proving outcomes. Flags may be present but not visible downstream; commitments may sit in free-text notes that later teams do not read. The audit should check whether the record at intake is sufficient to explain why the customer was routed as they were, and whether that rationale survives into later systems.

The second seam is handover. Evidence includes the ownership transfer, the context passed on and the commitments that must travel with the case. Handover is where activity metrics can look healthy while the outcome is already at risk. A case can be closed and reopened several times, each closure counted as a success, while the customer experiences a single unresolved problem. The audit should sample handovers and test whether the receiving team could act without contacting the customer again for information already provided.

The third seam is escalation. Evidence includes the trigger for escalation, the decision taken, the time elapsed and the discretion applied. Escalation is also where decision rights become visible. If front-line staff can adjust a payment arrangement but not a remediation threshold, the evidence trail should show who decided and under what authority. Where discretion is exercised without a record, the firm cannot demonstrate consistency or fairness.

The fourth seam is oversight. Evidence includes the management information reviewed, the quality assurance sample, complaint root-cause analysis and the actions taken as a result. Oversight evidence is often the weakest because it is aggregated. A dashboard may show that 92 per cent of cases met a service level, but not whether the 8 per cent that did not were disproportionately vulnerable customers or complex cases. The audit should ask whether oversight evidence can be disaggregated to the journey level and whether it links back to the original outcome claim.

A five-step audit sequence

1. Define the claims and the evidence standard. Agree with the business owner what would count as sufficient proof for each claim. Record the standard in advance so that the audit does not become a retrospective negotiation.
2. Select journeys and sample cases. Choose journeys that cross operational boundaries. Sample closed cases, open cases, complaints and near misses. A sample of 20 cases per journey is often enough to reveal systemic gaps without becoming a statistical exercise.
3. Walk the trail backwards. For each case, reconstruct the customer's path from outcome to intake. Record where the evidence is missing, inconsistent, inaccessible or held only in someone's memory.
4. Score evidence quality. Use a simple scale: contemporaneous and attributable, partial, or absent. Distinguish between evidence that exists but is hard to retrieve and evidence that was never created. The first is a data access issue; the second is an operating model issue.
5. Map the decision rights. For each gap, identify who could change the journey design, the data capture, the handover protocol or the remediation threshold. An evidence gap without an owner becomes a permanent feature of the model.

Evidence quality: the difference between activity and outcome

Activity measures are not useless, but they answer a different question. Average handling time tells the firm how long calls take; it does not say whether the customer understood the next step. First-contact resolution tells the firm how many cases were closed at the first point of contact; it does not say whether the customer had to return through another channel. Satisfaction scores can be influenced by the moment of survey rather than the outcome received.

Outcome evidence tends to be case-based and longitudinal. It includes repeat contact rates for the same issue, the proportion of commitments honoured at the next stage, remediation completion and the customer's demonstrated understanding where relevant. The audit should test whether these indicators can be produced at journey level and whether they are reviewed alongside the activity measures. Where they cannot be produced, the firm has a design problem, not merely a reporting problem.

Decision rights that keep the trail intact

An evidence trail is only durable if someone owns it. The audit should recommend a named evidence owner for each priority journey, usually the service owner rather than the risk function alone. That owner should have the authority to change data capture at intake, handover templates, escalation logs and the review cadence for outcome indicators. The risk and compliance functions can set the standard and challenge the evidence, but they should not be the only people able to fix the model.

The audit should also test whether the outcomes review forum can make binding decisions. If the forum can note gaps but not change thresholds, journey design or resource allocation, it is an oversight ritual rather than a control. Decision rights should be explicit: who can approve a change to a vulnerability flag, who can revise a handover protocol, and who can commission a remediation exercise when the evidence shows a pattern of poor outcomes.

Field notes for the audit team

Keep the fieldwork close to the customer record. Interview front-line staff about the cases they remember, not only the cases the system flags. Ask to see the evidence a reviewer would need, not the evidence the team normally produces. Where records are held in different systems, test whether a reviewer can assemble them within a reasonable time. And record the gaps in a way that distinguishes between a one-off documentation failure and a repeatable operating model weakness.

The proof behind the promise is rarely a single document. It is a chain of records, decisions and reviews that allows the firm to say, with confidence, what happened to a customer and why. Auditing that chain is a governance exercise, but its findings usually point to service operations: intake design, handover discipline, escalation authority and the quality of oversight. Fixing those seams is how Consumer Duty evidence becomes more than an annual reporting task.