Measured in Public: A Practical Case for Publishing Assistance Service Metrics
Assistance is an unusual product. The customer buys a promise, but the promise is tested only when something goes wrong: a boiler fails in February, a flight is cancelled, a card is lost abroad. At that moment, service quality is revealed in minutes and hours, not in policy wording. Yet most UK assistance groups still publish very little about how they perform when the moment arrives. That is a missed opportunity.
The argument for voluntary disclosure is not that regulation demands it. Much of the relevant data is already reported to regulators in specific contexts, and Consumer Duty has raised expectations that firms can evidence good outcomes. But evidence submitted in private is not the same as evidence published in public. A group that chooses to publish a small, consistent set of service metrics can turn operational performance into a visible asset for customers, distribution partners and investors.
The trust gap is a reporting gap
Assistance brands ask customers to trust a service they hope never to need. That is a difficult proposition in any market, and it is harder in the UK where consumers are accustomed to comparison sites, review platforms and public service rankings. A customer considering an assistance product cannot easily judge whether a 24-hour helpline is genuinely 24 hours, whether a repair network covers their postcode, or whether a claim is likely to be paid without friction.
Investors face a similar gap. They can read the combined ratio, the retention rate and the expense base, but those numbers describe the economics of the business more than the experience of the customer. Service-level data can bridge the two. Response times, claims acceptance rates and complaint volumes are not vanity metrics. They are evidence of how the operating model behaves under stress.
What a credible disclosure set looks like
A useful disclosure set should be narrow enough to be reliable and broad enough to be meaningful. For most UK assistance groups, that suggests four families of metrics.
- Speed of response: median and 90th percentile time to answer a call, time to dispatch an engineer, and time to complete a straightforward repair or replacement.
- Claims and case outcomes: acceptance rates, average settlement time and the proportion of cases resolved without escalation. Acceptance rates need careful framing so that customers understand the difference between a valid claim and a policy exclusion.
- Complaint and redress volumes: complaints per thousand cases, the proportion upheld, and the average time to resolve. These numbers are uncomfortable to publish, which is precisely why they carry credibility when presented with context.
- Customer effort and retention: repeat contact rates, first-time-fix rates and renewal rates for the relevant cohort. These are commercial indicators as much as service indicators.
The precise definitions matter. A median response time that excludes out-of-hours calls is not wrong, but it is incomplete. A claims acceptance rate that pools all products without segmentation can mislead. The discipline of publication forces a group to settle definitions internally, which is itself a useful governance outcome.
The risks are real, but manageable
The objections are predictable. Competitors may use the data selectively. A bad quarter may invite headlines. Regulators may treat voluntary disclosure as an implicit standard. Distribution partners may ask why the metrics are not better. These are legitimate concerns, but they are not reasons for silence.
On regulation, the key is to be clear that voluntary service reporting is not a substitute for statutory reporting or for the firm's obligations under Consumer Duty. It should be presented as additional evidence, not as a regulatory filing. On competition, the risk is reduced by publishing a consistent series rather than a one-off showcase. On market expectations, the answer is to set the cadence in advance and to explain seasonality, weather events and product mix when numbers move.
The greater risk may be internal. Publishing service data exposes the gap between what the board is told and what the customer experiences. That is uncomfortable, but it is also the point. A group that cannot measure its own service performance reliably is unlikely to improve it consistently.
A board framework in five tests
Before committing to publication, directors should apply five tests.
- Materiality. Does the metric reflect something customers and investors genuinely care about, or is it easy to measure?
- Consistency. Can the group publish the same definition every period, with a clear explanation of any change?
- Context. Will the disclosure include the operating conditions, product mix and volume that make the number intelligible?
- Assurance. Has someone independent of the operational team checked the data, or at least reviewed the methodology?
- Accountability. Is a named executive responsible for the numbers and for the response when they deteriorate?
These tests do not require a large reporting function. They require discipline. A modest dashboard, published twice a year with a short narrative, can be more credible than a sprawling annual report full of selectively favourable statistics.
Start narrow, then widen
For a UK assistance group such as CPP Group plc, the sensible starting point is a small set of metrics that are already used in board reporting and that customers can understand. Publish the definitions, explain the limitations, and commit to a timetable. Then widen the set only when the data quality and the narrative are strong enough to withstand scrutiny.
The transparency dividend is not automatic. It comes from the credibility that accumulates when a group publishes the same measures through good periods and bad, and explains what it is doing about the bad. In a market where trust is the product, that discipline may prove more valuable than another marketing campaign.
