A customer contacts a company about an incorrect charge. The immediate task is to investigate the account, explain the finding and provide an appropriate remedy. A second task begins when the case is resolved: understanding whether the same problem could affect someone else.
Businesses often measure the first task more clearly than the second. They know how many complaints were closed and how quickly staff responded. They may know much less about whether the product, communication or operating process that caused the problem has changed.
That gap makes complaints a strategic issue. They can reveal friction that a sales report or satisfaction score does not explain. The value comes from fair investigation and disciplined learning, rather than assuming every dissatisfied customer has identified a systemic failure or that every closed case represents an improvement.
A complaint has two different purposes
The customer needs a response to their circumstances. The business needs to understand what happened and whether its wider arrangements remain appropriate. These purposes overlap, but one should not be used to delay the other unnecessarily.
ISO 10002:2018 provides guidelines for handling product and service complaints within organisations, including analysing feedback to improve quality. Its published scope covers organisations of different sizes and sectors. The standard offers a management reference; it does not establish that a particular company's process meets all applicable legal obligations.
In an illustrative subscription business, a refund might resolve one billing error. If the error arose from unclear renewal settings, future customers may encounter it again. The remedy and the system change therefore need separate owners and completion evidence.
The reverse is also important. A planned product improvement does not settle the individual customer's position. Staff still need to investigate the facts, explain the outcome and provide relevant information about further options. Organisational learning should complement fair treatment rather than become a substitute for it.
Start with what customers can actually report
A low complaint count can indicate reliable service. It can also reflect difficulty finding the process, uncertainty about whether a concern qualifies or a belief that raising it will not help. Numbers need to be interpreted alongside the customer journey.
A business can examine where feedback enters the organisation. Concerns may appear in service calls, cancellation requests, branch conversations or messages to account managers. If only one formal channel reaches the complaints team, management may see an incomplete picture.
Accessibility matters. A process that works for a confident customer with time and digital access may be harder for someone facing illness, language barriers or an urgent financial problem. Alternative support should follow the organisation's obligations and the circumstances of its customers.
The OECD's Recommendation on Consumer Dispute Resolution and Redress, adopted in 2007, provides an international policy reference for accessible ways to resolve consumer disputes. It is a recommendation to governments and should not be presented as a single worldwide rule governing company response times.
Classification should support investigation
Labels make large volumes of information easier to manage, but they can conceal the issue if they are too broad. A category such as service delay may contain failures in staffing, missing documents, technology or communication. Each could require a different response.
An illustrative delivery company may receive complaints described as late orders. Some arise from a supplier shortage; others reflect a promised date the business could never meet. Combining them into one category can encourage a broad remedy that misses the underlying problems.
The record should distinguish the customer's description from the cause established through investigation. The first is valuable evidence of the experience. The second is an analytical conclusion that may need revision. Treating an initial label as a confirmed cause can lead the business in the wrong direction.
Consistency is important without requiring artificial certainty. Staff should be able to record that the cause remains unresolved and explain what evidence is missing. A completed field is not necessarily a reliable finding.
Root cause analysis needs a testable explanation
The UK Financial Conduct Authority's explanation of complaints root cause analysis distinguishes symptoms from underlying causes through a hypothetical case study. The distinction is useful beyond financial services, while the FCA's regulatory context remains specific to the firms and activities within its remit.
Suppose an illustrative lender finds repeated errors in letters sent after an account change. Retraining staff may appear to be an easy response. An investigation might instead show that an old template is selected automatically when a field is missing. Training alone would leave the relevant mechanism unchanged.
A useful explanation identifies what failed, why the process allowed it and what evidence would show that the correction works. The inquiry can have several contributing causes. A confusing screen, inadequate instructions and a poorly designed approval step may interact rather than provide competing explanations.
The investigation should avoid starting with a preferred answer. If every finding is recorded as human error, management should ask whether the analysis has examined the conditions in which people work. Accountability for individual actions can coexist with scrutiny of the process.
Serious cases cannot wait for a pattern
Repeated complaints can reveal a recurring issue, but repetition is not the only basis for action. A single report involving potential customer harm, safety or a significant control failure may require prompt escalation under the business's procedures and relevant obligations.
This creates a distinction between statistical significance and decision significance. A small number may be insufficient to estimate how widespread a problem is while still providing a reason to investigate immediately. Waiting for a trend could expose more customers to an avoidable risk.
The escalation route should be clear to frontline staff. They need to know when to seek specialist help and what information to preserve. A team should not have to resolve every uncertain concern before it can alert the people responsible for assessing seriousness.
Management also needs a way to revisit earlier assumptions. If a new case reveals that a previously isolated error has a broader cause, the organisation should examine the affected process and population rather than limit its attention to the latest complainant.
Good measures explain more than speed
The FCA's review of complaints and root cause analysis, updated in July 2026, discusses management information, governance and practices observed in a targeted review. Its findings are useful examples, rather than a universal performance benchmark for all businesses.
Closure time remains relevant, but it needs context. A quick answer that causes the customer to contact the company again may create more work and frustration than a carefully investigated response. A difficult case may take longer for legitimate reasons while still requiring clear communication throughout.
Businesses can examine repeat contact, reopened cases, remedy accuracy and recurrence after a claimed fix. These measures answer different questions. None should be treated as a complete account of fairness or service quality by itself.
Comparisons also need denominators. A growing business may receive more complaints while the number relative to relevant transactions declines. Conversely, a stable count can conceal deterioration if activity has fallen sharply. Product mix, channel changes and differences in severity should be considered before drawing conclusions.
External findings are evidence with boundaries
The Financial Ombudsman Service's annual complaints data and insight for 2025/26 offers a public source of information about financial-services disputes referred to that UK service. It can help firms identify issues for further examination, but it does not represent every customer interaction across the market.
An external decision or complaint category may reveal a question worth asking internally. The business should examine whether its own products, communications and procedures share the relevant characteristics. A superficial similarity does not establish identical circumstances or the same outcome.
Equally, an absence of external complaints should not become proof that an internal process is effective. Customers may resolve issues directly, leave the service or never pursue a dispute. Internal evidence and external insight work best when their different scopes are understood.
For companies outside UK financial services, the broader lesson concerns independent scrutiny. Sector-specific decisions can illuminate the quality of reasoning and explanation without transferring the underlying regulatory rules to an unrelated jurisdiction or industry.
Improvement needs an accountable owner
Complaints teams can identify an issue without having authority to change the product or system responsible for it. A recurring finding needs a route to the person who can act, with a decision recorded when the proposed response is accepted, changed or rejected.
An illustrative bank may find that customers misunderstand a fee explanation. The complaints team can provide evidence, but product, legal and communications teams may need to revise the material. The action should identify the scope of the change and how the business will check customer understanding afterwards.
Completion should be defined by the intended effect. Uploading a new document does not establish that it is reaching customers or that the confusing explanation has disappeared from every channel. A proportionate follow-up can check whether the relevant customer journey has changed.
Where the business chooses not to act, the reason should remain visible. The evidence may be inconclusive or another response may be more appropriate. A documented decision allows later information to be assessed against the earlier judgment rather than treated as a completely new issue.
Technology should preserve the customer's meaning
Automation can help organise case records, find related issues and summarise recurring themes. Its usefulness depends on how accurately it preserves the circumstances and language of the complaint. A short summary can lose a qualification that changes the investigation.
An AI-generated classification should therefore be treated as an input to an accountable process. Staff need access to the underlying evidence where appropriate and a way to correct the record. The business should also consider confidentiality and access before combining sensitive case information across systems.
Testing can examine whether unusual or serious concerns are missed when they do not resemble common cases. A system that performs well on frequent categories may still require a separate escalation route for low-volume but consequential issues.
The business case should include investigation quality and correction effort. Faster classification is not necessarily an improvement if staff spend more time repairing inaccurate summaries or if the resulting management information hides an emerging problem.
Questions business leaders often ask
Do fewer complaints always mean better service
No. Counts depend on activity levels, customer access and willingness to complain as well as service quality. Several sources of evidence are needed to interpret a change.
Does resolving a complaint fix its cause
Not necessarily. A fair individual remedy and a wider process correction are separate actions. Each needs an owner and evidence that it has been completed appropriately.
Can a small company apply these ideas
Yes. The process can be proportionate to its size and risks. A clear record, an escalation route and follow-up on recurring problems can be useful without a complex reporting system.
Customer complaints become useful business intelligence when the organisation can connect experience with a supported explanation and an effective response. The customer receives a fair investigation, while management gains a clearer view of the promises and processes that need attention. The measure of progress is what changes in practice after the case is closed.
References
ISO — ISO 10002 2018 Guidelines for complaints handling in organizations
https://www.iso.org/standard/71580.html
OECD — Recommendation of the Council on Consumer Dispute Resolution and Redress — 2007
https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0356
Financial Conduct Authority — Understanding complaints root cause analysis
Financial Conduct Authority — Complaints and root cause analysis good practice and areas for improvement — updated July 2026
Financial Ombudsman Service — Annual complaints data and insight 2025 26