The Most Common Complaint Handling Mistakes — 8 Traps That Cost the Most

Most problems with complaints don't come from bad intentions or an incompetent team. They come from a process that took shape "somehow" back when there were five tickets a month — and was never rethought once there were fifty. Below are eight mistakes that repeat themselves most often in companies and online stores, along with what each of them actually does to your costs, your deadlines and your customer's patience.

A complaint handling mistake is rarely a single bad decision. It's a gap in the process that repeats on every ticket: no single register, no confirmation, no deadline tracking, no record of decisions. The result is always the same — the case drags on, the customer calls, and nobody can reconstruct what actually happened.

1. Tickets arrive through five channels and never meet

The most common starting point: complaints come in by email to a general address, through a contact form, by phone, via website chat, and occasionally through a sales rep who got a private message. Every channel has a different "owner" and there is no shared register — or the register is a spreadsheet updated whenever someone remembers.

On top of that comes a channel that plays by its own rules: the marketplace. Selling on Allegro or Amazon means some complaints never reach your inbox at all — they sit in the seller panel, where the platform runs its own response-time counters and its own service-quality rating. A company selling in several places at once therefore handles complaints across three different interfaces, with three different clocks, and sees the whole picture nowhere.

The consequences are predictable: some tickets are never recorded anywhere, two people reply to the same one, and when someone asks "how many open complaints do we have?", nobody knows. When an employee goes on holiday, their inbox takes a few open cases with it.

How to avoid it: one register of tickets that every complaint enters regardless of the channel it came from. The web form and the complaints mailbox should open a case automatically instead of producing yet another email to be retyped by hand.

2. The customer gets no confirmation and no case number

The customer submits a complaint and… silence. They don't know whether it arrived, who is handling it, or how to ask about it. Two days later they send the same thing again — now you have two cases instead of one. Two days after that, they call.

This mechanism shows its teeth most clearly in January, when the wave of returns and complaints from the Christmas sales peak arrives all at once. The team is the same size as it was in November, the volume is several times higher, and every unconfirmed ticket generates a second ticket and a phone call. Duplicates can grow to the point where nobody knows how many cases there really are to assess — which is exactly the moment deadlines start slipping.

This is the cheapest mistake on the list to fix, and at the same time the one that damages the relationship fastest. An automatic confirmation with an RMA number settles three things at once: the customer knows the ticket arrived, they have a reference to quote, and you have one identifier around which all further communication is organised.

How to avoid it: a confirmation sent automatically the moment the ticket is registered, with the case number, a short description of the next steps and a realistic note on response time.

3. Nobody tracks the deadlines — the customer does

Complaint deadlines are set by law and don't care how busy the team is. Yet in many companies the only reminder mechanism is the customer, calling to say two weeks have passed. Cases sitting "awaiting the service desk's opinion" can stand for weeks because nothing and nobody signals that the clock is running.

The consequences of missing a deadline can be serious and depend on the legal basis of the complaint — that's a subject for a separate article and a conversation with a lawyer, not for a rule of thumb on a blog. There is, however, a purely operational point worth keeping in mind: a company selling to both consumers and businesses runs two different regimes at the same time. A consumer claim about goods not conforming to the contract, a B2B complaint settled under a commercial agreement, and a warranty repair carried out by the manufacturer have different deadlines and different addressees — yet in an inbox they all look identical. Without marking the case type at registration, nobody will separate those three paths.

Operationally, then, one thing matters: the deadline has to be visible in the system from day one and tied to the right ticket type, not reconstructed from the date of an email.

How to avoid it: a deadline counter attached to the ticket, an SLA with an automatic warning before it expires, and escalation to a supervisor when a case stalls. The key point is that the reminder must arrive before the deadline, not after.

4. The form collects too little, so staff have to ask three times

The customer describes the fault in one sentence, gives no proof-of-purchase number and attaches no photo. Staff write asking for more, wait two days, get part of the information, and ask again. Before the case even reaches the service desk a week has gone — a week consumed entirely by email ping-pong.

Usually this isn't the customer's fault but the form's, which asks for a "problem description" and nothing else. A well-designed form enforces the minimum needed to assess the case: product identification, proof of purchase, a description of the fault and photos — all at once, in a single step.

There's a separate case, typical of parcel-locker-heavy markets like Poland: the customer drops the goods into a locker before writing a word about what's wrong. The parcel reaches the warehouse with no document, no description of the fault and sometimes no sender details — and lands on the "to be clarified" shelf, where it can sit for a week. The cure isn't another set of instructions on the website but reversing the order: first the ticket and the case number, then the return shipment, with the number stuck on the parcel.

How to avoid it: required fields matched to the ticket type (different for a return, different for a technical fault), a mandatory attachment on quality complaints, and a short hint next to the description field explaining what's worth writing.

5. Agreements live in people's heads and inboxes

"But we agreed with the customer that we'd send a replacement" — except the agreement was made over the phone, there's no note, and the person who took the call is off sick today. Reconstructing the case means searching inboxes, asking on the team chat and trusting someone's memory.

This one hurts twice: once in day-to-day handling, and again when a case turns into a dispute and you need to show when and what exactly the company told the customer. In B2B there's a third dimension: agreements are often struck between a sales rep and a buyer, outside the complaints team — a compensating discount or a goodwill replacement gets settled over the phone and never leaves one person's calendar. The service desk then assesses the case as if that conversation had never happened.

How to avoid it: all correspondence, attachments, status changes and internal notes in one place — on the ticket, in chronological order. A phone call leaves a trace too: a short note added to the case as soon as the handset goes down.

6. A rejection with no understandable justification

A rejected complaint with the message "mechanical damage — claim not accepted" is an almost guaranteed escalation. Not because the decision is wrong — it's often perfectly sound — but because the customer gets nothing that would let them understand it.

The difference between a rejection that closes a case and one that inflames it comes down to three elements: what exactly was found, on what basis, and what the customer can do next. The same applies to positive decisions — "complaint accepted", with no word on what happens now and when, generates another phone call.

How to avoid it: response templates with room for the specific findings from the assessment, consistent decision wording across the team, and always one sentence about the next step.

7. The customer has no idea what stage their case is at

Between accepting a complaint and issuing a decision, a couple of weeks typically pass. If the customer gets no signal in that time, they fill the gap with their own guesses — and a call to your support line. Statistically, it's exactly these "what's happening with my complaint?" calls that eat the most support time while adding nothing to the cases themselves.

This mistake is especially expensive because it scales linearly: twice the complaints means twice the status calls. The team grows and the metrics don't.

It's also worth noticing what the customer has grown used to. The same person who tracks their purchase to the hour and can see on a map which parcel locker it's sitting in suddenly gets a black box and a "please be patient" when they file a complaint. The contrast does more damage than the waiting itself — because the communication standard here was set by the courier industry, not by your competitors.

How to avoid it: an automatic notification on every meaningful status change, plus a status page where the customer checks the stage themselves using the ticket number. When a shipment is involved, the tracking number should land in the same place — and the "where is my parcel?" question disappears with it.

8. Nobody looks at the data — and the same faults keep coming back

The last mistake is the most insidious, because it doesn't hurt immediately. The company handles complaints efficiently, case after case, but nobody asks the question: which products fail most often, from which supplier, how many months after purchase, and what does it cost us? A complaint is treated as an incident to be closed, not as data.

And yet it's the only place in the company where product quality is visible through the customer's eyes. Three tickets of the same type in a month is a signal to talk to the supplier or to change the product description — not to issue a third credit note.

A concrete, endlessly repeating example: an importer brings in a batch of goods, and a few weeks later individual units start coming back with the same fault. If the causes are recorded as free text — once "doesn't work", once "failure", once "damaged while unpacking" — nobody spots the pattern and the case dissolves into a dozen separate repairs. If the cause is picked from a defined list and tied to the product code, the same situation ends with a claim against the supplier rather than a cost on your side. The same principle applies to size-related returns in fashion: a recurring "too small" on one model isn't a service problem, it's an error in the size chart on the product page.

How to avoid it: report ticket causes and categories from day one — even a simple "fault / product / month" breakdown reveals patterns within a quarter. There's one condition: the cause has to be picked from a defined list, not typed differently every time.

How a system eliminates these mistakes

The common denominator of the eight points above is that none of them is a people problem. They're all gaps in the process — and a process can be closed with a tool.

• One register instead of five channels — the complaint form and the complaints mailbox open a case automatically, so nothing is lost along the way.
• A number and a confirmation immediately — the ticket gets an identifier the moment it's registered, and the customer gets an automatic reply.
• Deadlines under control — case duration is counted by the system rather than reconstructed from an email date, so overruns can be caught before they happen.
• A complete case history — correspondence, attachments, photos and notes in one thread, available to the whole team and not just to their author.
• Repeatable responses — decision and message templates keep the wording consistent and cut the time it takes to prepare a reply.
• Visibility for the customer — status-change notifications and a status page reduce the number of "what's happening with my complaint?" calls.
• Data instead of impressions — ticket categories and causes collected systematically produce a report you can take to a supplier.

In Reklamator these map onto specific pieces of configuration: complaint forms with required fields and a GDPR clause, business rules that automate assignments and notifications, message templates and the mailbox integration that lands emails inside the case rather than beside it. If you'd rather walk through it in order, we've described it as a step-by-step process.

---

FAQ

Which mistake should I fix first if I can't fix them all at once?

The first one — a single ticket register. As long as complaints live in five channels, none of the other mistakes can be removed for good, because there's nothing solid to work on.

Does a small company need a system with only a dozen or so complaints a month?

At that volume the biggest gain isn't automation but order itself: one place, a case number and a visible deadline. That's exactly the threshold where a spreadsheet still works but is starting to drop details — and implementing a system is cheapest then, because there's no history to migrate.

How should the response deadline be counted?

The rules depend on the legal basis of the complaint and the status of the person filing it (consumer or business), so don't treat this article as legal advice. Operationally, what matters is that the date of receipt is recorded automatically and that the system flags an approaching deadline on its own.

Does the customer need a reply on every status change?

Not every one, but every meaningful one: ticket accepted, goods sent to the service desk, decision issued, return shipment dispatched. Too many notifications are as tiring as none — a sensible minimum is 3–4 messages per case.

What about complaints from Allegro and other marketplaces?

The reply to the customer still has to go through the platform's panel — there's no way around that. The register, however, can be shared: opening a case in your system as soon as a marketplace complaint comes in gives you one history, one place to report causes, and the certainty that a ticket won't be lost simply because it arrived through a different channel than the rest.

What about complaints reported over the phone?

Register them exactly like any other — a case opened by the employee during the call, with a note on what was agreed. The entry channel can be anything; the register has to be one.

How do I measure whether complaint handling is improving?

Three metrics are enough to start: average time from ticket to decision, the share of cases exceeding the deadline, and ticket volume broken down by cause. The first shows efficiency, the second risk, the third the source of the problem.

---

None of these eight mistakes needs a revolution to disappear. What it does need is a decision that complaints stop being handled "on the side", in an inbox, and get a process of their own with one register, a deadline and a history. If you'd like to see how that looks on your own tickets, start a free trial account or begin with the basics of an RMA system.