The short answer
Record a return as a reported symptom before assigning a cause. Apply safety screening, identify the exact configuration and reproduce the permitted customer scenario where possible. Separate confirmed faults, setup-related findings and unresolved cases. No fault found means the evaluation did not reproduce a fault, not that the customer was wrong.
Introduction
A returned vacuum labelled does not work gives a service team very little to investigate. The appliance may fail to start, stop during a particular task, miss debris or simply behave differently from the buyer's expectation. Those are different symptoms, and treating them as one fault category can send both customer support and supplier discussions in the wrong direction.
Effective return triage records the customer's experience, screens for safety concerns, identifies the supplied configuration and evaluates the permitted scenario. It separates confirmed faults from setup-related findings and unresolved cases. No fault found should mean that the assessment did not reproduce a fault under its recorded conditions, not that the customer was mistaken.
This article proposes an intake and investigation workflow for distributors and private-label brands. It is not a report of RoboMeka return rates, a repair manual or a warranty policy. Consumer rights, commercial remedies and potentially unsafe products need the appropriate local and specialist handling; a convenient internal code does not decide those questions.
Capture the symptom before guessing the cause
Ask what happened, when it happened and what the customer was trying to do. Preserve the customer's own description alongside a structured category. A vacuum that stops only with one attachment should not immediately become a failed battery record, and a robot that cannot complete one room should not automatically become a navigation hardware fault.
Record the operating mode, relevant surface, approximate sequence and visible or audible messages. These details should be requested proportionately, without making the customer recreate a hazardous situation. If photographs or short videos are useful, explain what would help and avoid collecting unrelated household footage or personal information.
Do not promise a diagnosis from an initial symptom. A helpful first response can acknowledge the problem, state any immediate safety boundary and explain the next step. It should not tell a customer that maintenance caused the issue before the product, instructions and circumstances have been reviewed.
Keep customer evidence separate from interpretation
An intake record might say that the customer reports a stop after cleaning a rug, while the agent's preliminary hypothesis concerns a blocked floorhead. Store those as different fields. If later assessment finds another cause, the original report remains intact and the team can understand how the interpretation changed.
Put a safety screen ahead of ordinary troubleshooting
Reports of smoke, unusual heat, damaged batteries, liquid entering electrical areas, exposed wiring or other potential hazards require an appropriate escalation path. Do not continue a routine power-on script merely because the service system expects a functional test. Follow the manufacturer's instructions and qualified safety guidance for the specific situation.
The distributor needs a documented way to tell front-line staff which cases must stop normal use or testing and who takes responsibility next. The process should also address receiving and handling potentially unsafe goods. It should not rely on an ordinary retail employee improvising packaging, charging or disassembly decisions.
Keep hazardous fault recreation outside the customer workflow. Do not ask someone to defeat a sensor, open a battery, operate near an unprotected drop or touch a moving mechanism to prove a complaint. Evidence collection must remain within an approved, safe process, even when the customer wants an immediate explanation.
Identify the complete returned configuration
Record the model and available serial or revision identifiers, purchase channel and supplied accessories. For cordless products, the charger and battery belong in the compatibility picture. For robots, the dock, software version and relevant app environment may affect the symptom. An appliance body examined with an unrelated workshop accessory may not reproduce the reported condition.
Document missing items rather than silently replacing them with whatever is available. If a known-good approved accessory is used for comparison, label that intervention and retain the original configuration record. The result can then show that behavior changed with a particular substitution without pretending the returned bundle itself passed unchanged.
The replacement battery compatibility guide explains why similar appearance is not enough to justify a substitute. Use only authorized combinations and procedures. A service investigation should not create a new compatibility risk while trying to diagnose the original complaint.
Separate user-permitted checks from service work
Some symptoms have manufacturer-defined care or setup checks. Others require authorized service. Build the triage script around that distinction and the exact product instructions. An approved cleaning step for one model should not be copied to a different design merely because both products are cordless vacuums.
The Dyson V10 operating manual explains that a blockage can lead to an automatic cut-out and provides model-specific precautions for checks. This is an example of why a stopped appliance is a symptom rather than a complete diagnosis. It is not a repair procedure for RoboMeka products.
Record each permitted action and the observation before and after it. If a documented maintenance step changes the behavior, that is useful evidence. It does not automatically establish misuse or decide who pays for the remedy. Those conclusions require the relevant facts and applicable policy, not a shortcut in the technical record.
Design a triage record that supports a decision
Use fields that distinguish what was reported, what was observed and what remains unknown. A long form is not necessarily a good form. Each field should help route the case, reproduce the scenario, protect the customer or communicate a useful finding to the responsible supplier team.
| Record element | Example of useful information | What it must not imply |
|---|---|---|
| Reported symptom | Stops when a particular approved task begins | A confirmed component diagnosis |
| Configuration | Appliance, dock, accessory and visible software identifiers | That another version behaves identically |
| Safety screen | Concern recorded and case routed appropriately | Permission for improvised testing |
| Evaluation conditions | Surface, mode, starting state and permitted sequence | A complete representation of every household |
| Observation | Event reproduced, not reproduced or assessment interrupted | That an unobserved event cannot occur |
| Next action | Escalation, further evidence, approved remedy or pending decision | An automatic waiver of customer rights |
For an upcoming product range, send your after-sales and return-intake requirements. Defining the evidence, escalation roles and service boundaries before launch helps the buyer understand what support must accompany the hardware.
What does a fair reproduction attempt look like?
Reproduce the relevant permitted task using the identified configuration and a suitable controlled environment. State what could and could not be matched. A short workshop run on a clean hard floor does not fully evaluate a complaint about intermittent stopping during a different approved application.
Start with the customer's reported sequence rather than a generic success demonstration. Record whether the appliance was freshly charged, which mode was selected and whether any preparation differed. Where exact household conditions cannot be recreated, describe that limitation and choose an appropriate next investigation instead of overstating the result.
Use repeated observations where justified by the symptom, but do not invent a universal number of runs that proves reliability. Intermittent behavior may depend on conditions not yet identified. The assessment should be designed by competent personnel and remain within safe operating limits, with a clear stopping and escalation rule.
Preserve evidence before changing software or settings
Where feasible and authorized, record relevant versions, alerts and settings before an update or reset. A changed configuration may resolve the symptom, but it can also remove information about the earlier state. Explain the purpose of the action and respect account access and data-handling requirements rather than treating a reset as a harmless administrative step.
Use error codes as routing information
An error code can narrow an investigation, but its meaning belongs to a particular model and software context. Capture the exact code and accompanying message. Avoid reducing several different charging or docking messages to a single generic label simply because the customer describes all of them as not charging.
iRobot's charging-error reference distinguishes individual charging messages and routes users to relevant guidance. It illustrates the value of preserving the precise identifier. Do not transfer its code meanings to another brand, and do not assume that an app notification alone establishes which physical component has failed.
Ask the supplier for an approved symptom-to-guidance mapping that can be maintained as the product changes. The mapping should identify when front-line support can help and when technical escalation is required. It should not encourage agents to give increasingly invasive advice simply because the first permitted check did not resolve the issue.
Classify outcomes without hiding uncertainty
Use distinct outcomes for a reproduced fault, a documented setup or care finding, a missing-information case and an unresolved intermittent report. A no-fault-found result belongs to a defined assessment, not to the customer's character. Include the conditions and limitations that make the result understandable to another reviewer.
Separate the technical outcome from the commercial disposition. A replacement may be provided while an investigation remains open; a product may be held for specialist assessment before any remedy is selected. Keeping those decisions distinct prevents return statistics from confusing customer-service choices with engineering conclusions.
If a case changes category later, retain the history. A fault that appears only during an extended investigation should not be counted twice as two independent returned products. Conversely, closing an initial inquiry does not erase a subsequent recurrence. A stable case identifier helps connect the events without suppressing new evidence.
Feed useful findings back to the supplier
Escalate a concise package containing the symptom, configuration, relevant observations, permitted checks already completed and the exact question needing an answer. Large collections of unlabelled images can slow an investigation. The supplier should be able to identify the affected revision and understand why the buyer considers the issue unresolved.
For a recurring pattern, group cases by relevant identifiers before proposing a production cause. Different lots, accessories or software versions may require separate analysis. A cluster is a reason to investigate, not proof of a particular defect. The golden-sample and change-control workflow helps connect service findings to controlled product revisions.
If a corrective change is proposed, define how its effectiveness will be assessed and how it will reach affected stock or future orders. Shipment inspection can verify specified characteristics and packing changes, but cannot automatically reproduce every intermittent field condition. Keep the evidence scope visible instead of describing a passed inspection as a universal cure.
Report return data with useful denominators
A count of returned units is difficult to interpret without the population, time window and event definition. Separate units shipped to distributors from units sold to end users where the distinction matters. Products still in a warehouse have not had the same opportunity to produce a usage-related complaint as products already operating in homes.
Define whether the report counts inquiries, physical returns, unique units or repeated contacts. Explain exclusions and unresolved cases. Do not present a no-fault-found percentage as a proven measure of customer error, and do not claim a reliability improvement merely because the intake team changed its category names.
Use the findings to improve specific work: clearer setup information, a reviewed accessory, better diagnostic guidance or a product investigation. The multilingual manual approval guide provides a way to test whether recurring confusion reflects an information problem that can be corrected without blaming the reader.
Define who closes an unresolved case
An unresolved case needs an owner and a next decision, not an indefinite holding status. State what additional evidence would be useful, whether it can reasonably be obtained and when the responsible team will review the case again. The customer-service response and the technical investigation can proceed on different schedules, but their records should remain connected. Do not make a customer wait for an engineering conclusion merely because an internal report has no convenient category for uncertainty.
For a hypothetical robot that loses a room occasionally, the workshop may complete the available permitted checks without reproducing the event. A useful closing note describes that limited result, the unmatched household conditions and the agreed escalation or remedy. An unhelpful note simply says tested OK. The more precise record supports a later recurrence and allows another reviewer to understand what evidence was considered, without needing to repeat every conversation or infer that the first team disproved the complaint.
Frequently asked questions
Does no fault found mean that the appliance is definitely healthy?
No. It means that the assessment did not identify or reproduce a fault under its recorded conditions. The event may be intermittent or depend on an untested configuration. Describe the method and limitations, preserve the customer's report and decide whether further investigation is justified rather than treating the label as proof that no problem exists.
Should customers perform every check before a return is accepted?
Do not assume that they should. Checks must be safe, appropriate to the model and consistent with applicable rights and policies. Some reports require immediate escalation instead of troubleshooting. The workflow should make reasonable evidence collection possible without requiring hazardous actions, specialist tools or an unreasonable burden on the customer.
Can a distributor test with a different charger or battery?
Only use approved combinations and authorized procedures. If a permitted comparison uses another identified accessory, record it as an intervention rather than treating the result as a test of the original bundle. Similar voltage markings or connector shapes do not establish compatibility, and improvised substitutions can introduce risk or obscure the original symptom.
How should an intermittent robot complaint be handled?
Record the sequence, environment, visible messages and relevant versions without promising an immediate cause. Reproduce the permitted scenario where practical and state what could not be matched. Preserve evidence before authorized resets or updates, and escalate when necessary. A successful short run should not be presented as disproof of an intermittent customer experience.
What belongs in a supplier escalation package?
Include the identified unit and configuration, original symptom, safety-screen outcome, relevant observations, approved checks already attempted and the decision or technical answer required. Keep photographs and logs labelled and proportionate. Remove unrelated personal information, and distinguish confirmed findings from hypotheses so the supplier can investigate without inheriting an unsupported diagnosis.
Conclusion
Good triage turns a vague complaint into a clear, bounded investigation. It respects the customer, preserves uncertainty and gives support, quality and supplier teams a common record. The practical improvement is better decisions about individual cases and recurring patterns, not a cosmetic reduction in the number of complaints labelled as faults.
Send RoboMeka your proposed support workflow, including sales channels, product families, available service capability and the evidence your team needs for escalation. These requirements can then be discussed alongside product selection and spare-part planning before the range reaches customers.