By Sajad, Founder at Cellbot — 25 years in the tech repair industry

Published: 2 January 2025 · Editorially reviewed: 26 August 2026

An RMA, or return merchandise authorisation, is a supplier's reference and process for deciding what happens to goods a buyer wants to return. For a repair shop, an RMA should connect one failed part to its purchase, test evidence, supplier decision, physical movement and final refund, credit or replacement.

The RMA is not the customer repair job. It is a separate supplier-side control. Closing the customer job while leaving its failed screen in an unlabelled drawer hides cash, stock and quality problems.

Decision rule: quarantine the part first, read the supplier's current terms, file complete evidence within the stated route and keep the case open until both the goods and the financial outcome reconcile.

!Repair-shop RMA control loop from failed-part quarantine to supplier case, return, decision and financial reconciliation

What does RMA mean?

RMA commonly means return merchandise authorisation or return material authorisation. Oracle's current supplier-recovery documentation describes it as a transaction in which a buyer arranges to return goods to a supplier for repair, replacement, refund or credit. A supplier may use another label — returns request, warranty return, credit request or return portal — while applying the same basic control.

An RMA reference does not guarantee credit. It normally identifies an authorised case so the supplier can receive, inspect and decide the return under its terms. The shop still needs to prove which part was bought, why it is being returned, whether it meets the policy and what resolution was received.

Keep four records separate

RecordPurposeClose it when
Customer work orderAuthorises and records work on one customer deviceThe agreed job, evidence and payment outcome are complete
Customer warranty caseRecords a reported problem after repairThe customer remedy and root-cause decision are complete
Supplier RMARequests a supplier decision on a purchased partThe return and supplier resolution are confirmed
Stock and finance entryChanges usable quantity and records money owed or receivedQuantity, credit, refund or replacement has reconciled

One event may create all four records, but their identifiers and outcomes are not interchangeable. Link them rather than forcing them into one status.

Classify the failure before contacting the supplier

“Faulty” is too vague to support a good supplier decision or useful quality data. Record the observed condition without guessing at blame:

  • wrong item: the delivered SKU, variant, colour or specification differs from the order;
  • short or missing: the delivered quantity or parcel contents do not match the order;
  • transit damage: packaging or goods show damage at receipt;
  • dead on arrival: the part fails the documented pre-installation test;
  • installation-stage failure: the part fails during fitting or immediate testing;
  • early-life failure: the installed part later develops a confirmed fault;
  • customer or technician damage: evidence indicates damage outside a supplier defect claim;
  • ordered in error or unwanted: the part is not defective and is subject to a different return route; or
  • uncertain: the evidence does not yet distinguish the cause.

Do not alter the category merely to fit a preferred refund route. A supplier may test the item and reject a claim that conflicts with its condition.

Quarantine first, then preserve evidence

Remove the part from usable stock as soon as it is suspected. Attach a unique quarantine label and keep it with any required packaging or accessories. Do not let a second technician fit it to another customer device while the case is being considered.

Create an evidence packet containing only what is relevant:

  1. supplier, order number, order date and invoice line;
  2. part SKU, description, quantity and any batch or serial identifier;
  3. date received and date the problem was detected;
  4. failure category and a short factual symptom;
  5. the test performed, known-good comparison and result;
  6. clear photographs or video where they show the issue;
  7. packaging, label and parcel evidence for shortage or transit damage;
  8. linked customer job and warranty-case identifiers, without unnecessary customer data; and
  9. the supplier policy version or URL checked for the case.

Evidence should support the decision, not become a data dump. HMRC's VAT record guidance explains the wider duty to keep business and VAT records; a supplier RMA register should also link any credit note or refund to the original transaction.

Download the repair-shop RMA control register. It keeps the operational, physical and financial stages in one row while preserving their separate statuses.

Read the live supplier terms, not a remembered window

Return rules vary by supplier, reason and product. Repair Outlet's current policy distinguishes unused returns from faulty-item claims and asks customers to follow its returns process. PhoneParts publishes a different route and conditions. Those examples show why a shop should not turn “14 days” or “30 days” into a universal rule.

Before filing, confirm:

  • who is eligible to make the claim;
  • the relevant period and what starts it;
  • whether a portal, form, email or prior authorisation is required;
  • testing, packaging, accessory and label requirements;
  • whether opening, fitting, adhesive removal or physical damage affects eligibility;
  • who pays or arranges carriage;
  • the return-to address and dispatch deadline;
  • possible outcomes and any inspection charge; and
  • what evidence is needed for shortage, damage or fault.

Save the policy URL and the date checked. Do not copy a supplier's rules into a static staff sheet and assume they remain current.

Use a six-stage RMA process

1. Open and triage

Create the RMA record on the day the issue is identified. Assign an owner, quarantine location and next-action date. Link the purchase line and any customer case.

Decide whether the supplier-return route is appropriate. A dangerous or recalled product may require a safety process instead. GOV.UK's product-safety guidance tells businesses to assess risks and take appropriate corrective action; do not send a suspect battery through ordinary post merely because a returns label exists.

2. Test once, consistently

Use a test appropriate to that part and failure mode. Record the method and result. For a display, that might include visual inspection, connector condition, test-device identity, touch zones, brightness and image stability. Avoid repeated fitting that can create new damage and destroy the original evidence.

3. Submit a complete case

Use the supplier's stated channel. Include the purchase reference, SKU, quantity, concise symptom, test result and requested evidence. Ask for the case reference, permitted resolution and instructions rather than assuming that approval has been granted.

4. Dispatch only as instructed

Match the physical item to the case before sealing the parcel. Record quantity, condition, package reference, carrier and tracking. Royal Mail publishes prohibited and restricted-items rules, including battery conditions; check the selected carrier and service for the actual contents.

Do not combine cases unless the supplier permits it. If several items share one parcel, retain a parcel manifest showing every RMA reference and item.

5. Record the supplier decision

Possible outcomes include accepted refund, account credit, replacement, repair, partial credit, no-fault-found or rejection. Capture the decision date, reason and supporting document. A replacement is incoming stock, not cash recovered; a credit note is not complete until accounts can identify and use it.

If the supplier rejects the case, record why. Repeat rejection reasons may reveal a technician, packaging, purchasing or supplier-quality problem.

6. Reconcile and close

Close only when all applicable controls agree:

  • the quarantined part has left or been disposed of as instructed;
  • the carrier outcome is known;
  • the supplier has issued its decision;
  • the refund, credit or replacement has been received and matched;
  • usable and quarantined stock quantities are correct; and
  • the linked customer warranty case has its own outcome.

An “approved” portal status is not a reconciled credit. Assign ageing actions so open cases cannot disappear after dispatch.

Measure recoveries without inventing a defect-rate benchmark

Supplier quality should be measured from your own purchase and return records. Compare like with like by supplier, SKU family, grade and period.

Useful measures include:

  • claim rate = units submitted for supplier review ÷ units received;
  • accepted claim rate = accepted units ÷ decided units;
  • first-response time = first supplier response − submitted time;
  • decision cycle time = decision time − submitted time;
  • recovery value = refunds + usable credits + replacement value actually received;
  • open value = claimed value not yet resolved; and
  • repeat-failure rate = replacements that later enter a new failure case ÷ replacements received.

Do not publish one industry's alleged failure percentage as your purchasing benchmark. A screen grade, supplier, model mix, installer and definition can change the result. Your own evidence is more useful.

Worked reconciliation example

Suppose a shop quarantines two £35 screens from one £350 order. The supplier accepts one as defective and rejects one for physical damage, then issues a £35 credit note.

Supplier claim | Two units submitted; one accepted, one rejected

Physical stock | Two units removed from usable stock; neither returned to sale

Finance | £35 credit received and matched to the original supplier

Customer cases | Each linked case closed independently according to its evidence

Learning | Review handling evidence for the rejected unit; do not relabel it a supplier defect

The recovery is £35, not £70. The rejected unit remains a real cost and a process signal.

Diagnose the pattern behind repeat RMAs

A high count alone does not identify the cause. Review failures in a simple sequence:

  1. receiving: was the delivered part, batch and packaging checked promptly?
  2. storage: were moisture, static, bending and bin damage controlled?
  3. pre-fit test: was the part tested before adhesive or irreversible work?
  4. installation: do technician, tool or method patterns appear?
  5. customer use: is the post-repair symptom genuinely the same failure?
  6. supplier: does one SKU, batch or grade differ from comparable purchases?

Change one controllable variable, then observe the next comparable batch. Switching supplier and installation method simultaneously makes the result harder to interpret.

The repair-shop inventory guide covers stock controls; the warranty policy template separates customer remedies from supplier recovery.

Common RMA control failures

FailureWhy it mattersMinimum correction
Faulty parts in an unlabelled drawerValue, deadlines and causes disappearUnique case, quarantine location and owner
Generic “doesn't work” noteSupplier and shop cannot reproduce the issueObserved symptom, test and result
RMA closed at dispatchRefunds and credits go unmatchedClose at physical and financial reconciliation
Customer data sent to supplierUnnecessary privacy exposureSend only identifiers and evidence needed for the part claim
Static supplier-window spreadsheetStaff use obsolete rulesLive policy URL and checked date per case
Replacement counted as cashRecovery reporting is overstatedRecord refund, credit and replacement separately
Rejected claim silently written offThe process cannot improveCoded reason and monthly pattern review

RMA FAQs

Is an RMA the same as a refund?

No. An RMA is a return-control process or reference. The eventual outcome may be a refund, account credit, replacement, repair, partial settlement or rejection.

Is an RMA the same as a customer work order?

No. The repair work-order guide covers the authorised customer job and its evidence. A supplier RMA concerns an item the shop bought and may be linked to that job.

Should a shop send a faulty phone battery back by post?

Not automatically. Check the carrier's current dangerous-goods rules, the battery's condition and the supplier instructions. Isolate and escalate damaged, swollen, leaking or overheating batteries rather than putting them into ordinary post.

When should an RMA be closed?

After the physical item and the supplier's financial or replacement outcome have both reconciled. “Sent”, “received” and “approved” are useful stages, not final proof of recovery.

Can RMA tracking stay in a spreadsheet?

Yes, if the register has one owner, controlled statuses, linked evidence, ageing and reconciliation, and the volume remains manageable. Move to a connected system when duplicate entry or missed follow-ups become the real constraint. Cellbot's current capabilities and plan limits are on the features page.

Method, sources and update record

This guide treats the RMA as a linked operational and financial control, rather than assuming one supplier policy or an industry-wide defect rate. DataForSEO's fresh UK desktop results on 26 August 2026 showed AI Overviews for both “what is an RMA” and “what is a work order”. Generic logistics definitions dominated the RMA results; maintenance and template pages dominated work-order results. Cellbot did not appear in either sampled top ten. Exa was used to inspect current supplier-policy examples and the existing Cellbot copy, not as rank evidence. DataForSEO references: keyword overview 08261636-1339-0607-0000-17d2a50ef773; RMA SERP 08261636-1339-0139-0000-076152b808d0; work-order SERP 08261636-1339-0139-0000-febd5608e1e3.

Update note: Rebuilt on 26 August 2026. Unsupported defect-rate percentages, universal return windows, revenue-loss claims, anecdotes and unverified automation claims were removed. The new version adds failure classification, evidence, supplier-policy control, physical and financial reconciliation, a worked example and an editable register.