Before you place the next order, know exactly what you are buying.
A fixed-price, buyer-side factory audit built specifically for LED lighting suppliers. Six capability domains, evidence from the production floor, and a verdict you can act on — not a checklist with photos.
Three ways LED buyers get hurt before the first container.
These are the failure modes we see most often in European lighting procurement. If one of them sounds familiar, this service was built for you.
The certificates look perfect.
ISO 9001 from a body that cannot be found, test reports naming a different model, a quality manual written for the audit — not for production.
The first order was fine. The second wasn't.
Mid-production, a new driver series appears, the capacitor brand changes, and nobody informed you. Reorder risk lives in change control.
The last audit report told you nothing.
Equipment lists, photo albums, tick boxes. Descriptive, not judgmental. You still don't know whether to place the order.
From a factory name to a verdict, in five steps.
Everything before the audit is desk work and included. The audit itself is one day on site. The report reaches you three working days after.
Supplier Fact Sheet
Free, within 48 hours. Registration data, export footprint, litigation and penalties, certificate database checks.
Scope & fixed quote
We collect your audit requirements and return a fixed all-in price. No daily rates, no extras.
On-site audit
One day at the factory. Lead auditor with IATF 16949 background plus a lighting specialist. Six capability domains. Unannounced option available.
SAR verdict
Delivered within 3 working days. Quadrant judgment, red flags R1–R9, Risk-Equivalent Control options, order-fit profile.
Valid for 12 months
The report stays valid for a year. Re-audit within that window: −40–50%.
The whole cooperation on one page.
Lanes are the parties. Left to right is time. Dashed orange lines are money flows — note that none of them passes through our lane.
Three documents. One verdict.
No 40-page photo albums. Each document has a job to do in your buying decision.
We don’t show client names. We show the instrument.
Below is the on-site audit checklist our assessments are actually written on — an anonymized excerpt from a pure-assembly downlight audit. Capabilities 1, 3 and 5 shown; 2, 4 and 6 removed. Scroll in the frame to read it the way the factory floor saw it.
How to read the codes
Every ⚠ and ✗ carries a photo reference. Findings are written on the floor, in one line, same day.
What this is — and isn’t
The checklist structure, wording and codes are our real working document (v1.1). The findings shown are representative of a typical pure-assembly downlight audit — anonymized, not a client’s file. We issue deliverables in English; the sample keeps the working language.
What’s removed
Capabilities 2, 4 and 6 — supply chain & incoming inspection, testing & NPI, technical file & compliance. Also removed: factory identity, the scores, the quadrant verdict and the photo log. Clients receive all of it, plus the SAR report three working days after the audit.
| Case file | SAR-25■■-■■ · linked: QUO-25■■-■■ · SFS-25■■-■■ |
| Factory | [Redacted] — LED downlight pure-assembly plant, Pearl River Delta |
| Target product | LED downlights 7–12 W · pure assembly (no SMT on site) |
| Order type | Standard repeat order |
| Audit | [Redacted] · 1 day on site · announced |
| Auditors | Lead quality engineer · lighting specialist |
| Version | v1.1 — Six-Capability Base (internal working document) |
| # | Check item | How to verify | Finding | Photo |
|---|---|---|---|---|
| 1.1 | Design in-house, co-developed, or build-to-print? Design inputs (spec interpretation, DFM review) on record? | Ask the design process; pull the latest DFM / design-review record | ⚠ Build-to-print. No DFM records; drawing interpretation confirmed by e-mail only. | P01 |
| 1.2 | Driver selection logic: current margin vs rated; who decides the margin and by what rule | Driver datasheet vs actual load; name the decision-maker and the rule | N/A Selection fixed by the customer’s spec — no in-house decision to audit. | — |
| 1.3 | Thermal design: Tj / Tc determination method; design basis for the thermal path (housing, TIM, heatsink) | Thermal calculation or measurement report; check Tc-point marking | ⚠ No thermal design on site; assembly file carries no Tc-point marking. | P02 |
| 1.4 | Structural details: sealing method & gasket compression · breather · surge element · cable gland | Teardown a sample and check each: compression even? MOV rating vs driver spec | ⚠ MOV rating matches driver spec; gasket compression unchecked at any station. | P03 |
| 1.5 | Optics: lens / reflector bought-in; procurement spec bans regrind? | Check the resin UV grade spec; ask whether the purchase spec prohibits regrind | ⚠ Purchase spec does not prohibit regrind resin. | P04 |
| 1.6 | Golden-sample control: approval, sealing, validity, change trigger | Inspect the golden sample and its register: who approved? when last re-checked? | ⚠ Register exists; last re-check 14 months ago. | P05 |
| 1.7 | Sample consistency: submitted sample, FAI sample, production unit mutually verifiable | Pull one unit from each and compare key features | ✓ All three match on key features. | — |
| # | Check item | How to verify | Finding | Photo |
|---|---|---|---|---|
| 3.1 | Process flow obtained and walked end-to-end (incoming → assembly → aging → packing) | Walk the actual flow; note any step the flow chart omits | ⚠ One rework loop after aging is missing from the flow chart. | P06 |
| 3.2 | In-house vs subcontracted steps mapped (SMT · die-casting · coating · optics · assembly) | Ask, then confirm on the floor — cross-check red flag R7 | ✓ SMT and die-casting subcontracted — disclosed; consistent with R7. | — |
| 3.3 | Real line count and current utilisation vs claimed capacity | Count lines running today; ask what is idle and why | ✓ 4 lines claimed, 2 running today (~55%); idle lines awaiting season. | — |
| 3.4 | Equipment condition as found (nameplates · maintenance state · wear · dust) | Photo nameplates of key equipment; note mismatch vs claims | ⚠ One of two torque drivers past its calibration due date. | P07 |
| 3.5 | Jigs, fixtures and test stations present for the target product | Ask them to show the fixture for YOUR product, not a similar one | ⚠ Press fixture shown — shared with another model, unlabelled. | P08 |
| 3.6 | Operator skill at critical stations (soldering · sealing · TIM application) | Watch 10 minutes at each critical station; note method vs work instruction | ✓ TIM applied manually; operator method matches the work instruction. | — |
| 3.7 | Control plan / work instructions actually used on the line | Take the control plan to the line: does the operator/inspector know it? | ⚠ Inspector knows the control plan; line operators do not. | P09 |
| 3.8 | WIP and finished-goods identification and housekeeping | Photo; check labels carry model + lot | ✓ Labels carry model + lot; housekeeping orderly. | — |
| 3.9 | Traceability test: finished unit → component lots (record the depth achieved) | Pick a finished unit at random and trace back; record where the chain breaks | ⚠ Unit → lot traces; chain breaks at the bought-in light engine (see Capability 2). | P10 |
| 3.10 | Best batch size (their sweet spot, from the floor — not their claim) | Ask current running batch size and changeover frequency | ✓ Sweet spot 500–2,000 pcs; changeover ≈ 40 min. | — |
| # | Check item | How to verify | Finding | Photo |
|---|---|---|---|---|
| 5.1 | ECN process in use: change request → review → verification → notification → release record | Pull the last 2 ECNs; check every step has a record and notification actually went out | ⚠ One of the last two ECNs lacks the customer-notification record. | P11 |
| 5.2 | Substitution control: AVL management and change approval | Trace one recent substitution case: AVL update, verification test, customer confirmation | ⚠ AVL updated after the new part was already in use. | P12 |
| 5.3 | Change vs compliance: how a component swap is kept from breaking RoHS / certification | Check the compliance re-check in the latest change record | ✗ No compliance re-check on record at the last component swap. | P13 |
| 5.4 | One 8D traced end-to-end (containment → root cause → verification → closure) | Pick a real customer complaint; verify the fix is verifiable, not a promise | ⚠ Containment done; root cause stops at “operator error”. | P14 |
| 5.5 | Complaint register: intake, statistics, closure rate, repeat-issue identification | Is the register alive — statistics and reviews, or just a log? | ✓ Register alive — monthly statistics and review. | — |
| 5.6 | Field-failure sample retention and analysis capability | Ask whether failed units are kept for analysis; see the tools and records | ⚠ Failed units kept in one box; no analysis tools or records. | P15 |
| 5.7 | Post-change batch segregation from pre-change batches | Check warehouse and WIP: can pre/post-change batches be told apart at a glance? | ✓ Pre/post-change batches distinguishable at a glance. | — |
One price, stated before we start.
The price depends on six scope factors (factory size, product range, certification status, travel distance, order type, language). You see the number before you commit.
Travel within the Pearl River Delta lighting cluster is included. No daily rates. No hidden fees.
Boundaries that make the verdict trustworthy.
A report is only as good as the conflicts it refuses. These refusals are contractual, not stylistic.
Start with the free Supplier Fact Sheet.
Ask about a factory you are about to order from.
One factory name is enough to start. Tell us what you buy and what worries you — the reply tells you what the assessment would cover and what it costs.
Supplier Assessment
One form for this service. The engineer who would run the case reads every message — no autoresponder, no newsletter, no sales sequence.
- 01First reply within 24 hours, Mon–Fri.
- 02Scope and a fixed price before any work starts.
- 03Your details stay yours — GDPR-compliant, deletable on request.