The supplier looked strong before the visit. Its materials presented broad in-house capability, the factory photographs matched the address, and the quotation included the required surface finish without highlighting an external dependency.
On site, the assembly and core fabrication processes were exactly where expected. Operators were building relevant products. Records were available. There was no fake factory and no obvious misrepresentation.
The risk appeared only when we followed one product characteristic backward from final inspection.
The finish was cosmetic in the quote and functional in the product
The quotation treated the finish as a line-item specification: type, color, and unit adder. For the buyer, the finish also affected adhesion, wear, environmental resistance, and the reliability of a later assembly step. A visually acceptable part could still fail in use.
We asked to see the finishing process. The team explained that it was performed by a long-term nearby partner. Outsourcing itself was not concerning. The next questions were who approved the partner, which specification traveled with the parts, what records returned, and how the factory confirmed functional performance before assembly.
The supplier controlled the relationship more strongly than the process
The companies had worked together for years. Deliveries were fast, commercial communication was easy, and visible defects were handled promptly. This was valuable relationship control.
Process control was thinner. The purchase instruction defined the general finish, but not every critical parameter. Incoming inspection emphasized appearance and basic dimensions. Functional validation had been performed during an earlier sample phase but was not tied to routine batch evidence. A change at the external provider could therefore remain invisible until field performance shifted.
The supplier had a stable subcontractor. It did not yet have a complete control system for the risk that subcontractor owned.
We traced the boundary through five tests
| Test | What we found | Decision impact |
|---|---|---|
| Specification flow | General finish requirement transferred; critical functional criteria were incomplete | Buyer and supplier needed a controlled process specification |
| Source approval | Long commercial history; limited documented technical qualification | Relationship was credible but not sufficient evidence |
| Change control | Provider changes were unlikely but notification rules were unclear | Source and material changes required buyer approval |
| Incoming acceptance | Appearance checked; functional risk not routinely tested | Sampling and periodic validation needed redesign |
| Failure recovery | Defects could be returned quickly; root-cause ownership crossed companies | Corrective-action path needed named owners and records |
The finding did not automatically disqualify the supplier
A simplistic audit might classify the external process as a failure of vertical integration. That would miss the commercial decision. The supplier remained capable in fabrication, assembly, project communication, and final delivery. The external provider might even be technically stronger than an in-house line.
The issue was whether the lead supplier could accept responsibility for the finished outcome and build sufficient control around the external process. If it could, the networked model was workable. If it wanted the buyer to accept the risk while keeping the sub-tier invisible, the model was not.
The shortlist became conditional
The supplier stayed in consideration with four conditions:
- The critical finish specification would be rewritten around functional requirements, not appearance alone.
- The external provider and source location would be approved, with changes requiring notice.
- Incoming and periodic validation would cover the dominant failure mechanism.
- A corrective-action path would connect the buyer, lead supplier, and process provider without ambiguity.
The visit therefore changed the supplier model rather than producing a simple pass or fail. The company moved from “vertically integrated manufacturer” to “lead assembler responsible for a controlled external process.” That description was more accurate and more useful for the contract and quality plan.
The factory looked credible online because it was credible. The visit mattered because credibility at the company level did not answer control at the process level. One vertical slice exposed the difference.