The first 45 minutes inside an unfamiliar factory are high leverage. The visitor is alert, the supplier has the right people available, and the planned narrative has not yet consumed the schedule. Used well, this period creates a map for everything that follows.
Used poorly, it disappears into company history, customer logos, a sample-room walk, and a fast tour of machines with no connection to the buyer’s product.
Minutes 0–10: draw the process boundary
Begin with one product family close to the buyer’s requirement. Ask the team to describe the flow from incoming material to shipment. Mark which steps occur on site, at sister companies, at routine subcontractors, or only for certain product versions.
Then add the information flow. Where does the drawing enter? Who reviews it? How does a revision reach production? Where are inspection criteria created? Who can approve a deviation? A material-flow diagram without decision ownership is only half a map.
The objective is not a perfect flowchart. It is to locate the process steps and interfaces where the visit should go deeper.
Minutes 10–25: follow one vertical slice
Select one characteristic that matters to the buyer: a tolerance stack, surface finish, bond, seal, electrical test, software version, structural joint, battery traceability, or another dominant risk. Follow it through the actual process.
Look for continuity:
- Does the released requirement match the operator’s instruction?
- Is the process parameter controlled or left to experience?
- Does the fixture protect the intended datum or function?
- Can the inspection result be traced to the unit or batch?
- What happens when the result is outside the limit?
This slice is more informative than counting equipment because it tests whether the organization can translate a customer requirement into production control.
Read work in process as a live record
Work in process shows the operating factory rather than the prepared factory. Observe how material is identified, protected, queued, moved, and separated. Are accepted, waiting, rework, and rejected conditions unmistakable? Can someone explain why inventory is accumulating at a particular step?
A crowded floor is not automatically bad, and a clean floor is not automatically controlled. The question is whether the physical state agrees with the stated process. A neat rack with ambiguous status can be riskier than a busy area with clear ownership and flow.
The most useful observation is often a mismatch between the process people describe and the process the product appears to be experiencing.
Minutes 25–35: meet the owner of the exception
Sales can explain the offer. Ask who decides when the process does not meet the requirement. That person may be a quality engineer, production manager, application engineer, or owner in a smaller factory.
Use one recent example. What failed? How was affected material contained? Who decided whether to rework, scrap, use-as-is, or ask the customer? What prevented additional units from moving? What changed before production resumed?
The answer reveals authority, escalation speed, and whether the organization learns through controlled changes or relies on individual experience.
Look for the distance between promise and execution
Notice how many translations—organizational as well as linguistic—separate the commercial promise from the person performing the work. Can the engineer who made the DFM recommendation explain it on the floor? Does the production supervisor know the customer-specific requirement? Can the quality team see the specification used in the quote?
Every handoff can work. Each one needs an information path and owner. The risk grows when no one can connect the original requirement to the current operation without returning to the sales office.
Minutes 35–45: make an evidence-gap list
Do not rush to a rating. Separate what was observed from what was stated and what remains unproven.
| Status | Example | Next action |
|---|---|---|
| Observed | Current inspection record traced to a unit and station | Assess whether the test covers the buyer’s risk |
| Stated | External finishing provider is stable and approved | Review approval and incoming-control records |
| Unresolved | Who owns software changes after sample approval | Meet responsible engineer and review release path |
| Contradiction | Sales described an internal process; floor team described outsourcing | Map actual boundary and commercial responsibility |
This list should reshape the remaining agenda. Spend less time where evidence is already adequate and more where the decision still carries uncertainty.
What not to conclude too early
Do not confuse hospitality with priority, equipment with process capability, certificates with current compliance, famous customers with fit, or a fluent sales team with engineering access. Each can be a useful signal; none completes the assessment.
Likewise, do not reject a factory because one area looks imperfect. Ask whether the imperfection threatens the buyer’s dominant risks and whether the factory sees and controls it. Mature operations are not problem-free. They make problems visible and manageable.
Once the map exists, the showroom, management discussion, and detailed floor tour become more useful. They can be tested against a specific operating model instead of absorbed as isolated signals.