Case files
One company. One decision. The evidence that changed it.
Anonymized company assessment
Robotics · Integration risk · East China
Case 01
The robot worked. The deployment system did not.
A technically capable robotics company passed the demo. A closer review of interfaces, field ownership, spares, and commissioning showed why the buyer still should not proceed unchanged.
Anonymized product case
From impressive prototype to a pilot that could survive production
The prototype functioned. The BOM, tolerances, firmware ownership, and service plan did not. Here is how the China workstream was reframed.
Supplier selection case
Why a 70-person specialist beat a 1,200-person contract manufacturer
The larger company won the capability presentation. The specialist won the risk-adjusted production decision.
Deep analysis
Frameworks for reading a factory beyond the tour.
Not checklists for counting machines. These pieces focus on process ownership, failure behavior, commercial boundaries, and evidence quality.
Operating model
Manufacturer, integrator, or orchestrator? The ownership map that matters more than the label
A supplier can subcontract and still be excellent. The question is who controls each critical process when something changes.
Commercial analysis
A factory quote is a boundary document, not a price
A method for normalizing tooling, validation, packaging, yield, engineering changes, and field obligations before comparing bids.
Evidence design
The evidence architecture of a useful factory visit
How to connect claims to records, records to processes, and processes to people who can explain exceptions.
Failure behavior
Nine questions that reveal what happens after production goes wrong
The strongest supplier signal is not a perfect answer. It is a specific account of a real failure and the control that changed afterward.
Product & commercialization
Beyond supplier discovery
What changes when the job is not merely to buy a part, but to make a system manufacturable, deployable, and commercially supportable.
Prototype-to-pilot decisions
Separating the working demo from the product architecture that must survive sourcing and assembly.
China’s real advantage is iteration velocity
Why dense engineering and supplier networks can matter more than headline unit cost.
Arrive with a decision tree, not a factory list
A visit plan should define what each meeting must confirm, disprove, or redirect.
Regional intelligence
A city is useful only when you understand its production logic.
Dongguan systems brief
A production network, not a city of standalone factories
Tooling, secondary processes, assembly, and same-day problem solving—plus the control risks that density can hide.
Factory-reading method
Start with the process map, not the showroom
A disciplined way to spend the first 45 minutes inside an unfamiliar factory.
Have a live China question?