Why an evidence map matters
Remote factory visits get sold in two dishonest ways. The first oversells: "see the factory and know your supplier is safe." The second dismisses: "it is theatre, a prepared supplier shows you whatever they want." Both fail the same way — by refusing to sort what a visit shows into grades of evidence.
Sorted honestly, a live buyer-directed video visit produces strong evidence for some claims, weak-but-real evidence for others, and no evidence at all for a third group. Knowing which is which is what makes the tool useful, and this map is the sorting.
Strong: what a visit establishes well
A site exists at the stated address. The representative is physically there. For the class of fraud built on addresses that are vacant lots, residential blocks or other companies' buildings, this single fact is decisive — and it is established in the first five minutes.
Whose name is on it. Gate and building signage, photographed, checkable against the registered name. Not infallible (multi-tenant sites exist), but a committed, verifiable data point.
What was physically present that day. Machines, lines, inventory, people, your samples — at the moment of filming. Timestamped footage of six running injection presses is not a claim; it is a record.
What the supplier would and would not show. The pattern of access — offered freely, restricted with reasons, or refused — is itself hard evidence about the supplier's relationship to their own claims. It cannot be faked downward: a refusal on camera happened.
What was said, on the record. Answers given on the floor, on camera, in front of the operation being described, carry commitment that email answers do not.
Weak but real: what a visit suggests
Scale. Floor area, line count and rough headcount bound the plausible size of the operation. But visits happen on one day: a slow day understates, a rush day overstates, and the difference between "80 workers" and "150 workers" is not resolvable by camera pan.
That production is normal, not staged. Genuine ongoing production has a texture that is expensive to fake — work-in-progress at every station, materials mid-consumption, the unglamorous clutter of a real operation. But "expensive to fake" is not "impossible", and a visit scheduled a week in advance gives a motivated supplier a week. Treat tidiness with mild suspicion and coherence as the better signal.
That this factory makes your product category. Seeing your category on the lines today is good evidence of capability, weaker evidence that your order will run here — subcontracting after the visit remains invisible.
Process quality culture. ESD straps actually worn, in-process checks actually happening, work instructions at stations — visible artifacts of a quality culture. Artifacts, not verification: whether the system behind them operates is audit territory.
No evidence: what a visit cannot address
Stating these plainly is what separates a verification service from a reassurance service.
- Future performance. The visit records a day. It predicts nothing about month three of the relationship.
- Batch quality. Whether your goods meet your spec is a product-inspection question, answered by sampling actual production against the specification — a different service performed at a different time.
- Financial condition. Nothing about solvency, debt or cash is visible from a walkway.
- Regulatory compliance. Certificates on a wall are documents to verify with issuers, not compliance itself.
- Ownership and legal structure. Who owns the plant, on what terms, with what liabilities — registry and legal due-diligence questions.
- Intent. An honest-looking operation run by people planning to take deposits and underdeliver looks, on camera, like an honest operation.
Reading combinations, not items
The map's practical use is in combining grades. Some worked examples:
Strong + strong = act on it. Address confirmed, name matches, production running in your category, access freely given: the identity-fraud tail risk is substantially closed. Proceed to the next evidence class (inspection terms), not to blind trust.
Strong negative beats everything. The address is an office, or the name on the gate is a stranger's, or production access was refused wholesale: no quantity of pleasant office footage offsets it. Strong-grade findings, positive or negative, dominate weak-grade ones.
Weak evidence resolves by iteration, not squinting. Unsure if the floor you saw was staged? The answer is not re-watching the footage — it is an unannounced follow-up visit, an in-production inspection, or an audit. Weak evidence points at which stronger instrument to use next.
Absence of evidence is a finding with a grade of its own. "We did not see the warehouse" (refused, with reason recorded) is data. "We did not think to ask" is a hole in your checklist. The visit report should always distinguish the two.
The honest summary
A live video visit is the strongest cheap instrument for the questions that dominate first-order risk — does this company exist as presented, at this address, doing roughly what it claims — and close to worthless for the questions that dominate long-run supplier risk — will quality hold, will they stay solvent, will they ship on time in month twelve. Buy it for the first set. Budget inspection and audit for the second. And distrust anyone, including a vendor, who tells you one instrument covers both.