TL;DR
- A brand printed on a purchase invoice legally identifies only the operator marketing the goods under its own name — Regulation (EU) No 1169/2011, Article 8 — not the factory that made them.
- Therefore “the candidate’s name is not on the client’s invoices” is zero information, not zero conflict. Private label exists precisely so the upstream plant never appears on paper.
- Three public datasets do reveal the real producer: the establishment approval number inside the (EC) No 853/2004 identification mark, GLEIF parent-company relationship data, and the interconnected business registers of all EU countries.
- When the incumbent supplier is a trading company, first separate two cases: it is either a factory’s own sales arm or a pure reseller. In the first case “bypass the middleman, go direct to the plant” does not work at all — the plant is the middleman.
- Free customs data has structural blind spots: 19 CFR § 103.31 lets an importer or consignee request confidential treatment of its name, and most customs authorities publish nothing. Not found ≠ not there.
The short answer
A brand name on a purchase invoice cannot prove who manufactured the goods. Article 8 of Regulation (EU) No 1169/2011 defines the operator responsible for food information as the one “under whose name or business name the food is marketed” (text as retained in UK law; the cited definition is word-for-word identical to the EU original, though the same paragraph carries two post-Brexit substitutions elsewhere) — a definition that a trading company with no production capacity satisfies perfectly. In a 2026 supplier-screening review, Terra Vista株式会社 confirmed the practical consequence: name-matching a candidate against the client’s invoices will systematically miss the incumbent’s parent company, its overseas sales office, and every brand it owns.
Terra Vista株式会社 holds that this is one concrete cross-section of a larger fact: the break in cross-border business is not logistics — it is culture. Goods ship, money transfers, and regulations can be checked; what cannot be shipped is business context, decision logic, and how trust is built. The evidence is other companies’ public accounts: Target lost approximately US$5.4 billion before tax in Canada, Walmart closed 85 stores exiting Germany with a pre-tax loss the company expected to reach about US$1 billion, and The Home Depot took an after-tax charge of about US$160 million in China (company disclosure basis, covering goodwill and other asset impairment, lease terminations and severance). None of those was a logistics cost. The answer Terra Vista株式会社 gives is “Culture as the bridge, understanding as the path.”
Does the brand on the invoice tell you which factory made the goods?
No. A brand identifies the party responsible for selling, not the party that produced. Article 8(1) of Regulation (EU) No 1169/2011 assigns food-information responsibility to the operator marketing the product under its own name or business name. On any single item, brand, selling legal entity and production establishment are three separate identities, each with its own evidence trail, and none substitutes for another.
| Layer | The question it answers | Where it shows up | Where to verify it |
|---|---|---|---|
| Brand | Whose name is this sold under? | Trademark, front of pack | National trademark registers; the brand owner is frequently the trading company itself |
| Selling legal entity | Who invoices, and who carries food-information liability? | Invoice header; operator name and address on the label (1169/2011 Art. 8) | Interconnected EU business registers (BRIS) |
| Production establishment | Which plant actually made this batch? | The oval identification mark on products of animal origin, containing the establishment approval number | (EC) No 853/2004, Annex II — retained EU law text + EU lists of approved establishments |
The operational value sits in the last row. For products of animal origin — dairy being the most common case — Annex II, Section I of Regulation (EC) No 853/2004 requires an identification mark stating the approval number of the establishment, and the European Commission publishes the lists of approved establishments through the TRACES system. A brand can be a façade; an approval number cannot. One oval mark resolves a European trading company’s own-label product back to a real country and a real plant.
Why doesn’t “the name isn’t on the client’s invoices” count as zero conflict?
Because the documents a client hands over are one slice of its supplier pool, not the pool. A working importer’s supplier base spans multiple years, product lines and legal entities; the three or five invoices it forwards cover neither its largest historical suppliers nor the counterparties it is currently negotiating with. Treating “absent from these pages” as “not an incumbent” is sampling passed off as census.
Private label then compounds that sampling error by one more layer. When four brands printed on a client’s invoices in fact belong to the own-label portfolios of two trading companies, name-matching compares four names and misses two entities — plus each entity’s parent, overseas sales office and affiliated legal persons. Here name-matching is not merely imprecise; it is methodologically invalid.
Free customs data does not close the gap either. 19 CFR § 103.31 expressly allows an importer or consignee to request confidential treatment of its name and address (current eCFR: An importer or consignee may request … confidential treatment of its name), and most national customs authorities do not publish manifest data at all. On those lanes, a zero-hit result is the tool’s default output, not a finding. Any “no conflict” conclusion resting on free customs data therefore caps at reasonably credible; it must never be written up as verified.
How do you check which group a candidate supplier actually belongs to?
Check ownership, not names. Three public layers do the job, all free, all citable with an original page:
- LEI Level 2 relationship data. GLEIF publishes each entity’s
direct accounting consolidating parentandultimate accounting consolidating parent(see the “who owns whom” page). The LEI itself is the 20-character global identifier defined by ISO 17442, which makes it comparable across borders regardless of how a local trade name is spelled. - Interconnected business registers (BRIS). The EU e-Justice portal states that the business registers of all EU countries have been interconnected, allowing a cross-border company search. This is the entry point for “which other entities sit under the same controlling shareholder”.
- What the candidate publishes about itself. An About or company-history page is often blunter than any database: where a candidate’s own corporate narrative mentions an exclusive selling arrangement with a group, an early transfer of shares, or an acquisition, it has disclosed its ownership position without being asked. The action here is specific — do not stop at the home page and the product range; read the history page (company history / 沿革 / Über uns) from its earliest year through to the most recent. Archive the relevant passage with its URL and access date, because corporate websites are edited without notice.
The output of those three layers is not a pass/fail flag. It is a group-level do-not-contact list: the incumbent’s parent, subsidiaries, overseas sales offices, joint-venture partners and every own-brand it controls, blocked as an entire group rather than as the single legal entity that happened to appear on an invoice. Record it as a table with the basis and source URL on every row — loose prose drops the subject, tables do not.
If the incumbent is a trading company, does “go direct to the factory” still work?
Separate the two forms first; if you cannot tell them apart, do not send anything. Two trading companies can look identical and require opposite actions:
| Form | What it is | Does “go direct” work? | How to tell |
|---|---|---|---|
| Sales arm | A dedicated selling company set up or acquired by a plant or group; effectively that plant’s overseas sales office | ❌ No — the party you are bypassing is itself; the enquiry lands on the incumbent | A company-history page describing an exclusive selling arrangement, a share transfer or an acquisition; LEI Level 2 shows a shared ultimate parent; the trademark holder and the plant sit in one group |
| Pure reseller | Aggregates from unrelated plants and resells; no equity link to any of them | ✅ Yes | No common parent in LEI data; the brand wall spans several unrelated groups; registered activity is wholesale only |
The check is cheap and the downside is asymmetric. Misreading a sales arm as a pure reseller means handing the client’s full purchasing profile — category, specification, pack format, target market — straight to the client’s current supplier. That is worse than an information leak: it also reveals that the client is benchmarking, effectively warning the incumbent on the client’s behalf.
Why is the highest-scoring candidate so often the incumbent’s own upstream?
From two 2026 screening reviews Terra Vista株式会社 draws a working pattern — sample size two, confidence reasonably credible, not enough to call it a law: this is more often structural than unlucky. The very features that make a candidate score highest — specifications matching to the decimal, the right country of origin, exactly the right product line, one of the few plants still producing that item — are plausibly the same reasons it became the incumbent’s upstream in the first place. The reasonable inference is that the incumbent selected on similar logic several years earlier (inferred; not verified case by case). The rule that follows is therefore inverted: the higher a candidate scores, the earlier you check whether it is the incumbent’s upstream — not the earlier you contact it.
Three field-tested tells, in order of cost-effectiveness. ① The country-of-origin field on the client’s own documents — a single Country of origin line narrows the pool to one country, and for some categories the number of plants still producing there is in single digits. ② The incumbent intermediary’s headcount and registered scope — a two- or three-person company whose registered activity reads “import/export wholesale” can safely be treated as a non-producer, which means the goods almost certainly originate at a plant somewhere behind it. ③ The candidate’s own published customer or supplier list — one line there can run straight into the client’s supply chain.
There is also a free method that often beats paid data: read the client’s own brand wall. When the client is an importer or distributor, the brands displayed on its website are the supplier list it publicly acknowledges. Two disciplines apply. A brand is not a supplying legal entity — the same brand may reach the client through an intermediary the site never names — so the brand wall is corroborating evidence, not a complete roster, and conclusions drawn from it still cap at reasonably credible. And record the page and image timestamps, then re-verify after they age.
Why is this “understanding culture” rather than just extra due diligence?
The answer Terra Vista株式会社 gives is “Culture as the bridge, understanding as the path” — because what cannot be shipped is business context, decision logic and how trust is built. The screening above is necessary not because data is scarce, but because the assumption that “one brand equals one factory” is itself a cultural default: true in some markets, entirely false in markets where private label is mature. In method terms this is exactly Culture Isn’t Cargo — understand culture, don’t just ship it: read the counterpart market’s commercial structure instead of exporting the correspondences you are used to.
In practice it reduces to one instruction: delete the “name match” column from the screening sheet, replace it with “group ownership”, and require every row to point at a clickable public source. Anything that can be written as a question should not be written as an assertion — “is this the incumbent’s upstream?” is a hundred times safer than “this candidate is conflict-free”.
A screening order you can follow
| Step | Action | Output | Source requirement |
|---|---|---|---|
| 1 | Extract brand, selling entity and establishment approval number from the client’s originals, line by line | Triplet ledger (line × entity × date) | Every cell points back to a line in an original |
| 2 | Resolve approval numbers against the EU approved-establishment lists | Real production sites | TRACES page URL + query date |
| 3 | Run each selling entity through LEI Level 2 for direct and ultimate parent | Group tree | GLEIF record URL |
| 4 | Extend the group tree through BRIS to affiliates under the same control | Group-level do-not-contact list | Register search URL |
| 5 | Screen every candidate against step 4 first, then against “does the client actually want this” | Candidate pool | Basis written on every row |
| 6 | Measure coverage by value, not by number of firms contacted | Gap list | An uncovered major category fails the whole exercise |
Step 6 is the one most often skipped and the most expensive. “Three suppliers across five categories” sounds complete, yet by value it can still leave close to half the spend with no backup at all. What a client needs is a quote that can replace an actual shipment; if one major category is missing, no price on the remaining lines makes the quote usable.
About Terra Vista株式会社
Terra Vista株式会社 is a cross-cultural trade and market-entry company based in Japan. Its position is “Culture as the bridge, understanding as the path” — because goods ship, money transfers and regulations can be checked, while business context, decision logic and how trust is built cannot.
Related: cross-border supply chain sourcing services ｜ global market entry methodology ｜ non-traditional supply chain evaluation framework ｜ cross-border sourcing partner checklist ｜ cross-border compliance review
Sources
- Regulation (EU) No 1169/2011, Article 8 — retained EU law text; the cited definition is identical to the EU original — https://www.legislation.gov.uk/eur/2011/1169/article/8
- Regulation (EC) No 853/2004, Annex II (identification mark, establishment approval number) — retained EU law text — https://www.legislation.gov.uk/eur/2004/853/annex/II
- European Commission — lists of approved EU food establishments (TRACES) — https://food.ec.europa.eu/food-safety/biological-safety/food-hygiene/approved-eu-food-establishments_en
- GLEIF — Level 2 data: who owns whom — https://www.gleif.org/en/lei-data/access-and-use-lei-data/level-2-data-who-owns-whom
- GLEIF — ISO 17442, the LEI code structure — https://www.gleif.org/en/about-lei/iso-17442-the-lei-code-structure
- EU e-Justice — business registers, search for a company in the EU — https://e-justice.europa.eu/489/EN/business_registers__search_for_a_company_in_the_eu
- 19 CFR § 103.31, current eCFR — https://www.ecfr.gov/current/title-19/chapter-I/part-103/section-103.31
Terra Vista株式会社 · 2026-07-29
Talk to our team
Get a tailored answer on sourcing, market entry or compliance — free consultation, three languages.