What KYB Search is
KYB Search is where you run a Know Your Business check on a company: one case per company, holding everything gathered against it (the register's own record, the documents its registers will sell you, the documents you bring yourself, who ultimately owns it, what the watchlists and the news say about the people behind it) and ending in a report that states where every value came from.
The shape of a case
A case is opened by searching a business register for a company, or, where no register answers for it, by entering what you know by hand. From that moment it is a workspace of its own, organised as tabs:
| Tab | What it answers |
|---|---|
| Fields | Everything this case was opened to establish, who says so, and whether they agree |
| Prescreening | What a model found out about the company on its own, in a few minutes, with links |
| Ownership & beneficial owners | Who owns the company, drawn as the chain of stakes leading to the people at the end |
| Documents | What the company's registers will sell, and what you uploaded yourself |
| Extract | What each of those documents actually says, read with a schema, with the page it was read from |
| Screening | What the subjects are on record for: the watchlists, and the news |
| Activity | What was done on this case, and what each value was before it changed |
The Fields tab opens first, and that ordering is the point of it: it is the only tab that answers is this case done. Every other tab is somewhere you go because Fields told you to.
One idea to internalise first
A field is a question, not a place to look.
The same question, what is this company's registered address or what did it turn over, is answered by several different things: the register's own record, a document that was read, an online research pass, and you, when you open the document and type what it says. The platform asks all of them, keeps every answer, and attributes each one.
Nothing is merged and nothing is discarded. The strongest answer becomes the value; the others sit beside it saying whether they agree. A field three sources agree on is corroborated, a stronger claim than any one of them makes alone. A field two sources disagree about is a discrepancy, and that is the finding: it is lifted to the top of the Fields tab rather than left to be found by scrolling.
The precedence, which is also what a report prints:
| Rank | Source | Why there |
|---|---|---|
| 1 | Your own correction | You opened the document and said what the value is. Nothing automatic outranks that. |
| 2 | The registry | Structured, traversable, and the only source here that is accountable. |
| 3 | A document | The paper is stronger than the register; the model's reading of it is weaker, and the reading is what reaches us. |
| 4 | A web search | The widest source and the least checkable. It earns its place by answering what the others leave empty. |
Ties inside a rank break by recency, which is what makes this year's filing beat last year's without anybody having to say which is which.
Read more: The Fields dossier.
Nothing acquires anything without being asked
Almost everything you can read is free and happens on its own: searching, listing what documents a company's registers hold, watching work settle, reading anything already on the case, previewing a report.
Everything that acquires something is behind a confirmation that names exactly what it is about to do: opening a case, taking a level of an ownership walk, ordering a document, reading one with a schema, screening a subject, running a check or a research pass.
The one thing that can act without a click is the unattended ownership walk, and it is bounded by a budget you set before it starts.
Where to go next
- New here? Your first case walks the whole path once, in order, from the search box to an issued report.
- Mid-case with a question? The Platform section has one place per part of the product.
- Need a short answer? Statuses, Glossary, FAQ.