A shortlist you can defend beats a long list you can only apologise for.

With the round, not the market. Cheque size, stage, sector, geography, investor type and instrument are the facts that decide whether a conversation is possible at all. Applied first, they turn tens of thousands of firms into a set small enough to read.

The common failure is the opposite order: collect names that feel prestigious or familiar, then look for a reason each one might fit. That produces a long list with no internal logic, and no way to tell a weak entry from a strong one later.

It is evidence that a firm has previously operated in the shape you are describing. It is not an endorsement, a prediction, or a signal that the firm is currently deploying. Coverage and freshness vary by record, so the facts that decide whether to spend a week on an introduction should be checked against the firm’s own source.

Treating a row as a verdict rather than as evidence is how a research tool becomes an excuse. The insights pages state the same limit about every chart in the product.

A name on its own decays quickly. Two weeks after it was saved nobody remembers which constraint put it there, so the follow-up is written from scratch and reads like it.

Saved lists, pipeline stages and activity history exist to keep the research and the follow-up in the same place. The note that explains why a firm is on the list is the note that makes the outreach specific.

No. The register is large because the market is large, not because a founder should contact all of it. Broad, undifferentiated outreach burns the list once, damages the sender’s reputation, and is the behaviour the terms of use exist to prohibit.

The corpus contains real people who did not ask to be researched. That is why personal contact data stays behind sign-in and off public pages, why profile owners can claim, correct and remove their records, and why an objection is treated as a stop rather than a suggestion.

The same discipline that makes a shortlist defensible makes a database defensible: say where a fact came from, and be willing to withdraw it.