A good research record begins before the first summary, with a source another person can find and identify.

A folder full of documents is not yet a source register. The folder tells you where files happen to be. The register tells a colleague what each file is, which period it covers, and how it entered the research process. That distinction becomes useful the first time someone challenges a sentence in a brief.

A source register does not need an elaborate database. A small structured table can work. What matters is that its fields answer the questions a reviewer will ask, and that the references survive the handoff from the person who collected the material to the person using the conclusion.

Separate document identity from the filename

A filename can change when someone downloads a report or moves it into a shared repository. Give each source a stable internal identifier. Keep the original title, company, document type, publication date, reporting period, and public or authorized location in separate fields.

For US filings, the SEC’s EDGAR research guide provides an entry point to company disclosure documents. Preserve the filing reference rather than relying on a search query that may produce a different result later. If your team stores a permitted local copy, connect that copy to the original reference.

Keep publication date and reporting period apart

A report published in one month may describe a period that ended earlier. Those dates answer different questions. The reporting period supports a comparison of business performance; the publication date establishes when a researcher could have had the information.

Do the same for a presentation that discusses several periods. Do not assign every statement the date of the current quarter simply because it appears in a current deck. Let the claim carry its own period when that distinction matters.

Record the version you actually used

A corrected transcript, amended filing, or revised presentation can change a passage that a brief already cites. Keep a version field and a short change note. When replacing a working copy, retain enough information to explain why an earlier research note used different wording.

This is also a place to distinguish an original company document from a transcript or data product supplied by another provider. Record the provider and access conditions. A source being easy to download does not establish your right to redistribute it.

Make ownership visible

Each research source should have a clear next action: ready for review, awaiting a correction check, or superseded by a later version, for example. Assign responsibility for that action. Avoid a vague “complete” status that hides whether someone has checked the document identity, the extracted content, or both.

A compact handoff can include:

  • A stable source identifier and the original location.
  • The publication date and relevant reporting period.
  • The version used by the research note.
  • A review owner and any unresolved access or extraction issue.

Test the register with a real citation

Choose one sentence from a brief. Ask a colleague to use its reference to find the exact document and relevant passage without help from the author. If the colleague has to search a chat history or guess between nearly identical files, improve the reference before adding more automation.

The aim is a repeatable research path. Gatani’s research automation workflows can connect document intake, source organization, and the review handoff. A useful starting point is the source your team most often has to find twice.

This editorial collection was published on . Archive dates organize the series.