The three parts of an accession number encode the entity that submitted the filing, the year of submission, and a sequence number for that submitter within the year. The first segment is the CIK of the filer or filing agent that transmitted the submission, the middle segment is the two-digit year, and the final segment is a sequence counter. In EDGAR URLs the hyphens are often removed, producing a continuous string of digits.
Each accession number maps to a single submission, which usually bundles a primary document together with its exhibits and, for periodic reports, the associated XBRL data. Because the accession number points to one exact filing, it is the most precise way to cite a source — far more specific than referencing a company and a year alone.
When verifying a figure, the accession number lets you navigate directly to the original document in EDGAR. GeminIQ's approach of sourcing from as-filed XBRL data means any value can ultimately be traced back to the specific filing identified by its accession number.