A loss run is the insurance equivalent of a credit report: a documented history of every claim a policyholder has filed, including dates, amounts paid, reserves, and claim status. Carriers generate them; underwriters and agents rely on them.
For commercial insurance especially, loss runs are central to underwriting. When an agent markets a commercial account to multiple carriers, each underwriter wants loss runs — usually three to five years — to judge whether the risk is profitable. A clean loss run wins better terms; a history of frequent or large claims raises premiums or triggers declines.
The friction is that loss runs are notoriously messy. They arrive as PDFs in inconsistent formats, sometimes scanned, often requiring manual rekeying into a submission. Pulling loss runs from prior carriers can take days. This makes them a prime target for document-extraction AI: tools that ingest loss-run PDFs and output structured, comparable data eliminate hours of manual work per submission.
For agents, mastering loss runs — requesting them early, reading them critically, presenting them well to underwriters — is a core competency that AI tools increasingly augment rather than replace.