Which records actually decide claims?

The records that decide claims are the ones made at the time, by the people who were there, to run the job. They form a hierarchy. At the top, the contractual paperwork: instructions, confirmations of verbal instructions, notices and their correspondence, because entitlement usually turns on whether these exist and when they were sent. Beneath, the records of fact: site diaries, allocation and day sheets, delivery tickets, programme updates and photographs, proving what happened, who was where, and what it cost. Then the corroborating sources nobody thinks of as claim records, weather data, crane and hoist logs, gate records, payroll, valuable because they were kept by systems with no stake in the argument. What unites them is timing: a record made before anyone knew there would be a dispute carries weight no reconstruction can match. So keep the paperwork the contract names and the daily records that prove the facts, as the job runs, not afterwards.

Picture two subcontractors on the same delayed job, each claiming eleven weeks of prolongation at the final account. The first brings a claim narrative and a master programme marked up in hindsight. The second brings a diary naming the areas shut and the hours lost each day, a register of every late drawing with the dates it was chased and arrived, the notices sent inside their windows, and photographs of the blocked access taken on the mornings it was blocked. The first becomes an argument about whose story to believe, and settles at a heavy discount. The second barely needs arguing: the paying side reads the file, tests a sample, finds it holds, and negotiates the number rather than the facts. The delay was identical; the records decided the money.