Documentation and evidence standards for a defensible claim file

A defensible claim file is built from several distinct kinds of records, each supporting a different part of the estimate. Dated photographs establish what was observed and when, tied to a specific room, elevation, or component rather than filed as a generic batch. Measurements — room dimensions, roof squares, moisture readings — establish the quantities used in the scope. Invoices and receipts, where repairs have already occurred or materials have already been purchased, establish actual incurred cost as distinct from an estimated cost. The scope itself, with units of measure on every line item, establishes what work was determined necessary. And a stated pricing source for each line item establishes where the dollar figure attached to that quantity came from.

None of these records substitutes for another. A photo without a corresponding measurement documents that damage exists but not its extent; a measurement without a photo documents a quantity with no visual support for what was measured; a scope line item without either documents a conclusion with nothing underneath it.

The core discipline of a defensible file is keeping each link in this chain traceable to the one before it. A photo supports an observation (this drywall shows water staining consistent with a ceiling leak). The observation supports a scope line item (replace 40 SF of drywall ceiling). The line item carries a quantity derived from a measurement and a price derived from a stated source. When every link is present and traceable, a reviewer questioning the final total can walk backward through the chain to the original photograph and check each step independently, rather than having to trust the total on faith.

A break anywhere in this chain weakens the entire line item that depends on it. A correctly priced line item with no supporting photo is exactly as unverifiable as an unpriced line item with excellent photos — in both cases, a reviewer cannot confirm that the claimed work and cost correspond to actual documented damage.

Photos should be dated automatically at capture rather than relying on a manually entered date, and should be associated with a specific room or exterior location rather than left as an undifferentiated set. Measurements should record not just the final figure but how it was obtained — measured directly, derived from another recorded value, or estimated — so a reviewer can weigh its reliability appropriately. Moisture logs, for a water loss, should record date, location, material, and reading, compared against a documented dry standard for that material, not simply a final note that an area was "dried." Invoices should be matched to the specific scope line item they support, rather than filed as a lump total with no line-level correspondence to the estimate.

Pricing source should be recorded per line item — whether a figure came from a priced catalog, the estimator's own prior data, a manual entry, or an unverified source — along with enough information to identify why that source was used for that specific item. A single unexplained total price with no per-line source is much harder to defend than the same total broken into sourced, individually justified line items.

An estimate is rarely finished on the first pass — supplements, corrections, and additional findings are normal and expected. What makes revision defensible is that each version is preserved rather than overwritten: the file should show what the estimate said before a change, what it says after, and why the change was made, tied to whatever new evidence or finding prompted it. An estimate with only its current, final state gives no way to tell whether a disputed figure was the original finding or a later correction, and no way to check whether the correction was itself justified by new evidence.

This matters particularly for approved or reported estimates. Once a scope has been approved and priced, further edits should be visible as a distinct revision rather than a silent change to the approved version, so that anyone reviewing the case later can see exactly what was approved at each point and what changed afterward.

Beyond the content of each record, the metadata around it — who created it, when, and under what review state — is itself part of the defensible file. A photo's capture timestamp, a measurement's entry method, a line item's approval state, and a price's confidence and source all function as an audit trail: a record not just of what the claim file currently says, but of how it came to say that, step by step. A file that can answer both "what is being claimed" and "how do we know" for every line item is materially stronger than one that can only answer the first question.

This kind of metadata also protects against a more subtle failure: a claim file that looks complete because every category of record is present somewhere, but where nothing actually links to anything else. A folder of photos, a separate spreadsheet of measurements, and a scope typed independently of both can satisfy a checklist of "what documentation exists" while still failing the real test, which is whether a specific line item can be traced to the specific evidence that justifies it.

PaxExact enforces an explicit human approval gate before any scope becomes a priced, reportable estimate — unapproved items are excluded from totals and reports by the platform, not by policy alone. Estimates carry revisions with per-line approve/reject state and block silent edits to stale lines, and the approved case file exports as a PDF carrying evidence, scope, and pricing state together, though the export format itself is still being extended.