Estimate review: a reviewer's checklist

The first check is whether each quantity on the estimate matches an actual measurement or count taken at the property, rather than a plausible-looking round number. A 12-by-14 room should show a floor quantity that follows from that measurement, adjusted for waste factor where the material calls for it, not a generic 168 square feet asserted without the dimensions behind it. Where a measurement was later revised — a wall remeasured, a room reclassified — the quantity on the line item needs to reflect the current, accepted measurement, not a stale one carried over from an earlier draft.

This check is worth doing even on lines that look routine. A waste factor applied inconsistently — 10% on one flooring line and 15% on an adjacent one with no stated reason — is a small discrepancy on its own, but it is exactly the kind of inconsistency that erodes confidence in the rest of the document once a reviewer notices it.

Unit mismatches are common and expensive: material priced per square foot appearing on a line quantified in linear feet, or a per-each item quantified as a linear run, will produce a total that is wrong by a large, non-obvious factor. Checking that the unit of measure on the line item matches the unit the price is actually quoted in is a fast, high-value check.

Duplicate line items — the same scope of work appearing under two different descriptions, sometimes from two different trades' inputs — inflate a total without any single line looking wrong on its own. The reverse error, an omitted line item for a photographed and clearly scoped area, understates the estimate just as materially and is often harder to catch because there is no line to question, only an absence.

A practical way to check for both is to walk the estimate against the room-by-room evidence in order, confirming that every photographed area of damage has exactly one corresponding set of line items, rather than reading the line items in isolation and trying to recall whether each has already been covered elsewhere.

Every priced line should show where its unit price came from — a published pricing catalog, the estimator's own prior observed data, a manual entry, or an explicitly unverified placeholder — and the date that price reflects. Construction material and labor costs move over time and by region, so a price pulled from a stale list, or from a different region's pricing, is a defensible-looking number that is nonetheless wrong for this claim.

Where a price carries low confidence or a small sample size behind it, that should be visible to the reviewer rather than presented with the same apparent authority as a well-supported price. An unpriced line item should say so explicitly — it should never silently show as $0, which reads as free rather than as unknown.

Regional variation deserves its own check: a price sourced correctly for one metro area can be materially wrong if the loss is in a different region with different labor and material costs, even when nothing else about the line item looks unusual.

Overhead and profit (O&P) is typically applied when a loss requires coordinating multiple trades, not automatically on every claim regardless of scope, and the basis it is calculated on (which categories of cost it applies to) should be stated rather than assumed. The same discipline applies to tax treatment and to which basis — replacement cost or actual cash value — a given total represents; a total that does not say which basis it uses is not yet a complete answer.

Depreciation, where it applies, should show its method (straight-line over an estimated useful life, or a specific published schedule) and the age and condition it was calculated from, not just the resulting percentage. A reviewer checking a depreciated total needs to be able to re-derive it from those inputs, not simply accept the final number.

A defensible line item can be traced back to the observation that justified it — a photo, a measurement, or a specific damage note — rather than standing alone as an assertion. A reviewer should be able to click or reference from any line item to the evidence behind it; a line with no linked evidence is a line the reviewer has no way to verify except by trusting the estimator's word.

An estimate that has been revised should carry a visible history of what changed between versions and why — a measurement correction, a scope addition after further inspection, a price update — rather than presenting only the latest total with no record of how it got there. This matters most when a total changes materially between two dated versions the reviewer might be comparing.

Finally, the math itself: quantities times unit prices should sum correctly to each subtotal, rounding should follow one consistent rule throughout (not different rounding on different lines), and the stated total should equal the sum of its parts, including any applied O&P and tax, with no unexplained gap between the line items and the number at the bottom of the page.

Estimate review, revisions, and line-item approve/reject are in production today: estimates carry a revision history, each line item carries a review state, and stale lines are blocked from silent edits. Totals are computed server-side in cents with a consistent rounding rule, and they report whether they are complete or partial and whether they are replacement-cost-only, rather than presenting a single number without that context.