Field capture guide

Capture the claim so the report writes itself.

LossNarrative will not invent facts. If a measurement, count, or date isn't in your photos, voice memo, or field data, the report says [ADJUSTER TO CONFIRM] instead — which protects you, but means the quality of your report is decided on the roof, not at the keyboard. This page is the capture routine that produces a first-pass, desk-review-ready narrative. It's about 10 minutes of extra discipline per inspection.

1Before you climb: claim setup
  • Exact property address. Not approximate — the address drives the automatic weather lookup. A correct address pulls storm reports and wind data for the date of loss into your report, with sources cited.
  • Exact date of loss. Same reason. If the reported DOL doesn't match any storm activity near the property, you want to know that before you write the cause-of-loss section, not when the desk reviewer does.
  • Claim number, policyholder, carrier. Pick the carrier template that matches the file — it changes report structure and language (e.g., "the insured," "subject property," coverage-analysis subsections).
2Photo protocol: the shot list

Every photo gets an automatic caption that feeds the narrative, so shoot deliberately — each photo should show one thing clearly. Work this order and you'll have 20–40 photos, which is what desk reviews expect on a roof claim:

  1. Risk photo — front of the property with the house number visible. Establishes you were at the right address.
  2. All four elevations — north, east, south, west, even the undamaged ones. Undamaged elevations are evidence too.
  3. Per-slope test squares— chalk a 10'×10' square on every slope, mark each impact, photograph the whole square, then close-ups of representative hits. A test square on the damaged slope only is the single most common kickback trigger — document the clean slopes with a chalked, empty square.
  4. Overview → close-up pairs — for every damage item: one wide shot locating it, one close-up showing it. A close-up nobody can place on the building proves nothing.
  5. Scale reference in close-ups— tape measure, gauge card, or chalk circle with a coin. "Bruised mat" with nothing for scale is an argument; with a tape measure it's a fact.
  6. Pitch and layers — pitch-gauge photo, and an eave/rake edge shot showing the number of shingle layers.
  7. Accessories and collateral — vents, flashing, gutters, downspouts, window screens, A/C fins, fencing, soft metals. Collateral damage corroborates the storm.
  8. Interior, per room — wide shot of the room, then close-up of each stain or damage item, with a measurement in frame where possible. If the interior is clean, one wide shot per inspected room documents that too.

On a phone, use the in-app camera button — photos compress and upload from the driveway; captions are usually done before you leave.

3The voice memo: say every number out loud

The transcript is the backbone of the narrative. Talk for 60–120 seconds like you're briefing another adjuster. The discipline that matters most: numbers must be spoken to end up in the report."A bunch of hits" becomes [ADJUSTER TO CONFIRM]; "eleven hits per square on the north slope" becomes a finding. Follow this structure:

Script skeleton (adapt, don't read):

  1. Property: "Single-family ranch, architectural shingle roof, about 8 years old, roughly 28 squares, 6/12 pitch, one story."
  2. Per-slope findings, with counts: "North slope test square: 11 hits per hundred square feet, mat bruising, granule loss. East slope: 9 hits. South and west slopes: clean, zero hits in the test squares."
  3. Accessories: "Turbine vent on the north slope dented — replace. Gutters at the northeast corner dented but functional."
  4. Interior, per room, with measurements: "Master bedroom ceiling: water stain about eighteen inches diameter around the bath vent. No other interior damage. No visible mold."
  5. Cause statement: "Damage pattern is directional, consistent with the reported hail event — primary hail on north and east, secondary water intrusion through the damaged vent boot."
  6. Recommendation intent: "Recommend full replacement of north and east slopes, replace turbine vent and bath vent boot, drywall patch and paint roughly two square feet in the master bedroom."
  • Name the slope or room before each finding — "on the east slope…" — so nothing gets attributed to the wrong location.
  • Say ages, sizes, and counts as numbers: "eight years," "twenty-eight squares," "six lifted shingles."
  • State what you did not find: "no interior leaks," "south slope clean." Negative findings prevent supplement disputes.
  • You can review and correct the transcript before generating — do it; a mis-heard number propagates into the report.
4Field data: the numbers that must be exact

The claim page has a Field data card for the values that have to be exactly right: per-slope test-square counts, roof age, size in squares, pitch, stories, interior status. Anything you type there is treated as authoritative — it overrides the voice transcript if they disagree, and it drives the pre-generation completeness check. Enter 0 for a slope you inspected and found clean; leave it blank only if you genuinely didn't inspect it. A zero is evidence. A blank is a re-inspection trip.

5What gets reports kicked back

The app runs these checks before you generate and flags gaps on the claim page. The recurring desk-review kickbacks, in rough order of frequency:

  • Test square missing on one or more slopes — including the undamaged ones
  • No photos of undamaged elevations (reviewer can’t verify the directional pattern)
  • Close-ups with no overview shot or scale reference
  • Roof age, size, or pitch absent — stalls depreciation and pricing
  • Interior not addressed at all (even "inspected, no damage" needs documenting)
  • Cause-of-loss reasoning missing: findings listed but never tied to the reported event
  • Date of loss that doesn’t match any storm activity near the property

Carrier-specific checks in the app reflect common desk-review expectations — always verify against your own panel guidelines, which control.

6What LossNarrative does — and deliberately doesn't
It does
  • Caption every photo in adjuster language
  • Turn your memo + field data into the six-section narrative, in your carrier's format
  • Pull public weather records for the DOL and cite them
  • Structure the scope as an estimate-ready table you can carry into your estimating platform
  • Flag documentation gaps before you leave the site
It doesn't
  • Invent measurements, counts, dates, or causes — gaps become [ADJUSTER TO CONFIRM]
  • Write your Xactimate/Symbility estimate or line-item codes
  • Make coverage determinations
  • Sign the report — you review, edit, and sign; you are the author of record