How to Document Hail Damage for an Insurance Estimate: The Test Square, Done Right

Most hail packets fall apart for the same reason: the crew found the damage but never proved it. They walked the roof, saw the bruising, chalked a square, snapped a dozen blurry photos, and wrote an estimate. Then a desk reviewer three states away opened the file, saw a pile of close-ups with no slope labels, no orientation, and a test square that could have been anywhere on any roof in the country, and the whole thing stalled. Not because the damage wasn't real. Because the documentation didn't carry its own weight.

The test square is the center of gravity for a hail inspection, and it is also the single most misused tool in the trade. A square done right tells a clean story: here is a defined 100-square-foot area on a labeled slope, here is the date and direction, here are the functional hits counted and circled, here is what that density means across the whole plane. A square done lazy is just chalk on shingles. The difference is not talent. It is process.

What follows is the field workflow sharp restoration crews actually run, slope by slope, photo by photo, count by count, all the way through to an Xactimate-aligned estimate that a homeowner can hand to their carrier. The goal is not to win an argument with an adjuster. The goal is to document so thoroughly that there is no argument to have. And before any of that, one hard line worth stating up front, because it protects your license and your business: you document and you estimate your own scope of work. You do not file, handle, negotiate, or interpret the homeowner's policy. The homeowner files; the insurer decides coverage. Everything below lives strictly on the document-and-estimate side of that line, and there is a full section later on exactly what you must never say.

What a test square actually is, and what it is not

A test square is a defined 100-square-foot area, traditionally a 10-foot by 10-foot box, that you mark on a single roof slope so you can count hail impacts inside a known, repeatable boundary. The point is density: how many functional hail strikes appear in 100 square feet of that plane. Because a "roofing square" is also 100 square feet, the count inside your test square scales directly to the slope. If you find eight functional hits in your square and that slope is 14 squares, you are documenting roughly 112 strikes across the plane. The math is honest and easy to follow, which is exactly why reviewers trust it when it is done correctly.

Here is what a test square is not. It is not proof on its own. It is not a magic number that forces an outcome. And it is not a substitute for understanding the difference between damage that matters and damage that doesn't. A chalk box with random circles inside it is worthless. A chalk box on a clearly labeled slope, photographed with orientation, with each circled mark verified as functional damage and counted, is one of the most persuasive pieces of evidence you can produce.

The word doing all the work in that last sentence is functional. Not every mark on a shingle is a hail hit, and not every hail hit is functional damage. Functional hail damage is impact that compromises the shingle's ability to do its job and shorten its service life: fractured or bruised mat, displaced or fractured granules exposing the asphalt, and a soft, give-when-pressed spot you can feel with your thumb. Cosmetic marks, blistering, manufacturing flaws, foot traffic scuffs, mechanical damage, and normal granule loss from age are not functional hail hits, and if you count them as such, you have handed the reviewer a reason to throw out your whole square. Discipline about what counts is what separates a credible inspector from a kid with a crayon.

The impostors: what gets miscounted as hail

More squares fail for over-counting than for under-counting, and the over-counts almost always come from the same handful of look-alikes. Learn to name each one on sight, because a reviewer who finds one of these circled as a "hit" will assume the rest of your circles are just as loose:

• Blistering. Round pops in the surface caused by trapped moisture or volatiles off-gassing from the asphalt. They look like impacts but they open outward and lack the bruised, depressed mat underneath. Blisters are scattered without directional bias and often appear on slopes the storm never touched.
• Granule loss from age. A worn roof sheds granules everywhere, fairly evenly, with no circular impact pattern and no soft spot beneath. Even wear is age. Circular wear with a give underneath is impact.
• Foot traffic. Scuffs and granule displacement that track in lines along the natural walk paths, around penetrations, and up the gentlest route to the ridge. Linear and clustered on paths means feet, not hail.
• Mechanical and tree damage. Gouges, scrapes, branch rub, and tool marks. Irregular shapes, often with directional dragging, never the clean round signature of a stone.
• Manufacturing defects. Craze cracking, factory edge issues, and shading that came off the line that way. These follow the manufacturing pattern, not the storm.
• Shiners and exposed nail heads. Sometimes mistaken for impacts in a photo. They are fastener-related, not weather.

When you photograph a square, it is worth deliberately shooting one or two clear examples of the impostors you ruled out, labeled as such. Showing a reviewer "here is a blister, which we did not count" demonstrates that your count is the product of judgment, not enthusiasm, and that single move does more for your credibility than ten more close-ups of good hits.

Shingle type changes the signature

The same storm leaves a different mark on different materials, and a sharp inspector reads the signature to the material:

• Three-tab asphalt. Thinner mat, so functional bruising shows and feels obvious, and the soft spot is easy to find with a thumb. Older three-tab gets brittle, so test-bending a tab to show fracture is a fair, factual demonstration of pliability.
• Architectural/laminate asphalt. Thicker, layered. Impacts can mark the top layer while the bruise hides in the mat, so the feel test matters even more here. Look for the dark circular spot plus the give.
• Wood shake/shingle. Hail damage shows as splits with sharp, fresh edges, often with impact marks at the split origin and a corresponding indentation. Fresh, light-colored split faces distinguish hail splits from old weathering checks.
• Tile (concrete/clay). Hail tends to chip, crack, or shatter rather than bruise. Document chips and cracks with location and a scale; tile damage patterns read very differently from asphalt and should be described as what they are.
• Metal. Functional concerns are usually denting on the flat pans and damage to the coating or seams; cosmetic denting is a separate conversation. Document the dent pattern and any coating fracture.

Naming the material and reading the signature to it tells a reviewer you know the trade. A generic "hail hits" note on a tile roof, described as if it were asphalt, tells them the opposite.

Why "per slope" is the whole game

Hail does not fall evenly on a roof. A storm has a direction, and the slopes facing into it take the punishment while the leeward slopes may be nearly untouched. This is the single most important concept in hail documentation, and it is the one most often skipped. You do not document "the roof." You document each slope as its own case, because each slope has its own exposure, its own density, and its own answer to the question of whether it meets the threshold for replacement.

A west-facing slope might show eleven functional hits per square while the east slope shows two. That is not an inconsistency; that is physics, and it is exactly what an honest report should show. When a reviewer sees a per-slope breakdown that matches the storm direction, with the windward planes hammered and the leeward planes light, the documentation reads as real because it behaves the way hail actually behaves. When every slope magically shows the same convenient number, it reads as fabricated. Let the roof tell the truth slope by slope, and the truth is your strongest asset.

Before you climb: build the storm case from the ground

The inspection does not start on the roof. It starts with two questions you answer before you ever set a ladder: did a hail event of meaningful size actually occur at this address, and on what date. If you cannot anchor the damage to a real, dated storm, the best square in the world is documenting damage of unknown origin. Reviewers know this, and "when did this happen" is one of the first things that comes back on a thin file.

Pull the verifiable weather record

The authoritative public sources are the NOAA Storm Prediction Center and the NOAA National Centers for Environmental Information Storm Events Database. The Storm Events Database catalogs events with location, date, time, and magnitude, including reported hail diameter, and it runs from 1950 to the present. The SPC severe-weather plot lets you pull dated hail reports near a location. These are government records, they are free, and they carry weight precisely because they are not produced by anyone with a stake in your roof.

A few practical notes on weather records that pros learn the hard way:

• Reported hail size is a sample, not a survey. A 1.75-inch report a mile away does not prove 1.75-inch stones hit this exact address. Hail is spotty. The public report establishes that a hail-producing storm moved through on a date; the roof establishes what actually landed here. You need both halves.
• The size that matters is the size that damages the material in question. Soft three-tab can bruise at smaller stone sizes than a thick laminate. Note the reported size, but let the roof's condition, not the report, drive the functional-damage call.
• Finalized annual data lags. Official storm data for a full year is typically published the following year, so very recent events may show as preliminary reports. Use what is available and label it as preliminary if that is what it is.
• Multiple storms muddy the water. If two hail events hit the same address months apart, your weather pull should show both, and your roof read should be honest that you may be looking at combined exposure. Pretending there was one clean event when the record shows two is the kind of thing that unravels a file.

Anchor the date honestly

The homeowner's memory of "sometime last spring" is not a date of loss. Cross-reference what they remember against the dated storm reports for the area and land on the most defensible date the record supports. If the homeowner genuinely cannot recall and the record shows three candidate storms in an eighteen-month window, say so in your notes rather than inventing certainty. A report that says "hail reports of 1.25 inch on this date approximately 0.8 miles from the address, consistent with the impact pattern observed" is bulletproof. A report that asserts an exact date with nothing behind it is a liability.

Photograph collateral on the way to the ladder

Soft metals are your honest second witness, because they record impacts the same way the roof does and they are hard to fake. Before you climb, walk the property and document the soft-metal story:

• Gutters and downspouts (dents on the round, upward-facing surfaces)
• Roof vents, turbines, and pipe-jack caps (dings and dents)
• HVAC condenser fins (flattened or bent in a directional pattern)
• Window screens, fascia wraps, and aluminum trim
• Mailbox, grill lid, garage door, and AC disconnect cover
• Painted surfaces showing spatter marks (the dark dots where hail knocked oxidation off a chalky painted surface)

When the soft metals around the home show directional denting that lines up with the slopes taking the worst of it on the roof, your case becomes coherent. The collateral and the roof corroborate each other, and corroboration is what makes documentation hard to dismiss.

The on-roof workflow, step by step

This is the core. Run it the same way every time so nothing gets skipped and every file looks like it was built by the same disciplined hand.

Step 1 — Safety and access first

Falls are the leading cause of death in construction, and steep, possibly storm-loosened roofs are exactly where they happen. Tie off, use proper anchorage, and respect the conditions. OSHA's general residential fall-protection trigger is six feet, and a hail roof you have never walked is not the place to freelance. If the pitch, the height, or the surface condition makes a safe walk impossible, document from a drone and a ladder-assisted edge inspection rather than putting a person where they shouldn't be. A great packet is worth nothing if someone gets hurt building it.

When you do go to a drone, document the way you would on foot, adapted to the platform. Fly each slope as its own pass, capture a high-resolution orthogonal shot looking straight down at each plane so impacts are not foreshortened, and get low, close obliques over the area you would have squared. A drone cannot do the feel test, so it does not replace the thumb on a borderline roof, but on a steep or fragile roof it gets you honest, high-resolution per-slope imagery, and the orthographic overhead shot is excellent for showing impact distribution across a whole plane. If the count is going to hinge on the feel test, note in the file that the roof was inaccessible for a safe walk and document why. Reviewers understand a steep-and-fragile note; they do not understand a slope that simply went undocumented.

Step 2 — Establish slope identity and orientation

Before a single chalk mark, you label the roof. Decide and record your slope-naming convention and use it consistently across photos, the square diagram, and the estimate. Two conventions work well:

• Cardinal: North, South, East, West (and the in-betweens for hips). Cleanest when slopes align roughly to compass directions.
• Front/Rear/Left/Right: oriented as you face the front of the house from the street. Cleanest on complex roofs where cardinal directions get muddy.

Pick one, write it down, and never mix them. Then shoot an orientation photo for each slope: a wide shot that shows the whole plane, a recognizable landmark (the front door, the driveway, a chimney), and ideally a compass or your phone's compass reading in frame. The reviewer needs to place every close-up you take onto a specific plane of a specific roof. Orientation photos are the map that makes that possible.

Step 3 — Find the worst representative area on the slope

Walk the slope and find where the damage concentrates. You are looking for the area that honestly represents the worst of that plane, not a cherry-picked sweet spot and not a barren patch. The square goes where a reviewer would expect to find damage if the slope is genuinely hit. Avoid placing it across a ridge, valley, or penetration, because edges and transitions skew the count and invite a fair challenge. Keep the square in a clean field of the slope's main field shingles.

Step 4 — Mark a true 10x10

Chalk a real 10-foot by 10-foot box. Use a chalk line or a measuring tape and lumber crayon, not a freehand guess, because a sloppy square that is actually 8x8 understates your density and a 12x12 overstates it, and either one gives a reviewer a reason to doubt everything. Brilliant orange or blue lumber chalk photographs well against most shingle colors. Mark the corners clearly. The square should read instantly as a deliberate, measured boundary in the photo.

Some inspectors prefer a test square frame or marked grid for repeatability, and a couple of carriers reference 10x10 explicitly. Whatever your tool, the principle is identical: a known, measured 100-square-foot boundary you can count inside of and a reviewer can trust.

Step 5 — Find, verify, and circle each functional hit

Now you work the inside of the square methodically, left to right, top to bottom, so you cover every shingle and double-count nothing. For each candidate mark, run the functional-damage test before you circle it:

1. Look — is there displaced granule loss exposing the dark asphalt mat, a round dark spot, or fractured granules in a circular pattern consistent with an impact?
2. Feel — press the spot with your thumb. Functional bruising gives a soft, mushy response because the mat fibers are fractured underneath. A cosmetic mark stays firm.
3. Pattern-check — is it random and circular (hail) or linear, clustered on traffic paths, or aligned with mechanical contact (not hail)?
4. Cross-check size — do the impact diameters roughly match the hail size the weather record reported? Wildly mismatched sizes deserve a second look.

If and only if a mark passes as functional, circle it with your crayon. The circle is doing two jobs: it lets you keep an accurate count, and it shows the reviewer exactly what you counted and why. A square full of circled marks that each clearly read as hail is persuasive. A square with circles around granule loss that is obviously age or foot traffic destroys your credibility for the entire roof.

Resist the urge to inflate. If the slope honestly has four functional hits, document four. Padding a count is the fastest way to lose a reviewer's trust, and a reviewer who catches one bad circle re-examines every other claim in your file with hostility. Honest counts, even low ones, build the credibility that carries the slopes that genuinely qualify.

Step 6 — Photograph the square as a sequence

One photo of a square proves almost nothing. A sequence proves the whole thing. For each test square, shoot this set, in order, the same way every time:

1. Context shot — wide, showing the chalked square sitting on the labeled slope with an orientation landmark visible.
2. Full-square shot — the entire 10x10 box framed, with all circled hits visible, so the count is verifiable from the photo alone.
3. Mid-range cluster shots — sections of the square showing several circled hits together.
4. Close-ups with scale — individual impacts with a reference object in frame for scale: a chalk-marked coin, a tape measure, a hail gauge, or a keychain scale. A close-up with no scale is a close-up of nothing measurable.
5. The thumb/probe shot — a finger or chalk pencil on a soft, fractured spot showing the give, paired with the granule loss.

Keep the lens clean, shoot in good light, and avoid harsh shadow that hides texture. Blurry, dark, scale-free close-ups are the number-one reason a real damage file gets bounced for "insufficient documentation." The damage was there. The photos didn't carry it.

Step 7 — Repeat on every slope

This is non-negotiable and it is where lazy inspections cut the worst corner. You run a square on every slope, including the ones you expect to be light. The light slopes are not a weakness in your file; they are proof of your honesty and proof of storm directionality. A report that shows the windward slopes at nine and eleven hits and the leeward slopes at one and two is enormously more credible than a report that only documents the three good slopes and stays conspicuously silent on the fourth. Silence on a slope reads as something to hide. Document all of them.

On a simple gable roof, that is two slopes. On a complex cut-up roof with multiple facets, dormers, and hips, it can be eight or more planes, and yes, you document the meaningful ones. Group tiny facets sensibly, but never skip a major plane.

Reading density: what the count means per slope

Once you have a verified functional count inside a true 100-square-foot square, you have the density for that plane, and density is what you carry into the estimate. The arithmetic is direct: hits inside the square equal hits per roofing square, and hits per square times the slope's square count equals total documented strikes on that plane.

A worked example. Say you are on a moderately cut-up two-story with a laminate roof, and your per-slope squares come back like this:

The story this table tells is clean: the storm came out of the west and south, the windward planes are heavily impacted, the leeward planes are light, and the pattern is internally consistent with a real directional hail event. That coherence is worth more than any single number.

Now, about thresholds. You will hear that carriers commonly look for somewhere in the range of six to eight functional hits per test square as a working benchmark for a slope, and that some carriers set their own bar higher. Treat these numbers as carrier-specific internal guidelines, not as law and not as your decision to make. Your job is not to declare a slope "approved" or "denied." Your job is to document the actual functional density accurately and let the count be the count. Whether a given slope meets a particular carrier's internal threshold, and whether the roof is repairable versus a full replacement, is the carrier's coverage determination, not yours. State the facts; do not promise the outcome.

This matters legally as much as it matters technically. The moment you tell a homeowner "you've got nine hits, that's an automatic approval," you have crossed from documenting facts into promising a coverage result you do not control. Document the density. Explain what density measures. Stop there.

Repairability and matching, stated as fact not promise

Two technical realities often come up in hail documentation, and you can document both honestly without making coverage claims.

Brittleness and repairability. On older or weathered shingles, repairs can be impractical because surrounding shingles fracture when lifted to make a tie-in. You can document this directly: photograph a test repair attempt where adjacent shingles crack on lifting, and note the shingle's condition and pliability. That is a factual observation about the material. Whether it drives a repair-versus-replace decision is, again, the carrier's call.

Matching. If discontinued or unavailable shingles make a slope-only repair create a visible mismatch, document the specific product, its availability, and the mismatch with photos. Some states have specific matching provisions in their insurance regulations. You document the physical facts of the material and availability; you do not interpret the homeowner's policy language or any state matching statute for them. If matching law is relevant, that is a question for the homeowner and their carrier, and possibly the state department of insurance, not for you to adjudicate.

Which roofs are even worth this workflow: targeting before the ladder

Here is the operational truth nobody puts in the glossy guides: this documentation workflow is expensive in time and payroll, and running it on roofs that will not qualify burns your crew for nothing. The slopes that document well are the ones where two things are both true: a real hail event of meaningful size actually hit the address, and the roof was old or worn enough that the impact produced functional damage rather than bouncing off fresh, pliable shingles. A two-year-old roof in a light storm rarely produces a packet worth building. A fifteen-to-twenty-year-old roof under a genuine hail core often does.

The problem is you usually cannot see either of those things from the truck. Roof age is invisible from the curb; a re-roof done eight years ago looks identical to original construction, and county records show the year the house was built, not the year the roof was last replaced. Storm exposure is invisible too; "it hailed somewhere in the metro" does not tell you which specific addresses sat under the damaging core versus the edge that did nothing.

This is the gap RoofPredict is built to close, and it is worth being precise about what it does and does not do. RoofPredict scores roofs house by house using two signals fused together: a roof-age range estimated from aerial imagery, and storm exposure modeled per individual roof rather than looked up as a regional hail map. The distinction on the storm side is the part that matters for this workflow. A hail map shows you where it hailed across an area. RoofPredict models the storm onto each roof, so the output is closer to which specific roofs the storm likely wore out, paired with how old each one already was.

For a restoration crew, that turns a blind canvass into a ranked list. Instead of knocking a whole subdivision and running test squares on roofs that were re-roofed last year or sat on the harmless edge of the storm, you walk the addresses where an aging roof and a modeled hit overlap, which is exactly the population most likely to produce a documentable, functional-density packet. It also enriches a contractor's own CRM and mailing list with roof-age and storm signals, so the old estimates and past customers already in your book get flagged when a storm gives them a reason to call you back.

The honest limits, because overselling this would be its own kind of bad documentation: the age figure is a range, not a roof's birth certificate, and the storm signal is odds, not proof. It does not measure the roof, identify the shingle, or count hits for you; nothing replaces a person marking a true square and verifying functional damage by hand. What it does is point your inspection time at the roofs where this entire workflow is most likely to pay off, so the squares you run are the squares worth running. Targeting decides which doors get the ladder; the test square decides what is actually on the roof. You still need both.

Building the estimate: from square to Xactimate-aligned scope

Documentation and the estimate are two halves of the same packet. The photos and counts prove the condition; the estimate prices the repair of your scope of work in the carrier's own language so there is nothing to translate and nothing to argue. The most common reason a well-documented roof still stalls is an estimate that does not line up with how carriers actually price work.

Write it in the carrier's language

Xactimate is the estimating platform most carriers use, and writing an estimate aligned to its line items and pricing structure removes friction. You are not negotiating; you are presenting your repair scope in the format the file will be reviewed in. Practical alignment moves:

• Use accurate measurements. Pull squares, ridge, hip, valley, eave, and rake lengths from real measurements (a measured diagram or a verified aerial measurement report), not eyeball estimates. A scope built on wrong quantities invites a full rework regardless of how good your photos are.
• Match line items to actual conditions. Steep charges for steep slopes, multi-story access where it applies, the correct underlayment, starter, ridge cap, drip edge, and flashing line items, and the right waste factor for the roof's complexity. Each line should map to something visible in your photos.
• Document code-driven items factually. If the local building code requires something on a re-roof (ice-and-water shield in certain climate zones, specific deck attachment, drip edge, ventilation), cite the applicable code section and note that it applies. The IRC and your locally adopted amendments govern this. You are stating the code requirement as fact, not arguing the homeowner's coverage of it.
• Account for every layer and accessory the photos show. Pipe boots, ridge vents, valleys, chimney and wall flashing, and detached items that have to be reset all belong in the scope when they are part of the work, and your photos should already show them.

A scope line list that maps to the photos

The discipline here is one-to-one: every line on the estimate should have a photo behind it, and every roof component in the photos should have a line. A skeleton scope for a typical asphalt re-roof, with the documentation that backs each piece, looks like this:

The waste factor is where amateurs get quietly dinged. A simple gable roof carries a low waste percentage; a cut-up roof with multiple hips, valleys, and dormers carries a higher one, because more cuts mean more offcut. Pick the waste factor the roof's geometry actually justifies and be ready to point at the diagram that shows the complexity. A defensible quantity beats an aggressive one every time, because the moment a reviewer rejects one inflated quantity, they re-check all of them.

A short worked illustration ties it together. Take the four-slope roof from the density table: 16 + 16 + 7 + 7 squares is 46 squares of field area before waste. On a moderately cut-up roof you might carry roughly a 12 to 15 percent waste factor on the field shingles, while linear items (ridge, hip, valley, eave, rake) come straight off the measured diagram with their own small waste. The estimate then prices tear-off and install across that 46-plus-waste square count, adds the linear-driven accessories, layers in steep and two-story access where the photos show they apply, and lists code-driven items with their citations. Nothing on the page is a guess; every number traces to either a measurement or a photographed condition. That is what "Xactimate-aligned" actually means in practice, and it is why a clean estimate moves and a sloppy one circles for weeks.

Documenting the re-inspection, if there is one

Sometimes a carrier sends an adjuster to re-inspect, and your documentation should make that meeting productive rather than adversarial. You are allowed to be present as the contractor who will do the work and to point out the conditions you documented. What you are not doing is negotiating the claim or arguing coverage. The clean play is to have your packet in hand, walk the same squares you already marked (the chalk is often still there), and let the adjuster verify your count against the boundary you defined. If you and the adjuster reach different counts, document both readings as facts and note the difference; do not turn it into a fight. A calm contractor who can show a methodical, pre-existing packet and a chalk square the adjuster can re-count on the spot is in a far stronger factual position than one who shows up empty-handed and starts making claims about what should be approved. Let the square do the talking. Coverage is still the carrier's call.

The packet structure that gets read

Hand over a single, organized packet rather than a shoebox of photos. A structure that reviews cleanly:

1. Cover summary — address, date of loss anchored to the weather record, roof type, age range observed, and a one-line per-slope density summary.
2. Weather verification — the dated NOAA/SPC report(s) for the address with the reported hail size and distance.
3. Collateral evidence — soft-metal and spatter photos showing directional impact around the property.
4. Per-slope sections — one section per slope, each with its orientation photo, full-square photo, cluster and close-up shots with scale, the functional hit count, and the slope's documented strike total.
5. Repairability and matching notes — factual observations with photos where relevant.
6. Measurements — the measured diagram or aerial measurement report behind your quantities.
7. The estimate — Xactimate-aligned scope with line items mapped to the documented conditions.

Number your photos, label every slope consistently with whatever convention you chose, and make the packet so easy to follow that a reviewer who never visited the property can reconstruct the roof in their head. That is the bar: not "could a roofer tell there's damage," but "can a stranger at a desk verify the count and the scope from the file alone."

The compliance line: what you document, what you never say

This section is not optional reading, and it is not legal boilerplate. Crossing this line is how roofers lose licenses and how companies catch enforcement actions. The rule is simple to state and easy to violate when you are trying to be helpful: you may inspect, document, photograph, and write an accurate estimate to repair your own scope of work, and you may state facts about your scope. You may not, for compensation, negotiate, adjust, interpret, or handle the homeowner's insurance claim. That is the practice of public adjusting, it is licensed and regulated, and in some states even calling yourself an "insurance specialist" or advertising that you handle claims has been found to cross the line.

The safe frame, every time: you document thoroughly, you write an accurate estimate, and you hand it to the homeowner. The homeowner files the claim. The insurer decides coverage. You are a documentation-and-estimating expert about your own work, full stop.

Here is the do-not-say list, and it is worth posting where your sales team can see it:

• Do not promise approval or payment. Never "we'll get this approved," "this is a guaranteed claim," or "nine hits means automatic replacement." You document density; the carrier decides coverage.
• Do not touch the deductible. Never offer to waive, absorb, cover, eat, or "make the deductible disappear." In many states that is insurance fraud, plainly. The deductible is the homeowner's, period.
• Do not advertise or imply a "free roof." A storm does not equal a free roof, and saying so is both false and, in several states, specifically prohibited.
• Do not negotiate, handle, manage, or fight the claim. Never "we'll deal with the adjuster for you," "we'll fight the denial," "we maximize your settlement," or "we'll handle everything with insurance." That is unlicensed public adjusting.
• Do not interpret policy or coverage. Never tell a homeowner what their policy covers, what their matching rights are, or what a clause means. Hand them the facts; coverage questions go to their carrier or their state department of insurance.
• Do not represent the homeowner against the insurer. You are not their advocate in the claim. You are the contractor who documented the roof and priced the repair.

What you absolutely can and should say: "Here is thorough documentation of the functional damage we found, slope by slope, with photos and counts. Here is the dated storm record for your address. Here is an accurate estimate to repair our scope of work, written in the format your carrier uses. You file the claim with your insurer, and they'll make the coverage decision." That sentence is honest, useful, and entirely inside the line. Teach it to every rep, because the reps who freelance promises are the ones who get a company in front of a regulator.

What pros get wrong (and the fixes)

A punch list of the failure modes that turn real damage into denied files, and how to close each one:

• One square for the whole roof. The fix: a verified square on every meaningful slope, light ones included. Per-slope or it didn't happen.
• Counting cosmetic marks as functional hits. The fix: run the look-feel-pattern-size test on every mark before you circle it. One bad circle poisons the whole square.
• Scale-free close-ups. The fix: a reference object in every close-up. A coin, a tape, a gauge. No scale, no measurable evidence.
• No orientation. The fix: a context shot per slope with a landmark and compass, so every close-up can be placed on a specific plane.
• A guessed-size square. The fix: chalk a true measured 10x10. An 8x8 understates you; a 12x12 invites a challenge.
• No weather anchor. The fix: pull the dated NOAA/SPC record for the address first. Damage of unknown origin is weak damage.
• Inflated counts. The fix: document the honest number. A reviewer who catches padding distrusts everything else you wrote.
• An estimate that doesn't match the documentation. The fix: every line item maps to a photographed condition and a real measurement, written Xactimate-aligned.
• Crossing the compliance line to "help." The fix: document and estimate; never promise, negotiate, or touch the deductible. Helpful and illegal are not the same thing.
• Running squares on roofs that were never going to qualify. The fix: target before you climb so your inspection time lands on aging roofs that actually took a modeled hit.

A repeatable field checklist

Print this. Run it the same way on every roof.

Before the ladder

• Pull dated NOAA/SPC hail record for the exact address; note size and distance
• Confirm a defensible date of loss against the record
• Photograph collateral: gutters, downspouts, vents, HVAC fins, screens, painted spatter
• Confirm roof is safe to walk; set fall protection or plan a drone/edge inspection

Per slope, every slope

• Label the slope with your chosen convention (cardinal or front/rear/left/right)
• Shoot an orientation/context photo with a landmark and compass
• Find the worst representative field area (off ridges, valleys, penetrations)
• Chalk a true measured 10x10
• Work the square methodically; run the functional test on each mark; circle only functional hits
• Photograph the sequence: context, full square, clusters, close-ups with scale, thumb/probe shot
• Record the functional hit count and compute documented strikes for the slope

Back at the desk

• Pull real measurements (measured diagram or verified aerial report)
• Write the Xactimate-aligned scope; map every line to a photographed condition
• Cite applicable code-driven line items as fact
• Assemble the packet in order: summary, weather, collateral, per-slope, repairability/matching, measurements, estimate
• Hand it to the homeowner; they file, the carrier decides

The bottom line

The test square is not a trick and it is not a number you chase. It is a discipline. Mark a real boundary, count only what is functionally damaged, do it on every slope, photograph it so a stranger can verify the count, anchor it to a dated storm, and price the repair in the carrier's own language. Do that and you stop arguing about whether the damage is real, because the file proves it before anyone asks.

The roofs where this entire process pays off are a specific population: aging roofs that actually sat under a damaging storm. You can spend your crew finding them door by door, or you can point the inspection at the addresses where an old roof and a modeled hit overlap and run squares that are worth running. RoofPredict exists to make that targeting honest and specific, by scoring roofs house by house on an age range plus storm exposure modeled per roof, so the documentation work lands where it counts. It tells you which roofs are due. The square tells you what is on them. Bring both, document like a professional, stay strictly on the document-and-estimate side of the line, and let the evidence do the talking.

If you want to see which roofs in your area carry both an aging-roof age range and a modeled storm hit, book a demo and hand us a roof you already know. You decide if we got it right.

FAQ

How big should a hail test square be?

The standard test square is a true 10 feet by 10 feet, which equals 100 square feet, or exactly one roofing square. Mark it with a measured chalk line, not a freehand guess, because an undersized square understates your hit density and an oversized one invites a fair challenge to your whole count. The point of a fixed 100-square-foot boundary is that the number of functional hits inside it scales directly to the slope: hits in the square equal hits per roofing square.

How many hail hits per test square count as damage?

You will commonly hear a working range of roughly six to eight functional hits per 100-square-foot test square as a benchmark, and some carriers set their own bar higher. Treat those as carrier-specific internal guidelines, not as a rule you enforce. Your job is to document the actual functional density on each slope accurately, not to declare a slope approved or denied. Whether a count meets a particular carrier's threshold and whether the roof gets repaired or replaced is the carrier's coverage decision.

Do I need a separate test square for every slope?

Yes, every meaningful slope, including the ones you expect to be light. Hail falls directionally, so windward slopes take far more impact than leeward slopes, and documenting all of them is what proves the pattern is real. A report showing the front and side slopes hammered and the rear slope light reads as honest and consistent with a directional storm. Documenting only the good slopes and staying silent on the rest reads as something to hide.

What is the difference between functional and cosmetic hail damage?

Functional hail damage compromises the shingle and shortens its service life: a fractured or bruised mat that feels soft when you press it, displaced granules exposing the asphalt, and fractured granules in a circular impact pattern. Cosmetic marks, blistering, manufacturing flaws, foot-traffic scuffs, mechanical damage, and age-related granule loss are not functional hail hits. Only functional hits should be counted and circled inside your test square; counting cosmetic marks undermines the credibility of the entire square.

How do I verify the date a hail storm actually hit an address?

Pull the public record from the NOAA Storm Prediction Center severe-weather plot and the NOAA Storm Events Database, which catalog dated hail events with location, time, and reported hail diameter. Cross-reference what the homeowner remembers against those dated reports and land on the most defensible date the record supports. Note that reported hail size is a sample from a nearby report, not proof of the exact stone size at that specific address, so let the roof's condition confirm what actually landed.

What photos do I need to document a test square properly?

Shoot a sequence, not a single photo: a wide context shot showing the chalked square on a labeled slope with an orientation landmark, a full-square shot where every circled hit is countable, mid-range cluster shots, close-ups of individual impacts with a reference object for scale, and a thumb-or-probe shot showing a soft fractured spot. Scale-free, blurry, or unlabeled close-ups are the leading reason a genuinely damaged roof gets bounced for insufficient documentation.

Can a roofer handle the insurance claim for the homeowner?

No. A roofer may inspect, document damage, write an accurate repair estimate for their own scope of work, and state facts about that scope. A roofer may not, for compensation, negotiate, adjust, interpret, or handle the homeowner's claim, promise approval or payment, interpret policy or coverage, touch the deductible, advertise a free roof, or represent the homeowner against the insurer. That is licensed public adjusting. The homeowner files the claim and the insurer decides coverage; you provide documentation and an estimate.

Why does writing the estimate in Xactimate format matter?

Xactimate is the estimating platform most carriers use to review files, so a scope aligned to its line items and pricing presents your repair in the format it will be read in, removing friction and translation. That means accurate measured quantities, line items that map to photographed conditions, correct waste and access factors, and code-driven items cited as fact. You are not negotiating; you are presenting your repair scope clearly so there is nothing to argue about.

How do I decide which roofs are worth a full documentation inspection?

The roofs that document well share two traits: a real hail event of meaningful size hit the address, and the roof was old or worn enough for the impact to cause functional damage rather than bouncing off fresh shingles. Both are hard to see from the street, since roof age is invisible and county records show year built, not last re-roof. Tools like RoofPredict score roofs house by house on a roof-age range plus storm exposure modeled per roof, so you point inspection time at the aging roofs that actually took a hit. The age is a range and the storm signal is odds, so it targets your work; it does not replace marking a square and verifying functional damage by hand.

What is the most common reason a real hail-damage file gets denied?

Documentation that does not carry its own weight: one square for the whole roof instead of per slope, close-ups with no scale, no orientation photos to place the damage, no dated weather record anchoring the loss, cosmetic marks counted as functional hits, or an estimate whose quantities and line items do not match the photographed conditions. In most of these cases the damage was real; the file simply failed to prove it to a reviewer who never saw the roof in person.