Canvass Pro.

storm

How to Canvass a Hail Swath Door to Door

Turning a NOAA hail report into a worked territory, including how to confirm the storm date and what hail size actually means.

A hail swath is a lead list that expires. The storm draws a shape on the map, that shape contains a few hundred or a few thousand homes, and for a short window those homes have a reason to talk to a roofer that they did not have the day before.

Working it well is mostly a mapping and sequencing problem. Here is how to turn a weather event into a worked territory.

Confirm the storm before you knock it

Before anything else, establish that hail actually fell where you think it fell, and on what date. You will need the date for every conversation and every claim, and a rep who gets it wrong in front of a homeowner has lost the conversation.

Two public sources do this, both free.

The Storm Prediction Center's storm reports archive. The SPC storm reports page serves daily report maps and downloadable CSV and KML files of tornado, hail, and wind reports. The page notes that reports are organized "from 1200 UTC to 1159 UTC the next day," that you retrieve a past day by typing the date in YYMMDD form, and that data is "available back to 6/1/99." It also states that all reports are preliminary.

NOAA's Storm Events Database. The NCEI Storm Events Database is the archival version, described on its own page as "a publicly accessible source of National Weather Service (NWS) Storm Data." It is searchable by map or by state, offers bulk CSV, and currently contains data from January 1950 forward.

One detail on that page is worth knowing and almost nobody in the trade uses it: NCEI will issue a certified report. The page says to contact ncei.orders@noaa.gov "with the search results that you want certified." If a homeowner's claim turns into a dispute about whether a storm occurred, a certified NOAA record is a stronger document than a screenshot.

NOAA also publishes its raw severe weather database files at the SPC's warning coordination page, with hail and damaging wind events dating from 1955. If you work storms regularly, pulling the annual hail CSV and mapping it yourself is more useful than any third-party hail map, because you can see the individual reports and their stated sizes rather than an interpolated surface.

Know what the sizes mean, and what they do not

Hail reports come with a size in inches, and that number drives everything downstream. Two things are worth getting right.

What counts as severe. The Storm Prediction Center lists "Hail 1 inch in diameter or larger" as one of the criteria for a severe thunderstorm, and "Hail 2 inch in diameter or larger" as significant severe. The 1 inch line is relatively recent. NWS Service Change Notice 09-52 raised the warning criterion "FROM 3/4 INCH /PENNY/ DIAMETER OR LARGER HAIL TO 1 INCH /QUARTER/ DIAMETER OR LARGER HAIL," effective January 5, 2010. If you are looking at reports from before 2010, the threshold for a report existing at all was lower.

The coin references. An NWS reference sheet, the wind and hail reference chart, lists the standard comparisons: 0.75 inch penny, 1.00 inch quarter, 1.25 inches half dollar, 1.50 inches ping pong ball, 1.75 inches golf ball, 2.00 inches lime, 2.50 inches tennis ball. It notes that "Hail size refers to the diameter of the hailstone." This is a local office reference sheet rather than national policy, but the comparisons are the ones in general use, and a homeowner describing "golf ball sized" is giving you roughly 1.75 inches.

What the size does not tell you. Be careful here, because this is where the trade gets loose. The Insurance Institute for Business and Home Safety tested commercially available asphalt shingles with laboratory hailstones and reported in Hail Impact Performance of Asphalt Shingles that "All products tested with 1.5-in. laboratory hail were rated as Good or Excellent performers," and that "In their new state, both IR and non-IR asphalt shingles should offer good protection for 1.5 in. or smaller hail impacts."

IBHS also states its own limits: it did not test all available shingles, products were tested "in their new state," and the lab-to-field equivalence claim is limited to non-wind-driven conditions.

So the defensible version of this, the one a rep can say at a door without being wrong, is that hail size is a reason to inspect, not a determination of damage. A reported 1.75 inch stone three streets over is a good reason to look at a roof. It is not a finding that the roof is damaged, and a rep who says otherwise is making a claim neither they nor the lab data can support.

Draw the swath as a polygon, not as a radius

The common mistake is treating a storm as a circle around a reported point. Hail cores travel. A supercell lays down a long, narrow corridor, often miles long and a few blocks wide, and the edges matter enormously because a street inside the corridor and a street two blocks outside it are completely different conversations.

Practical method:

  1. Plot the individual hail reports with their sizes and times. The times tell you the direction of travel.
  2. Draw the corridor along the track, not a ring around the biggest report.
  3. Mark the heaviest reported sizes within the corridor, because those are the sub-areas worth working first.
  4. Pull every address inside the polygon. That is your lead list.

A swath drawn this way usually looks nothing like the circle a rep would have drawn by eye, and it will include streets they would have skipped and exclude streets they would have wasted a day on.

Sequence inside the swath

Once you have the polygon and the addresses, the order matters more than the total count, because the window is short.

Rough priority:

Highest reported hail size first. Not because size determines damage, but because that is where the density of real findings and real claims will be highest, and where your first jobs will come from.

Streets with uniform housing stock. A subdivision built out in a two or three year window has roofs of the same age and material. A rep who learns the conversation for that roof gets to have it a few hundred times, and a job on one street is a reference on the next.

Streets where you already have a job. "We are working on the roof at 412" is the strongest opener in storm canvassing and the only one that is both verifiable and locally relevant. Sequence outward from your first jobs rather than jumping across the swath.

Edges last. The boundary of the swath is where you will find the most homeowners with no damage, and the most conversations that go nowhere. Work the core first, work the edges when the core is done.

What to put on the map before the first knock

Three layers, and all three before anyone walks:

The swath polygon. So reps know where the storm is and where it is not.

Jurisdiction lines. Permit rules are municipal. A hail corridor will cross city boundaries without caring about them, and a rep working a swath can be legal on one street and in violation two blocks later. Know which cities your polygon touches, and get the permits before the week starts, not after a complaint.

Do-not-knock and no-soliciting. A storm does not clear a prior refusal. A posted no-soliciting sign is the same sign it was last week. This is the exclusion most likely to get overlooked in the rush of a storm week, and it is the one most likely to produce a complaint to a city that is already fielding calls about out-of-town roofers.

The honest frame

A hail swath gives you something canvassing almost never has: a genuine, verifiable, locally specific reason to be at someone's door. The homeowner heard the storm. They may have already seen their neighbor's tarps.

That is a real advantage and it is also the reason the category attracts bad actors, which is why several state agencies publish warnings about post-storm door-to-door contractors. The teams that do well in storm work are the ones whose reps can state the storm date, the reported hail size, and the source, and then say they are there to look rather than there to diagnose.

Everything else in storm canvassing is sequencing, and sequencing is a map problem.

Canvass Pro renders hail swaths as polygons under the door pins and turns the homes inside a swath into a routed lead list.

Sources

More on this