3-tab vs. architectural shingles: telling them apart from above
An applicant writes "architectural shingle, 2019" on the form. The roof in the imagery looks flat, uniform, almost grid-like from directly overhead. Something doesn't match, and you're trying to decide if it's worth a reinspection order or if you can just rate the file as submitted. Here's what the imagery can tell you, and what it can't.
What separates them on a roof plane
3-tab shingles are cut into equal, single-layer tabs. From the ground they read as a flat, repetitive pattern, and that repetition survives in overhead imagery too: you'll see fine, regular horizontal lines running across each roof plane at consistent intervals, with almost no shadow depth between courses. The color field tends to look flat because there's only one layer of asphalt and granules doing the work.
Architectural (laminate) shingles are built from two or more layers fused together, staggered to fake the look of wood shake or slate. That extra layer throws a shadow. At a ground sample distance of 0.3 m or better, you can usually pick up a visible texture, a dimensional surface instead of a flat grid. The shingle edges don't line up in neat rows the way 3-tab does, and the color field often shows two tones blended into the granule mix instead of one flat shade.
Neither pattern requires measuring an individual tab. What you're looking for is the overall surface character of the plane: flat and striped versus textured and shadowed.
Where shingle class matters on the file
Shingle class feeds into more than one line on the dec page. Architectural shingles generally carry a longer expected service life and, depending on the carrier's rating plan, can affect the wind and hail class assigned to the roof. A 3-tab roof nearing the end of its rated life is a different risk than an architectural roof installed the same year, even if the applicant describes both as "good condition." If the form says architectural and the imagery shows the flat, grid-like pattern typical of 3-tab, that's a discrepancy worth resolving before the policy binds, not after a claim.
The same goes the other direction. Some applicants undersell their roof, listing 3-tab out of habit or because that's what the prior policy had on file, when the actual install is a dimensional shingle. Catching that correctly matters for ITV and for not leaving premium on the table that the risk doesn't carry.
What the imagery won't tell you
Shingle class from overhead RGB imagery is a visual read, not a lab test. It won't give you the manufacturer, the exact product line, or a wind rating in miles per hour. It won't catch hail bruising that hasn't broken the surface yet, and it won't substitute for a close-up inspection when the file calls for one. What it does well is flag the mismatch between what's written on the application and what the roof plane actually looks like from above, before anyone signs off on the rate.
Age complicates the read too. Both shingle types fade and lose granules over time, and a heavily weathered 3-tab roof and a weathered architectural roof can start to look more alike at the edges than either does when new. That's one more reason to treat the imagery as a first pass that either confirms the application or sends the file to a closer look, not as the final word on roof condition.
Building this into the binding workflow
Most shops doing this today are pulling imagery manually, zooming in on each address, and making a judgment call roof by roof. That works until the submission volume picks up. Property Characteristics pulls a structured roof record for the address, including the geometry, material, and storey count that show up in the imagery, so the shingle class question gets answered as part of the file instead of as a side errand before binding.
If shingle mismatches between the application and the roof are costing you rework after bind, it's worth seeing what a structured record looks like on your own book.