Desktop underwriting vs. field inspection for property insurance
Every submission that lands on an underwriter's desk comes with a story the applicant tells about the property. Roof age, roof type, number of stories, condition of the structure. Most of it is accurate. Some of it is optimistic. The question that drives the desktop-vs-field debate is simple: how much of that story do you verify before you bind, and how?
What a field inspection gets you
A field inspection sends a loss control rep to the property, usually after binding on a 30 or 60-day contingency. They walk the roof, photograph the mechanicals, note the condition of the siding, and write it up. It's thorough. It's also slow, it costs money per address, and it happens after the policy is already on the books. If the rep finds a problem, you're now non-renewing or endorsing a risk you already took, which is a worse position than catching it at submission.
Field inspection isn't going away for complex commercial risk or anything that needs a human judgment call on interior conditions. For the bulk of residential and light commercial submissions, confirming the roof matches what the applicant told you doesn't need a person standing on the ground staring up at it.
What desktop underwriting can and can't do
Desktop underwriting today mostly means pulling whatever third-party property data report is attached to the submission and reading it. Those reports are useful for things like year built, square footage, and prior claims history. They're weaker on the thing that actually drives roof risk: what the roof looks like right now, from above, where you can see pitch, material, and condition without anyone standing on a ladder.
That's the gap imagery fills. A desktop review built on current very high resolution imagery (0.3 meters or better) lets an underwriter see roof geometry, the material covering it, and how many stories the structure has, pulled from the image itself rather than from a form the applicant filled in. Add visible site features like a trampoline, a pool, or a detached structure that didn't make it onto the application, and you've closed most of the distance between what they told you and what's there. Property Characteristics builds exactly this record, per address, from satellite or aerial RGB imagery, refreshed on an annual cadence, before the policy binds rather than after.
What it won't do is replace a field inspection where one is warranted. Imagery shows you the roof and the lot. It doesn't tell you the condition of the wiring, the age of the water heater, or what's stacked in the garage. For risks that need that level of detail, a field inspection or a physical loss control visit is still the right call. Desktop verification sorts submissions so that field inspections go to the risks that need them, instead of every file getting the same treatment regardless of what the imagery already shows.
Where this fits in the underwriting workflow
Timing changes first. Instead of binding on the applicant's word and inspecting afterward, you pull the property record at submission, alongside or instead of whatever third-party data report you're already using. If the roof geometry and material match the application, you bind with more confidence. If they don't, that's a rating conversation or a referral before the policy is in force, not an endorsement six weeks later.
For MGAs writing high volume on tight turnaround times, that ordering matters more than it sounds like it should. A field inspection queue backs up. An image-based record doesn't, because there's no rep to schedule and no ladder to climb. Your loss control program still handles the risks that need a site visit. The imagery decides, at submission, which files those are.
If you're trying to work out where a given book of business falls on the desktop-versus-field line, Property Characteristics is built around that exact decision: a structured roof and site record for each address, ready before you bind.