State of the Claim Q3 2026 recap

6 minute read

Finch ResearchFinch

Finch is a claims intelligence platform that turns residential and commercial property claim data into leverage. Drawing on 11,530 property claim files handled by public adjusters, with data through October 16, 2026, its Q3 report sets out 20 findings on costs, coverage and settlements. Each finding states the sample it rests on.

Database: 11,530 claim files, data through October 16, 2026. Findings come from public adjusters’ claim files and are not a representative national sample. Each finding states the sample it rests on. Claims opened through October 16, 2026; the database was queried on October 17 and 18, 2026.

This recap lists each finding by chapter. Every section links to its chapter in the full report, where the charts and the notes on method are.

The estimate gap

On 46 disputed property claims where both the insurer and the public adjuster itemized the work, the insurer usually listed the same repairs and priced them lower.

Dollars lost to lower pricing for every dollar lost to repairs left out

More than 5 to 1

In dollars, $7.2M lower pricing against $1.3M omissions: most of the gap sat in the price of repairs already on the list, not in what was omitted. On repairs both estimates listed, the insurer priced lower 236 times, higher 77 times and the same 35 times.

Where insurers price lower

Cabinetry: 55%

On a typical claim, the insurer’s cabinetry price was 45 cents for every adjuster dollar, and it was lower on 19 of 21 claims.

How big the price gap usually is

1.20x

On a typical repair both estimates listed, the adjuster’s price was 1.20 times the insurer’s. Most gaps were modest; the money sat in a few very large ones.

See the charts in chapter 01.

Where coverage meets the loss

Building code coverage

33.0%

About one in three claims with a readable policy limit had coverage for upgrades required by current building codes. The typical cap was 10% of the building’s coverage limit.

Exclusions

Wear and tear leads.

Most policies carried earth movement, war and nuclear hazard exclusions, but insurers rarely cited them. Wear and tear was different: insurers cited it on 36.7% of claims with a written coverage decision.

See the charts in chapter 02.

The first number is not the last

Where disputed claims settled

1.41x

On 61 finished claims where a letter stated both sides’ figures, the claim typically settled at 1.41 times the insurer’s figure. The adjuster had typically asked for 1.6 times.

First valuation to final amount

1.37x

On 171 finished claims, the final amount was typically 37% above the insurer’s earliest recorded valuation. More than one claim in four ended below it; one in ten ended at more than twelve times. This differs from 1.41x above because it covers different claims and starts from the insurer’s earliest figure rather than the one stated beside the adjuster’s request.

How long claims take

321 days

Half of claims took more than 321 days from intake to final insurer payment, across 5,050 claims that received payment. About 11% were still being paid when measured.

See the charts in chapter 03.

What the file holds

How much is in a claim file

77 emails, 16 files

The typical claim involved about 77 emails across 17 threads, counting emails visible only in reply history. The busiest quarter involved more than 180 emails; the top 5%, more than 460. Of those, a typical 52 are linked to the claim on the platform (63 counting every message in those threads); the rest appear only as quoted history inside replies. Among claims with files on the platform, the typical one held 16 files; across all stored files, photos and PDFs make up nearly half each; the largest 5% held at least 198. When claims get big, important details can be easy to miss.

Time and money recovered

7.5x

The longest-running quarter of claims had a typical payout 7.5 times that of the quickest quarter. The fastest quarter finished in about three months; the slowest took about two years. This does not mean waiting caused the extra money.

How many policy forms appear in these files

About 2,800

The records first appeared to hold 3,674 different policy forms. Once editions of the same form and spelling variants were grouped, about 2,800 distinct forms remained. Across 2,730 claims carrying a form number, one standard commercial form appeared under 52 different spellings.

See the charts in chapter 04.

Three more patterns in the record

Fire made up most of the claims whose estimates could be compared. Belongings tended to lose more to depreciation than buildings. And some claims appeared to reach the building limit, depending on how repairs were measured.

Fire and estimate gaps

69%

Fire made up 26 of the 45 claims with detailed estimates from both sides. On those, the insurer’s estimate was typically 69% of the adjuster’s, a gap of $45,160, close to the $43,137 across all 45. Insurers came no further from the adjuster on fire than on other claims.

Withheld for depreciation

15.8% vs. 10.7%

When values were reduced for age and wear, belongings lost a typical 15.8%, versus 10.7% for the building. A wider measure points the same way (19.3% on 623 claims for belongings vs 12.4% on 111 for the building). On claims stating both, belongings lost more on two claims in three (37 of 54).

When a building limit may cap a claim

12.8%

Using the settled replacement value, the most complete measure, 12.8% of 897 claims appeared to reach the insurance limit for the building. Other ways of measuring gave lower results.

See the charts in chapter 05.

Duration, duration, and more duration

Duration by state

395 days

Georgia claims took a typical 395 days to resolve, the longest in the book, followed closely by Tennessee and Delaware. California was quickest at 186 days, with Massachusetts and Louisiana close behind.

Longest by cause of damage

378 days

Half of fire claims took longer than 378 days to final insurer payment, against 321 days book-wide. Collapse followed at 348 days.

Shortest by cause of damage

143 days

Vehicle impact claims finished fastest: half took longer than 143 days. Flood took 233 days and water from outside 238 days.

See the charts in chapter 06.

How a claim moves

Which limits are reached most

More than a quarter

Loss of use and other structures each reached their limit on more than a quarter of claims (27–28%). Measured on what insurers paid, counting each payment amount once.

Replies are fast; many requests get none

1.2 hours

On claim email, a reply typically came in a little over an hour (1.2 hours, across about 184,000 replies). But roughly a third of the adjuster’s own requests never got one. The replies that came were quick; the gaps were in the requests that got none.

When people bring in an adjuster

10 days

The typical policyholder brought in a public adjuster 10 days after the loss. 44.2% did so within a week; 4.5% waited more than a year.

See the charts in chapter 07.

What complexity conceals can cost millions.

Connect your inbox and bring the full record into every conversation with the carrier.

In your demo

  • See Finch build a claim record from real correspondence
  • Compare every estimate version line by line
  • Ask a question and get an answer with citations

Book a demo

Meet with our team to see how Finch can work for yours.

Read-only inbox connection. Your data never trains shared models. Revoke access anytime.

Evaluating for your firm?