Most of the book has been here before
The larger part of what reaches an underwriting team is not a risk nobody has seen. It is a risk the team wrote last year, or wrote in March and has since agreed to change.
SnapLine reads a renewal submission against the expiring policy, and reads a mid-term endorsement or amendment against the policy as it currently stands. Both are live. In each case the comparison is done before the file reaches a person, so what has moved is the first thing on the screen.
Where the renewal hours actually go
Renewals rarely get budgeted the way new business does, on the reasoning that the file already exists and the risk is already known. The broker is sending back something close to what they sent before.
Then somebody opens the expiring policy in one window and the renewal submission in another, and reads both.
That comparison is the work. It is slow, it is repetitive, it scales with nothing except volume, and it is the first thing that gets shortened when the season compresses. A shortened comparison defers cost rather than removing it, and the place deferred cost tends to surface is a claim.
Read against the expiring policy
A renewal submission arrives the way any submission arrives. An email, a revised schedule, a loss run covering the period, and whatever got attached at the last minute. SnapLine extracts it the way it extracts new business, with provenance on every field, traced back to the region of the document it was read from.
What differs is what happens next. The extracted record is set against the expiring policy record, and the view an underwriter opens is the difference between them.
- Values that have moved, and what they moved from on the expiring schedule
- Locations added and locations removed, including the ones that have only changed how the address is written
- Occupancy, construction and protection detail that no longer matches what was written last year
- Terms in the covering email that contradict the expiring slip
- Fields that were present last year and are absent in this submission
- Fields that are identical to last year, which is sometimes the finding
The last of those does more work than it looks like it does. A field that comes back unchanged has not been confirmed. It has been carried, and the two are easy to mistake for each other at speed.
What moved
- building_declared_valueExpiring4,120,000This submission4,560,000MovedRevised schedule, line 1. The expiring figure is the one on the bound schedule, not the one on the broker’s covering email.
- locations_scheduledExpiring62 locationsThis submission64 locationsMovedTwo added. One of the sixty-two has only changed how the address is written, and is not a new location.
- sprinkler_protectionExpiringFull coverage, to specThis submission—no value in this submissionAbsent this yearPresent on the expiring schedule and not in this submission. The gap becomes a request back to the broker.
What did not move
- contents_declared_valueExpiring980,000This submission980,000Carried forward
- occupancy_codeExpiringWarehouse — general storageThis submissionWarehouse — general storageCarried forward
- 2021set
- 2022unchanged
- 2023unchanged
- 2024unchanged
- 2025unchanged
- 2026unchanged
Set once, at inception. Carried through five renewals since, and 2026 is this submission. Nothing in that run is a person looking at the building: a tenant can change, a use can change, a refurbishment can happen, and none of those is a submission event.
Unchanged is a fact about the paperwork. It is not a fact about the risk.
Endorsements land on a policy that has already moved
A mid-term change is small, which is why it gets handled quickly, and why it goes wrong quietly.
The request arrives as an email, sometimes with a marked-up document behind it and sometimes as a single sentence. Someone applies it. The question there is rarely time to ask is which version of the policy it was applied to: the bound version, the version after the endorsement in June, or a copy in a folder that was never brought up to date.
SnapLine reads the endorsement or amendment, extracts what is being changed, and applies it against the current state of the policy record rather than against whichever document happens to be open. Where a requested change conflicts with something already endorsed, the conflict is raised instead of being settled off-screen.
Inherited data does not age visibly
This is the part that reaches the loss ratio rather than the expense ratio.
An occupancy description written when the building held a warehouse tenant stays in place after a light manufacturer moves in, because nothing in a renewal ever asks that field to justify itself again. A declared value set before a refurbishment stays set afterwards, and the refurbishment was never a submission event in the first place.
Underinsurance and a wrong occupancy code are not usually the result of a bad decision. They are the result of a sound decision taken against a figure that stopped being true several renewals earlier, in the file everyone agreed was the straightforward one.
Reading the renewal against the expiring policy makes a carried-forward field visible as a carried-forward field, which is the point at which somebody can decide whether to chase it. Where Atlas is licensed, that field can also be set against an independent source with its age attached. Whether any of that changes an outcome on your book is a question about your book.
Atlas property intelligence is licensed separately from SnapLine and is not required for renewal comparison.
Why renewals are the easiest place to measure this
Reading documents accurately is a means to something. What an underwriting operation is actually buying is a lower cost of handling each file, fewer errors reaching the rating engine, a quicker answer back to the broker, and downstream data that agrees with the documents it came from. Those land on the expense ratio and on the loss ratio, and they are what a CUO or COO is asked about.
Renewals suit measurement because the volume is predictable and the baseline already exists. You know what last season cost you to get through.
This page carries no figures, because a number produced on our sample data would tell you nothing about your own. The mechanism is above; the measuring happens during a fixed-scope Proof of Value across twenty of your live submissions, scoped in writing beforehand, with a written report at the end.
None of this takes the underwriter out of the renewal. The comparison is prepared in advance, the decision stays with a person, and the record carries who made it.
Bring last year and this year
The demo runs on five sample submissions and you can upload up to three of your own into the same session. For this page the useful test is a pair: an expiring policy and the renewal submission that came back against it. Ideally the one where something had moved and nobody spotted it until the broker was chased.
If the comparison survives that, the next step is the Proof of Value across twenty of your live submissions, with a written report at the end that you can take to whoever signs.