Compressing FNOL-to-payment cycle time from 14 days to 36 hours

Every carrier we work with publishes the same internal goal somewhere on a slide: shorten cycle time. The number on the slide is always smaller than the number on the wall. The reason is not that adjusters are slow. The reason is that almost no time in a claim file is spent doing claims work. Most of it is spent waiting.

If you instrument a representative auto or property file from FNOL to payment and break it into time buckets, the picture is consistent: roughly one to three hours of active human handling, distributed across fourteen calendar days. The other 333 hours are queue time, vendor lag, document chase, internal handoffs, and waiting for systems to update. Those hours are the prize.

The anatomy of a 14-day claim

Most carriers measure cycle time as the wall-clock distance from FNOL to closure or to first payment. That single number tells you very little. Decompose it and the picture becomes operationally actionable. A typical mid-sized auto claim breaks down roughly like this.

1. Hour 0–4: FNOL is taken, basic facts are captured, and the claim is assigned. This bucket is largely fine on most books.
2. Hour 4–48: The first real lag. Coverage verification waits for someone to open the file. Photos and supporting documents trickle in from the insured. The DRP or shop receives the assignment but does not action it the same day.
3. Day 2–5: The estimate arrives. It is reviewed in batches because nobody reviews estimates one at a time. Supplements are requested and re-reviewed.
4. Day 5–9: Fraud and subrogation review, if anyone touches them, happen here. In most carriers, they happen on a sample of files and almost never on the full population.
5. Day 9–13: Reserve adjustments, internal approvals, payment staging.
6. Day 13–14: Payment is issued. The file is moved to closure.

The reason most cycle-time initiatives fail is that they attack the wrong bucket. Carriers buy faster intake tools and shave four hours off a workflow that already only takes four hours. The fourteen days do not move because the fourteen days were never about intake.

Where agents actually collapse time

Layerup agents do not replace adjusters and they do not speed up the moments adjusters are already in the file. They collapse the moments adjusters are not in the file. There are four big ones.

What changes when you do this

When the queue time and handoff time get squeezed simultaneously, the cycle time number on the wall starts moving — not by ten percent, by an order of magnitude. The carriers we work with see initial payment cycles compress from days to hours on the workflows they have deployed agents into. The improvement is not because anyone is working faster. It is because nobody is waiting.

Cycle time is the headline. The second-order effects are the prize.

Cycle time is what executives ask about first. It is rarely what they care about most. What they actually care about is the chain reaction.

• Loss adjustment expense per claim falls because the per-file handling time is lower and the BPO and overflow spend shrinks.
• Leakage falls because every estimate and every invoice is reviewed instead of a sample, and because reserves adjust on evidence as it arrives.
• Customer experience improves because the gap between FNOL and substantive action is hours, not days.
• Capacity to handle catastrophe surge improves because agent capacity is elastic and does not require hiring against a once-in-three-years event.
• Adjuster retention improves because the work that remains for adjusters is the part they signed up for: judgment calls, coverage interpretation, negotiation. Not document chase.

What not to do

A few patterns predictably do not move cycle time, and we have watched carriers spend years on each.

1. Buy a faster intake tool when intake is already four hours of a fourteen-day file. The hours are not in intake.
2. Migrate to a new policy admin system and assume cycle time will fall as a side effect. The system rarely was the bottleneck. The waiting was.
3. Layer a BPO under the existing process. The BPO inherits the same queue dynamics and adds a handoff. Cycle time stays flat. Cost per claim goes up.
4. Add an LLM that summarizes documents. Summarization is useful only if a human was already going to read the documents. The waiting was upstream of the reading.

What actually works

The pattern that works is small and repeatable. Pick one line of business. Pick the largest queue inside it. Deploy an agent that drains that queue continuously. Measure cycle time before and after, on the same cohort, with the same definitions. Then pick the next queue.

Most of our carrier deployments start with either FNOL coverage verification or estimate QA, because those are the two queues where the time math is most lopsided. Both can be live in weeks, not quarters, and both produce a defensible cycle-time number on the first cohort.

Thirty-six hours, not ten days

Thirty-six hours, FNOL to first payment, is achievable on the right cohorts in auto and property today. It does not require a core system replacement, a BPO overhaul, or a wholesale change in how adjusters work. It requires draining the queues, one workflow at a time, with agents that operate inside the systems the carrier already runs. That is the work.