Wholesale operations

Submission triage: the appetite guide nobody reads

Why underwriters burn hours on risks the shop would never write, and what a triage pipeline that reads the appetite first looks like.

By Arcturus Labs · October 2026 · 9 minute read

Team working through a queue at a table

The queue grows faster than the desk. Triage decides which hours matter.

If you run a desk at an MGA or a wholesale broker, the last three years happened to you: E&S premium fled the admitted market, carrier appetites contracted monthly, and submission volume grew fifty percent while headcount grew not at all. The pile is the business. The pile is also the problem.

The most expensive hours in that pile are the ones spent on risks the shop was never going to write. This guide is about those hours.

The problem is not volume. It is mix

Every desk has a real appetite and a published one, and the published one lives in a PDF that retail agents do not read. So everything arrives: the contractor that fits, the contractor with three open losses, the habitational risk your carriers exited two years ago, and the account that was declined by three other wholesalers before breakfast.

Volume you can staff for, eventually. Mix you cannot: an hour on an unwritable risk is not a slower day, it is an hour taken from the risk that would have bound. And while the desk reads the pile, the retail agent who submitted the good risk is watching the clock, because their client is watching theirs.

What triage should do

Why the PDF fails

The appetite guide is a static document in a dynamic market. Carriers pull back from classes mid-year; guidelines shift with reinsurance renewals; the guide is updated eventually, distributed eventually, read never. Meanwhile the retail agent's incentive is to submit and see, because submission is free for them and expensive for you.

A triage system is the appetite made executable: the current rules, applied to every inbound, automatically, before a human hour is spent.

The desk math is unforgiving. Underwriters describe weeks where half the reading time went to risks that never had a chance. Recover those hours and the same desk binds more business with the same people. That is the whole pitch for triage.

The reputation effect nobody prices

Turnaround is the wholesale product. Retail agents quietly shop the shops that answer fast, and they say so, to each other, constantly. A three-day declination is not a neutral outcome: it is a broker deciding who to call first next time. Same-day service on both the yes and the no is the cheapest marketing an MGA can buy.

How we fit into this

We build the triage pipeline: submissions read on arrival, matched against the live appetite, ranked for the desk, declined same day when they are dead, with every reason captured. It starts with the measurement: what fraction of the pile was unwritable, and what those hours cost.

Frequently asked questions

What is submission triage? +

Sorting inbound submissions against the underwriter's actual appetite before human hours go to them: writable risks ranked and worked, unwritable ones declined the same day. It is the difference between a desk that reads everything and a desk that works what can bind.

How can an MGA reduce unquoteable submissions? +

Make the appetite executable instead of published. Triage every inbound automatically against current class, loss-history, and territory rules, decline the dead ones same day with a reason, and feed the decline reasons back to retail agents. Shops that answer fast and explain why get better-fitting submissions over time, because agents learn what sticks.

What is an appetite guide in insurance? +

A summary of what an underwriting shop will actually write: classes of business, loss tolerance, territories, and exposure limits. The failure mode is that it is a static document that agents do not read and that ages badly as carrier appetites shift. The useful version is not a document at all but rules applied automatically to every inbound submission.

How fast should an MGA decline a submission? +

Same business day, always. A fast no lets the retail agent shop elsewhere while the risk is still live, which is the service they actually remember. A slow no costs the agent the account and costs you the next submission from that agent. Declination speed is quoted-turnaround reputation, built one no at a time.

What percentage of submissions are dead on arrival? +

Commonly a quarter to half of inbound at wholesale shops, depending on how published the appetite is. The number itself matters less than the fact that most shops cannot state theirs, because nobody counts. Measuring the dead-on-arrival rate is the first step, and it is usually the most alarming number on the desk.

What fraction of last week's submissions were dead on arrival?

We measure the submission pipeline first: hours on unwritable risks, turnaround on the ones that matter. Then we build the triage. Ask us how it works.

Book a call