The appetite graph: why it changes close rate by 38%

Submission speed is table stakes. Every insurtech tool on the market promises to cut submission time. Most of them can. But faster submissions alone don’t grow your book — speed-to-the-wrong-market is just a faster way to hear no.

The piece that actually changes close rate is appetite. Which carriers will quote this risk today, and why. That’s the number. Everything else is a distraction.

What “appetite” actually means

Every commercial carrier has an appetite — a shape of risk they want. Class of business, revenue band, geography, loss history, TIV limits, roof age on habitational, years in business, number of employees, the list goes on. It’s written down in bulletins and underwriter guides. It changes quarterly. Sometimes monthly.

In a typical 20-producer agency, that knowledge lives in three places:

• The head of your most senior CSR (usually accurate, always a bottleneck)
• A binder in the desk drawer (usually 6 months stale)
• Your wholesale broker’s memory (good for E&S, not for standard)

When a new ACORD comes in, the CSR best-guesses six carriers. Maybe four. Often she picks one market that declined last year and doesn’t know yet because the producer never told her. The account goes to six markets; three reply with “out of appetite” three days later. Two of those declines could have been predicted.

What an appetite graph does

An appetite graph is a live, structured model of every carrier’s current appetite across the entire US commercial market. Class codes × revenue bands × geographies × loss thresholds × underwriter preferences. Refreshed from:

• Published carrier bulletins and underwriting guides
• Wholesale broker appetite memos
• Your agency’s own declination history (carriers that declined you last year in this class)
• Every other pilot agency’s declinations (anonymized, across-network intelligence)

When you drop an ACORD into Bindflow, the account is scored against every carrier in the graph. You get a ranked list with reasons. Not a black-box recommendation — a list you can read and argue with.

A concrete example

Joe’s HVAC. Phoenix, AZ. $2.4M revenue. Five employees. Clean 5-year loss runs. Currently with Hartford, shopping because producer sees better pricing elsewhere.

Here’s what the graph says:

The CSR submits to four carriers. Three of those were already on her best-guess list. One (CNA Connect, score 84) she wasn’t going to include — and that’s the one that quotes aggressively. Bound in 9 days at 14% below Hartford’s renewal.

That’s one account. Do this across 180 new accounts a year at a 15-producer shop, and the quote-to-bind ratio moves from 35% to somewhere in the 45-55% range. Which on a book where average commission is ~$4,000 per bound policy, is $100-150K of recovered commission per year. Per agency.

Why most agencies don’t have this

Because the data is everywhere and nowhere. Bulletins come in PDFs. Wholesale memos come by email. Carrier portals change quarterly. Keeping a structured, accurate, agency-specific appetite model alive was a full-time job nobody was going to hire for at a 20-producer shop.

LLMs changed that. Ingesting bulletins, normalizing them against a class-code ontology, enriching with your declination history — all of that is now tractable with small teams and careful engineering. The Bindflow appetite graph ships with 60+ standard and E&S markets preloaded, and learns from your book in the first 30 days of use.

What the math suggests

Bindflow is pre-pilot — the first 10 agencies will produce the first real case studies. What industry data and existing research on submission-ops improvements suggests for tools in this category:

• +20-45 percentage points on quote-to-bind when submissions only go to markets actually in appetite
• ~2 fewer wasted submissions per account — producer stops spraying markets that were never going to quote
• 30-50% shorter time to first quote because carriers in appetite respond faster to clean submissions

These aren’t guarantees — they’re the thesis. The pilot structure (bindflow.co/pilot) includes a money-back refund if your numbers don’t move. The principle is simple: if you’re selling the right risk to the right market, the market says yes more often. The “right market” is a solvable data problem. The tool just makes it visible.

— Tom Hwang, founder, Bindflow