A quote takes 25 minutes and the prospect gives you 15
Personal lines runs 25 to 35 minutes of agency work, commercial runs 45 to 90, and a prospect with no proposal in 15 minutes has real odds of calling the next agency instead.
The short version
4 things that decide this
- 01Personal lines quotes typically take an agency 25 to 35 minutes of hands-on work; commercial lines run 45 to 90. About 40% of prospects abandon if they don't see a proposal inside 15 minutes, according to US Tech Automations' 2026 research, which calls the gap the single largest source of preventable revenue loss in agency new business.
- 02The time doesn't go to underwriting judgment. It goes to typing: re-keying the same submission data into the rater, then the AMS, then the proposal document, then sometimes accounting.
- 03Extraction from an ACORD form or a loss run can hand a producer a filled rater in under a minute instead of a blank one, without removing the coverage decision from a person.
- 04The fix isn't quoting faster by trying harder. It's cutting the number of times one submission gets typed by hand before a proposal exists.
Fifteen minutes is not a lot of patience
A prospect calls or fills out a form wanting a number. They aren't thinking about your rater, your AMS, or how many systems that submission has to pass through before a proposal exists. They're thinking about how long it's taking. Most of them stop thinking kindly about that somewhere around fifteen minutes.
US Tech Automations put a figure on it in 2026: about 40% of prospects abandon if no proposal shows up within fifteen minutes of asking. That's not a slow quarter, or a bad producer having an off day. It's the ordinary outcome of an ordinary process, and it happens at agencies that are otherwise running fine. Your quote gets done. It just doesn't get done before the prospect has already called the next name on their list.
Personal lines: 25 to 35 minutes. Commercial: 45 to 90
Personal lines quoting typically runs 25 to 35 minutes of your agency's hands-on time. Commercial runs meaningfully longer, 45 to 90 minutes, depending on the account's complexity. Neither number is mostly underwriting judgment. It's data entry, done more than once.
A submission arrives, often as an ACORD form or, on a renewal or a remarket, as a loss run attached to an email. Someone reads it and types the applicant's information into the rater. The rater spits out a number, and that number, with the applicant's details, gets typed again into your AMS to open the file. Then it gets typed a third time into whatever produces the proposal document your prospect actually sees. If the account binds, some of it gets typed a fourth time into accounting.
You've probably watched this happen at your own desk without naming it as the problem. It doesn't look like a bottleneck from the inside, because everyone involved is working the whole time. It looks like a bottleneck the moment you time it against your prospect's clock instead of your producer's.
The rater doesn't need to start blank
The fix that closes the fifteen-minute gap isn't a faster typist. It's skipping the repeat typing. Document extraction reads the ACORD form or the loss run the moment it arrives, and pre-fills your rater with what's already on the page. Your producer opens a quote that's most of the way done, not one that's empty.
That's a meaningfully different job than the one your producer does today. They check the extraction against the source document and make the actual coverage and pricing calls, instead of typing an address for the third time this hour. The rater still runs. The AMS still gets the file. What changes is how many times a person has to move data from one screen to another.
This is the same problem shape as reading a document and proving a number to a regulator, a shape we've built for before. A model reads the file. A person reviews what it found. The system only acts on what got confirmed. Extracting an ACORD form for a quote asks less of the model than an audit does, and it's held to the same rule. It reads. It doesn't decide.
- Submission arrivesACORD form or loss run, by email or portal
- ExtractionData pre-filled from the document, reviewed against the source
- Rater runProducer confirms coverage and pricing, not the typing
- Proposal generatedBuilt from the same confirmed data, not re-typed
- Producer bindsThe decision stays with the person who signs it
The producer still makes every coverage call. What changes is how many blank fields they're staring at first.
Faster intake isn't the same as automated underwriting
It's worth being precise about what changes here, and what doesn't. Extraction removes typing. It doesn't remove judgment. Your producer still decides which carrier fits, still checks the loss history against appetite, still makes the call on a marginal risk. What extraction removes is the fifteen minutes spent getting the data into a shape where that judgment can even start.
If you try to shortcut the judgment part instead of the typing part, you'll usually regret it. A rater fed bad data by an automated process with no review step produces a wrong number fast, which is worse than a right number slow. The point isn't to remove your producer from the quote. It's to stop asking them to do the part of the job a computer reads faster and more accurately than they type.
Some of the systems we have shipped
Related questions
01Does extraction replace the rater or the AMS?+
No. It feeds them. Your rater and your AMS stay exactly as they are. Extraction just means the fields inside them get filled from the source document, instead of typed a second time by a person.
02What if a submission has messy or incomplete data?+
Your producer reviews the extraction against the actual document before the rater runs, so a messy submission gets flagged for a human to complete rather than silently guessed at. That review step is what keeps a bad submission from producing a confident wrong number.
03Is this only useful for high-volume personal lines?+
Commercial submissions benefit at least as much. They carry more fields and more attached documents like loss runs, and the 45-to-90-minute range gives you more time to recover per quote.
04Does this help renewals too, or only new business?+
The same extraction applies to a loss run on a renewal or remarket, which is often the slower half of a renewal cycle. A renewal review that starts from pre-filled data leaves more time for the actual remarketing decision.
Related
- AI, automation and custom software for accounting, insurance and staffing firms →The wider build across intake, quoting and renewals.
- Workflow automation for firms →Close checklists, renewal sequencing and the recurring work automation removes.
- How to cut quote turnaround in an insurance agency →The direct steps, beyond just the diagnosis.
