Insurance · Warm Call
Warm Call Script for Insurance: Cashing a Peer Referral with a Claims Leader in 90 Seconds
A claims leader picks up your warm call knowing one thing: someone they respect at another carrier said your name. They cannot tell you what you sell. They probably skimmed the intro on a phone between a reserve review and a call with an agent who is threatening to move a $4M commercial book because a hail file has been open 41 days. That's the room you're walking into.
Warmth in this market is short-dated. Claims and underwriting leaders have been pitched by every analytics vendor with a demo deck since 2016, and a fair number of them have a shut-off fraud model sitting in a folder somewhere that drowned SIU in false positives. The referral buys you the pickup, roughly ninety seconds of suspended disbelief, and a permission slip to ask a direct question. It does not buy you interest, a problem, or a budget line — especially at a mutual, where every dollar of expense ratio is policyholder surplus and someone will say so out loud.
This playbook is a warm call script for Insurance sellers calling into carrier claims organizations — Chief Claims Officer, VP of Claims Operations, SVP Claims, Director of Claims Transformation, SIU Director, and the VP of IT who will inevitably get pulled in and ask where it touches the core. The job is to spend the referral fast, convert it into one testable reason you're relevant to their book rather than the referrer's, and get out with a dated second meeting that has a line of business attached to it.
The warm call script
Say it in your own words. The structure is the part that matters.
- 1
Pre-call: write the provenance line before you dial
One sentence, out loud, small and checkable: "[Referrer] mentioned you because [specific reason]." Good: "Marisol at Cornerstone Mutual pulled down our cycle-time back-test and said you'd taken over claims ops at Meridian in January when they consolidated the two claim centers." Dead: "Marisol thought you'd be interested in what we do." Also pre-decide three things: 1. **Exactly what the referrer said.** If Marisol said "you should call Dev," you say that. Not "Marisol said you'd really want to see this." Claims leaders at regional carriers sit on the same DOI working groups and the same carrier councils. They compare notes. Getting caught embellishing costs you two relationships in one call. 2. **Your relevance hypothesis, about their book, not the referrer's.** Built from a trigger: new in the seat, a Guidewire or Duck Creek migration announced, a rate filing on a personal auto line, an IA vendor RFP, a job ad for six field adjusters in the Gulf. 3. **The 15-second re-brief.** Assume the intro email was never read. Have a plain-language version ready that contains no product name. One more: know the season. If they took 4,000 hail claims three weeks ago, don't open with cycle time as an abstraction. Open with the surge.
- 2
The open — name, referrer, provenance, permission (25 seconds)
Four beats. No deviation, no warm-up. "Dev — Sam Whitfield, from Harlow. Marisol Reyes at Cornerstone suggested I call. She said you'd inherited claims ops at Meridian after the two centers merged in January, and that days-to-close on property was the first thing on your desk. She may have been overselling my usefulness here. Have you got four minutes for me to find out whether that's actually true, or should I come back?" When they say **"Right, Marisol said you'd call"** — that is the entire warmth budget being handed over. Do not spend it asking how they know Marisol. Five words and move: "Good, she flagged it. Then I'll be quick." What kills it here: "So just to give you a bit of background on us," "as per the email below," or any sentence about your company before you've said a sentence about theirs.
- 3
If they can't place the referrer
Happens constantly. Don't make them work for it. "No reason you would — it was a two-line note on a Friday. Short version: I work with claims orgs on the gap between when a file should close and when it actually closes — the property files that should be 12 days and sit at 28 because the adjuster is chasing a photo estimate, an appraiser assignment and a coverage question in three different queues. Marisol was dealing with that at Cornerstone. Worth four minutes, or not really?" No product name, no category name, no "AI-powered." Just the operational picture in their language. If it's not their picture, they'll tell you inside ten seconds and you've saved a week.
- 4
The relevance bridge — name the reason you might be irrelevant
This is the move the call lives or dies on. Referrer's world → the specific difference → a question that hands them control. "What Marisol was dealing with was a personal auto book where the average FNOL-to-close was sitting around 26 days and the DOI complaint count had roughly doubled year over year. That's her shop. You're mostly commercial property and farm, you've got a smaller field force, and from what I can tell your severity problem looks different from her frequency problem. So I honestly don't know if this lands with you. Where is your cycle time actually going right now — is it the adjuster's desk, or is it waiting on third parties?" Why this works: it proves you looked at their book, not the referrer's. Naming a reason you might be irrelevant is the fastest credibility move available on a warm call — it disarms the brace for the pitch. And it ends on a question about their operation, which is where you wanted to be. **Never** transplant the referrer's pain: "Marisol had a real leakage problem, so I imagine you do too." Different lines, different reinsurance structure, different reserving philosophy. They'll hear that you did no work on them.
- 5
Discovery — three questions, four at the absolute most
This is not a booked discovery call. You have four to twelve minutes and no agreed agenda. Eleven questions reads as an abuse of the referral and gets shut down with "can you just send me a deck." **1. Current state, mechanical.** "Walk me through a mid-size property file that goes sideways. Who actually touches it between FNOL and the check?" **2. Cost or friction, in their numbers.** "What's your average days-to-close on that line right now, and what did the board ask you about it last quarter?" Or, if the conversation has gone to severity: "When you look at ALAE per claim on that book — is that number moving, and in which direction?" **3. The priority test — the one that saves you a wasted meeting.** "Is that a this-year problem with a number attached to it, or a live-with-it problem?" **Optional fourth, high hit rate on warm calls because you're already a known quantity:** "Besides you, who'd have to care about this? Does your SIU Director own referral volume, or does that sit with you?" Listen for the correction. When they say *"honestly it's less cycle time, it's that we're leaking somewhere between two and five points on paid losses and nobody can tell me where"* — that's the call working. Write their exact phrasing down. It goes in the follow-up email and it becomes the second meeting's agenda line.
- 6
Reading the cool-off and naming it
Warmth in claims withdraws politely. The tells: - Answers shorten and get agreeable. "Yeah. Yeah, that tracks." - They narrate logistics: "Send me something and I'll get it to our innovation team." - They ask what it costs before you've established a problem. - They invoke the referrer as an exit: "Well, if Marisol rates you..." Stop talking and name it: "I'm getting the sense this isn't the live thing for you this quarter — which is completely fine, Marisol was guessing. Is it not the problem, or not the moment? If the pressure's coming from severity or from agent retention rather than cycle time, I'd rather know now than send you a deck you'll never open." A clean "not the problem" is a good outcome. It protects the referrer and it stops you burning a quarter on a ghost. And "it's actually retention — three of our top producers went to a carrier council meeting and hammered us on claims service" is a better opening than anything you had planned.
- 7
The close — a date, a length, a line of business, and a name
The number one failure is a pleasant call that ends in "send me some info." Friendly is not progressed. Close on their words, not yours: "Then here's what I'd suggest. You said the problem isn't the model, it's that the last one buried your SIU unit in false positives and got switched off. Give me thirty minutes and I'll show you how Cornerstone ran it — back-test only, last year's closed files, nothing touching the PAS, and the SIU Director set the alert volume before anyone saw a score. I'd want your SIU Director on that call, and if your VP of IT wants to sit in and ask where it touches the core, good — that's the right question and I'd rather answer it in week one. Thursday morning or Monday afternoon?" Must be present: **a date, a length, a reason built from their sentence, a named line of business, and who else attends.** If they genuinely can't commit: "Fine. I'll send you two paragraphs and one screenshot of the back-test output — no deck. Tell me Thursday whether it's worth thirty minutes." Then diarise Thursday on *your* side, not theirs.
- 8
Same-day loop-back to the referrer
Non-negotiable, routinely skipped, and the only reliable way a source produces a second referral. "Marisol — spoke to Dev, thanks for that. Their issue is different from yours; less cycle time, more that a prior fraud model got shut off and SIU won't touch another one without a back-test. Meeting him and his SIU Director Thursday. Appreciated the intro." Three jobs: thanks them, tells you whether you can use their name again, and keeps the channel open. Sources who never hear back stop referring — and in a market where the same forty claims leaders rotate through the same three conferences, that channel is most of your pipeline.
How the call actually sounds
Prospect on the left, the rep on the right.
Rep
Dev — Sam Whitfield, from Harlow. Marisol Reyes at Cornerstone suggested I call. She said you'd taken over claims ops at Meridian in January when the two claim centers consolidated, and that property days-to-close was the first thing on your desk. She may have been overselling my usefulness. Four minutes to find out whether that's true, or should I come back?
Buyer
Marisol, yeah. She mentioned something. Look, I'll be straight with you — I'm three weeks past a hail event in Kansas and I've got 3,800 claims in a queue built for about 400 a week. So four minutes, and then I'm gone.
Rep
Understood, and that's more useful to me than a calm Tuesday would be. Marisol's problem was personal auto — frequency, files sitting at 26 days, DOI complaints roughly doubling. That's her shop, not yours. You're commercial property and farm with a smaller field force, so I genuinely don't know if this lands. On the hail queue — what's the actual bottleneck? Is it adjuster capacity, or is it waiting on appraisers and estimates?
Buyer
It's both, but mostly it's that we've deployed about sixty independent adjusters at two-and-a-half times what an internal adjuster costs us, and I can't tell which of those 3,800 files needs a human on a roof and which one is a $4,000 shingle claim we could pay off a photo estimate. So everything gets the same treatment and everything takes the same time.
Rep
That's the specific version, thank you. What's your cycle time doing on that book right now versus a normal quarter?
Buyer
Normal is around 14 days for a mid-size property file. Right now I'd guess we're north of 40 and the complaints haven't even landed yet — those show up six weeks out. But before you go further: we tried this. Two years ago we bought an analytics tool that was supposed to triage and flag. SIU drowned in false positives, my adjusters stopped reading the alerts inside a month, and we turned it off. I'm not doing that again during a cat.
Rep
Fair, and I'd want to know why it failed before I'd argue with you. Two questions: what was the referral threshold set at, and did anyone in SIU get to choose the alert volume before it went live?
Buyer
Nobody chose anything. The vendor set the threshold, it fired on about eight percent of files, and my SIU Director had four investigators. You do the math.
Rep
So it was a capacity and threshold problem before it was a model problem. That's the failure mode I hear most. Here's what I'd propose instead, and it's deliberately small: no production connection at all. You give us last year's closed property files — the ones where you already know the outcome — and we score them backwards. Your SIU Director tells us the alert volume they can actually work in a week, and we size the threshold to that number, not to whatever the model wants to output. If it wouldn't have caught the confirmed files, you've lost nothing but an extract.
Buyer
Extract from where? Because this is the part where every vendor tells me it's simple and then wants to write back into the PAS. Our core is twenty-two years old, the two people who understand it retire inside five years, and the last integration that broke there took statutory reporting down with it. And we're mid-migration on top of that — there's no IT bandwidth until 2027.
Rep
Then let me be specific, because you're right to be scarred. Read-only, flat file against the closed claim file, nothing written back, no schema change, nothing near check issuance or the statutory extracts. And on the migration — that's a reason to talk now rather than later. Core migrations run two to three years and claims service usually degrades while they're running. What's your ALAE per claim done since the project kicked off?
Buyer
It's up. I'm not going to give you the number, but it's up, and my CFO has noticed. Doesn't change the fact that we're a mutual — I don't have a technology budget line like a stock carrier. Surplus is policyholder money and my board looks at expense ratio like a hawk.
Rep
Then I wouldn't frame it as tech spend, and I wouldn't bring you an ROI model with my assumptions in it. I'd bring it empty and let your chief actuary fill it with your paid losses, your LAE ratio, and whatever leakage number you'll admit to. The back-test calculates against actual settlements, not projections — which is the only version an actuary will sign. And I'd size the first phase to sit under whatever threshold doesn't need a board vote. When does your planning cycle close?
Buyer
End of October. Honestly, send me something and I'll route it to our innovation committee.
Rep
Happy to — though I'll say plainly that innovation committees are where claims vendors go to die, and I'd rather not waste your name on one. Before I send anything: is cat surge cycle time actually the thing you're being measured on this year, or is the real pressure coming from somewhere else — severity, or agent retention?
Buyer
It's the agents. Two of our biggest producers took us apart at the carrier council in June over how long farm claims sit. That's the conversation I'm actually in.
Rep
Then that's the meeting. Thirty minutes, back-test on last year's closed farm and property files, and the specific question is whether triage would have moved the files your producers complained about. I'd want your SIU Director on it, and your Head of Core Systems if he wants to interrogate the extract — I'd rather he did that in week one than week twelve. Thursday morning, or Monday after your queue meeting?
Buyer
Monday. After eleven. Bring the extract spec, not a deck — if it's a deck I'm cancelling.
Objections you will hear
What they say, and what you say back.
| Objection | How to answer it |
|---|---|
| “We run on a 20-year-old core. Integrations die there.” | Agree first, because they're right and they've been burned. "Fair — what broke last time?" Nine times out of ten a vendor tried to write back into the PAS. Then get specific about the surface: read-only against the claim file, flat file or message queue, no schema change, nothing touching check issuance or the statutory extracts. Then shrink the ask to the point of absurdity: one line of business, historical closed claims, no production connection at all until they've seen it score files where they already know the outcome. |
| “Every dollar goes through actuarial scrutiny. The ROI has to be bulletproof.” | Don't arrive with a calculator full of your own assumptions — a chief actuary will shred it in ninety seconds. Bring the model empty and ask them to fill it with their paid losses, their LAE ratio and their own leakage estimate. Then ask which number their actuary would actually book: cycle time reduction is hard to credit, but a measurable drop in ALAE per claim or a documented subrogation recovery lands in a reserve review. Offer the pilot as a back-test on closed claims so savings are calculated against actual settlements rather than projections. |
| “We bought a fraud analytics tool two years ago. SIU drowned in false positives and we shut it off.” | Ask what the precision was and where the referral threshold was set. Most of these failures are threshold and workflow, not model. Then invert the pilot: score last year's closed SIU-confirmed files and see whether it would have flagged them — and how many clean files it also flagged. Critically, let the SIU Director set the weekly alert volume their investigators can actually work, and size the deployment to that capacity rather than to whatever the model outputs. Referral rate only matters if the referrals get worked. |
| “We're mid-migration to a new policy admin platform. No IT bandwidth until 2027.” | That's an argument for talking now. Migrations run 24–36 months and claims service usually degrades while they're live — ask what cycle time and ALAE per claim have done since the project kicked off. Position whatever you sell as sitting outside the migration path so it isn't queuing behind the same integration team, and scope the pilot so the claims org owns it rather than IT. If it genuinely needs IT, say so and ask to be on the roadmap conversation for the quarter after go-live instead of pretending otherwise. |
| “We can't have a black box making claim decisions. We'd get destroyed in a market conduct exam.” | Concede the frame immediately: the model recommends, the adjuster decides, and every recommendation carries a reason code in the file notes. Then walk them through what an examiner would actually see — the audit trail, the disparate impact testing, and whether the model touches coverage decisions at all or only routing and prioritization. Best question available: "Does your compliance team already have a model governance standard?" If they do, get it and answer against their document rather than against a generic AI objection. |
| “We're a mutual. Surplus is policyholder money — we don't have a technology budget line.” | Reframe it as combined ratio, not tech spend. Every point of loss ratio and every dollar of expense ratio comes out of the same surplus, and slow claims cost them three ways: complaint handling, agent defection, and rate adequacy they have to claw back through a filing a regulator has to approve. Then get practical — "When does your planning cycle close, and what's the board's expense ratio target next year?" — and size phase one to fit under whatever threshold doesn't require a board vote. |
| “Send me something and I'll route it to our innovation team.” | Say the quiet part politely: "Happy to, though I'll be honest that innovation committees are where claims vendors go to die." Then anchor back to the operational number: "Before I send it — is cycle time actually what you're measured on this year, or is the pressure coming from severity or from the agent channel?" If they name a real pressure, book time with them and whoever owns that number. If they don't, you've learned this is a stall, and the right follow-up question is who does own days-to-close in the organization. |
Questions reps ask about this call
- How is a warm call script for Insurance different from a cold call script?
The referral buys you three things: the pickup, roughly ninety seconds of suspended disbelief, and permission to ask a direct question without earning it first. It does not buy you interest, a problem, or budget. In practice that means you skip the value proposition entirely and spend your opening on provenance — the specific, checkable reason the referrer named this person — then move straight to a question about their book. Two generic sentences and you've turned a warm call into a cold call the claims leader now feels mildly embarrassed to be on.
- What if the referrer works at a competing carrier?
Common in insurance, and it cuts both ways. Never repeat anything operational you learned from the referrer's shop — no cycle times, no leakage estimates, no loss ratio detail. Name the referrer and the general problem shape, then explicitly separate the two organizations: "That's her book, not yours — different lines, different field force, so I don't know if this lands." Claims leaders at regional carriers sit on the same DOI working groups and carrier councils. Discretion is the thing that keeps you referable.
- Which metrics should actually appear in the call?
Two, maybe three, and always as questions rather than claims. Average cycle time from FNOL to close by line is the one most claims leaders can quote from memory, along with the number their board asked about last quarter. ALAE per claim is the one a chief actuary will actually credit in a reserve review. If you're talking to an SIU Director, referral rate and dollars denied or recovered. If it's a mutual, expense ratio and combined ratio. Never quote a market statistic at them — ask for theirs.
- Should I call a claims leader during cat season?
Yes, but change the call. Three weeks after a hail event, a VP of Claims Operations is not thinking about a normal Tuesday — they're thinking about independent adjusters at two to three times internal cost and a complaint wave that lands in six weeks. Open by naming the surge and ask what the bottleneck actually is. Everything a carrier buys gets judged on whether it holds up during the surge, so a conversation held inside one is worth more than three held in March.
- How do I open when they clearly didn't read the intro email?
Don't make them remember. "No reason you would — it was a two-line note on a Friday." Then give a fifteen-second re-brief in their language with no product name in it: the property file that should close in twelve days and sits at twenty-eight because the adjuster is chasing a photo estimate, an appraiser assignment and a coverage question in three separate queues. Finish with "worth four minutes, or not really?" If that picture isn't their picture, they'll tell you immediately and you've saved a week.
- What counts as a successful outcome on a warm insurance call?
A dated meeting with a length, a named line of business, and at least one other attendee — typically the SIU Director, the Director of Claims Transformation, or the VP of IT who will ask where it touches the core. Bring that IT question forward rather than deferring it; it's the right question and you want it answered in week one. A clean "not the problem" is the second-best outcome: it protects the referrer relationship and saves you a quarter. "Send me some info" is a failure wearing a friendly jacket.