Insurance · Renewal Call
Insurance Renewal Call Script: Saving a Claims Platform Account at a Carrier That's Already Halfway Out
Six weeks out. The VP of Claims Operations who signed last year has taken heat internally for it, your platform went quiet for nine days in the middle of a April hail event that dumped 4,100 claims into a queue built for 400 a week, and a competitor rep has been sitting in their inbox with a number 27% lower. They took this meeting to say out loud what went wrong. If you open with the order form, you confirm the thing they already suspect — that you only show up when money is due.
This is not a renewal call. It's an autopsy with a signature at the end of it. Carriers have long institutional memory about vendors who disappeared during cat season, and claims leaders talk to each other at every industry council meeting. The good news is that the same person who is 80% gone will start negotiating with you the moment you say the ticket numbers out loud before they do — including the ones that make you look bad. Claims people spend their whole careers dealing with people who minimise. Being the vendor who doesn't is a genuine differentiator.
The numbers in the script blocks are bracketed — fill them with the real ones from your account before you dial. The numbers in the sample dialogue are an illustrative account so you can hear how it sounds at pace. Everything below assumes you sold a claims triage, SIU scoring, or subrogation identification platform into a mid-size carrier, that adoption stalled outside one team, and that your support org had a bad Q2.
The renewal call script
Say it in your own words. The structure is the part that matters.
- 1
Pre-call: the renewal autopsy (claims edition)
Do not walk into this with a CRM summary. Pull the raw material and read it. **Support:** every ticket for 12 months — ID, open date, first response, resolution, severity. Flag anything over 24 hours to first touch. Know your three worst by number and date, and know which of them landed during their cat event. A 60-hour first response on a normal Tuesday is a miss; a 60-hour first response on day five of a hail event is the thing they will still be telling other carriers about in three years. **Adoption by desk, not by account:** not '38% adoption.' It's 'SIU is at 92% weekly active on 11 seats, the auto desk peaked at 31% in February and has been at 18% since, and property never got trained at all.' Break it out the way they break out loss ratio — by line. **Your own churn:** CSM and AE changes on your side. If they've had two CSMs in fourteen months, you name it before they do. **Broken QBR promises:** the salvage module you said was Q3. The property desk training rescheduled twice. Write them down verbatim from the deck. **Their org:** did the Chief Claims Officer change? Did the person who signed still own the budget after the reorg? Is there a new Director of Claims Transformation who owns vendor consolidation now? Is the SIU Director — probably your only real power user — even invited to this call? **Their year:** pull whatever you know about their cat activity, their DOI complaint trend, and whether they've filed for rate. If they took a bad hail season, their combined ratio conversation is different and your renewal is competing with independent adjuster spend.
- 2
Internal alignment before you dial
Get written confirmation from your own side on three things, because a second broken promise ends this account permanently and it will be attributed to you personally: 1. **What support actually changed** — new tier, named engineer, segmented queue, and specifically what the surge protocol is when a customer's claim volume spikes 10x. Get the date it took effect. 2. **What CS resource you can commit** — named person, actual hours, for the adjuster desk that never got trained. 3. **Your floor** — on price, term, and seat count. Decide now whether you'd rather right-size to 40 real seats than discount 160 fictional ones. Concessions you invent on the call feel like guilt money and they read as guilt money. And clear one thing before the meeting: close their open tickets. Do it Thursday, mention it Monday in passing.
- 3
Opening: you go first, with numbers
'I know the contract's up on [date]. I'm not here to talk about the renewal yet — I want to talk about this year first, because from where I'm sitting it wasn't the year we sold you. Here's what I can see. Ticket [INC-####] went in on [April 14], during your hail event. First response was [61 hours]. It took nine days to resolve and your nightly claim extract wasn't scoring for six of them — in the week you needed it most. Between April and June your median first response from us went from about four hours to [38]. You've had two CSMs in fourteen months, both changes came from our side. We committed to the [salvage/total-loss] module in the Q3 QBR and it didn't ship. And the adoption number is the one that bothers me most: SIU is at [92%] weekly active, and the auto desk hasn't been above [31%] since February. You're paying for [160] seats and roughly [50] people use it. That's what I can see from my side. What am I missing, and what did it actually cost you?' Then stop. Do not fill the silence. The next ninety seconds decide the meeting.
- 4
The listening phase: let it be worse than you thought
Budget 40–50% of the call here. On a 30-minute call, that's fifteen minutes minimum. No 'but.' No 'to be fair.' Do not defend the ticket where their own IT team sat on the firewall change for four days — you may be right and you will lose the renewal being right. Ask what it cost them **politically**, not just operationally: - 'Who did you have to explain this to?' - 'Did this come up in the board deck, or just in your leadership review?' - 'Did any of those files turn into DOI complaints?' - 'What did your days-to-close do in April and May?' - 'Did any of this get back to the agents? Did it come up at the agent council?' Follow the emotional word. When they say 'we were flying blind through the surge' — 'tell me about that week.' Write it down visibly and read it back: 'So: the outage during the hail event, the Q2 response times generally, the auto desk rollout that never happened, and you found out we were down from your own adjusters rather than from us. Anything else?' Ask 'anything else?' twice.
- 5
One apology, specific, no hedging
'The Q2 response times were our failure. We reorganised support in March and your tickets got queued behind two enterprise escalations. That's an explanation, not an excuse — you bought a 24-hour response SLA and we missed it [eleven] times, and we missed it during the six weeks of the year when a claims org has the least tolerance for it.' One apology. Specific. Then move. Do not apologise again later in the call — repeated apology reads as a negotiating posture. **Do not offer a discount here.** A price cut is not an apology. It reframes a service failure as a pricing dispute, tells them your list price was always soft, and permanently resets your floor. If you owe them something, service credits for the affected months are the honest instrument — finite, specific, dated.
- 6
Proof of change: show, don't promise
They've heard 'we're really investing in support.' Only three things carry weight: **1. Structural change with a date and their own data.** 'We hired [14] support engineers in August and split the queue by segment. Since September 1 your median first response is [3.1] hours across [14] tickets — here's the list, you can check it against your own inbox.' Their numbers, not a company-wide average. **2. A carrier like them who had the same problem and stayed.** Ideally another regional carrier that went through a cat surge with you afterwards. Offer the call, don't just cite it: 'Their VP of Claims Operations will take twenty minutes with you — I'll make the intro today.' **3. Something already done, unasked.** 'I had the team clear the three open tickets in your queue last week — you'd have seen those close Thursday.' **And the one they'll actually test you on — the cat protocol:** 'The thing I can't undo is April. What I can tell you is what's different for next hail season: [named engineer] is assigned to your account, there's a volume trigger on your extract that pages our on-call when your daily FNOL count goes above [X], and we're writing a four-hour P1 response into the contract with a credit attached. If we miss it, you get paid, without asking.' If nothing has genuinely changed, say so, and change what you *can* control — your own coverage, a named escalation path, an executive sponsor. Do not manufacture progress. They will check.
- 7
Rebuild on what works: find the person, not the feature
They still like one or two things. Usually it's the subrogation queue or the SIU scoring, and usually it's because one person built a workflow on top of it. 'You said the subro identification queue is the one thing your team would miss. Who's actually running that, and what does it replace?' Quantify it in their language, and make them say the number: - 'Where was your subrogation recovery rate on auto before we turned it on, and where is it now?' - 'What's your SIU referral rate this year versus last? What's the dollar figure on files you denied or reduced off the back of a flag?' - 'What's ALAE per claim doing on the files that get scored versus the ones that don't?' Then ask the switching-cost question directly: **'If we went away Monday, whose job breaks?'** That person — usually the SIU Director, sometimes a subro manager — is your remaining champion and almost certainly wasn't invited to this call. Get them in the next one. This is your quiet competitive defence. The switching cost isn't the licence fee. It's rebuilding two years of referral thresholds and a subro workflow that one investigator tuned by hand, during the run-up to storm season.
- 8
Handling the competitor
Surface it yourself: 'I'd assume you've had [Competitor] in the inbox. I'd be surprised if you hadn't — what did you like about them?' Ask genuinely, then find out how far it's gone: - 'Have they done a demo with your SIU team, or just with you?' - 'Is the pricing in writing? Has procurement seen it?' - 'Have they been through your security and model governance review yet? Who signed off?' - 'Have they scoped the extract work with your VP of IT, given you're mid-migration?' A lunch is a different conversation than a signed order form sitting in legal. Reframe from price to total switching cost, and say it plainly: 'The licence delta is [$X]. What I'd want you to weigh against it is: re-tuning referral thresholds from scratch with your SIU Director, a new extract build with an IT team that's already committed to the core migration, six months where your subro queue is worse than it is today, and — honestly — the risk that you inherit a new vendor's version of the same support problem with none of the goodwill and no track record of how they behave in a surge. Ask them what their P1 response was during last year's storm season. Ask for the ticket data, not the SLA document.' **Do not match price here.** Price is the last section of this call.
- 9
The 90-day plan: build it live, in their words
This is the artefact that changes the decision. Four fields per line: **what, who by name on both sides, by when, and what 'done' looks like.** - **Adoption:** onboarding for the [110] auto adjusters who were never trained — run by [CSM name], desk by desk, weeks of [dates]. Done = [60] weekly active by day 60, measured in the usage report, not by feeling. - **Threshold reset:** [SIU Director] and [our data lead] re-tune the fraud score to the alert volume SIU can actually work — target [40] referrals a week, not 400. Done = a written threshold document signed off by SIU by [date]. - **Support:** named engineer [name], direct line, four-hour P1 response, effective at signature. Cat surge trigger on their FNOL volume. - **Governance:** 30/60/90 reviews, written usage report each time, first one dated [date], attended by [Chief Claims Officer] not just the VP. - **The thing you can't do:** 'The write-back into your PAS is not on the roadmap before Q3 next year. I'm not going to pretend otherwise, and I don't want you renewing on the assumption that it is.' **Ask them what they'll own.** 'A plan where only the vendor has obligations is a wish list — and part of why adoption stalled is that nobody on your side owned the auto desk. Who's that going to be, and can they mandate it or only suggest it?' **Tie a commercial consequence to it:** a break clause at six months, quarterly review, or SLA credits written into the contract. Signing a credit is worth more than a discount, because it proves you expect to be held to it.
- 10
Only now: the commercial conversation
Earn the transition out loud: 'If we can agree that plan is real, can I walk you through what renewal looks like?' **Structure before number.** - Right-size the seats to actual usage instead of discounting the whole thing: 'You're at [50] real users. Let's renew [70] — the [50] plus the property desk we're training — and put step-up pricing on the rest. I'd rather be right than big, and it kills the shelfware conversation permanently.' - Offer a shorter term if they need the comfort, longer term if they want price protection. Name the trade. - **Every discount is traded:** a reference call with another carrier's claims leader, a case study on the subro numbers, a multi-year, an earlier signature, or a meeting with their Chief Claims Officer and your exec sponsor. - If they've had a genuinely bad year, offer **service credits for the affected months** — specific, finite, and they don't reset your floor forever. For the mutual carrier, speak in combined ratio, not tech spend: 'Every dollar of this sits in the expense ratio, and every point of leakage or lost subro sits in the loss ratio — same surplus either way. Where does the board's expense target land next year, and what's the threshold that doesn't need board approval for phase one?'
- 11
Multi-thread and close
Six weeks is tight and most stalled renewals die in procurement, not in the business. - 'Walk me through what happens between now and [date]. Who signs? Does this go to procurement, and have they already got [Competitor]'s number? Does legal need to re-paper anything? Does the model governance review have to be refreshed for a market conduct standpoint?' - 'I want [Chief Claims Officer] and [SIU Director] in the plan review — the person who owns the number and the person who uses it every day. Can we do [specific date and time]?' Bring your exec sponsor. An exec-to-exec apology lands differently than yours. - **Book the next meeting on the call.** Never 'I'll follow up.' - **Temperature test, honestly:** 'On a scale where 1 is you've already decided to leave and 10 is you're signing — where are we right now, and what moves it a point?' Take the honest answer over the polite one, and if they say 3, ask what a 6 looks like. **Within 24 hours, send four things:** the written 90-day plan with names and dates, the ticket data you quoted, the reference intro, and a one-page summary of what they told you went wrong — in their words, not softened. That last document does more work than any proposal you'll ever send.
How the call actually sounds
Prospect on the left, the rep on the right.
Rep
Before we get anywhere near the renewal — I want to talk about this year, because it wasn't the year we sold you. Ticket INC-3319, opened April 14, day five of your hail event. Our first response was 61 hours. The nightly extract wasn't scoring for six days while you were taking 4,100 claims in a week. Q2 median first response from us was 38 hours against a 24-hour SLA — we missed it eleven times. Two CSM changes, both ours. And the auto desk never got above 31% weekly active. You're paying for 160 seats and about 50 people log in. That's what I can see. What am I missing?
Buyer
You're missing the part where I had to sit in front of our Chief Claims Officer in January and explain why we're paying six figures for something 18% of the auto desk opens. And then April happened. Our days-to-close on auto went from 14 to 26 in six weeks. The board asked about it. I don't get to say 'the vendor's extract was down' in a board deck — it's my number.
Rep
Tell me about April. What did the week of the 14th actually look like on the floor?
Buyer
We had independent adjusters out at two and a half times internal cost, files stacking up, and the one thing that was supposed to help us triage severity and route the simple ones straight through was just... not scoring. So everything went to a human queue. We ate the LAE. Then the DOI complaints came in six weeks later, like they always do, and we were up eleven year over year. And my agents — I had two of our biggest independent agencies call me directly about specific files. That's the part that actually threatens the book.
Rep
So: the outage during the surge, the Q2 response times generally, the auto desk rollout that never happened, and you heard we were down from your own adjusters instead of from us. Anything else?
Buyer
The salvage module. You committed to it in the Q3 QBR. I put it in a slide for my COO. It doesn't exist.
Rep
You're right, and I'm not going to soften it. The Q2 failures were ours — we reorganised support in March and your tickets queued behind two enterprise escalations, which is an explanation, not an excuse. And the salvage module slipped two quarters and nobody called to tell you before you'd already put it in front of your COO. That's on me personally. One apology, and then I'd rather spend the time on what happens next. Can I ask about the piece that did work?
Buyer
Look, the apology's fine, I appreciate it. But I've got a quote sitting in my inbox that's 27% under your renewal and they're offering to cover the migration. Ray in SIU likes your subro queue, sure. He's one person on a six-person unit. That's not a business case.
Rep
Ray built something, though. What was your subrogation recovery rate on auto before he turned that queue on, and where is it now?
Buyer
...It's up. Call it a point and a half on paid losses, maybe a bit more. And the SIU referral rate went from about 1.1% to just under 2%, though I'd argue some of that's Ray being Ray. But my chief actuary won't book any of that as savings on a renewal justification. Cycle time doesn't show up in a reserve review. Recovered dollars do, and he'll want them attributed.
Rep
Then let's give him something he can attribute. I'm not going to hand you an ROI model with my assumptions in it. I'd rather run a back-test on closed files — take last year's auto BI settlements and the subro referrals, and show what the queue identified versus what got recovered, against actual settlements, not projections. If your actuary can't defend that number, I don't want you renewing on it.
Buyer
That's fair. But here's the real problem. We're eighteen months into a core migration. My VP of IT has told me flat out: nothing new touches the claim extract until we're through go-live. He's already got two integrations that broke and took statutory reporting down for a day. If your renewal means IT work, it's dead before it starts.
Rep
It doesn't, and I'd want him on the call to confirm that rather than take my word for it. We're read-only against a flat file. We don't write back into the PAS, we don't touch check issuance or the statutory extracts, and there's no schema change. Which is exactly why I'd argue this is the wrong thing to drop during a migration — claims service usually degrades during one. What has your cycle time done since the project started?
Buyer
It's gone the wrong way. I'll give you that. But there's another thing and it's the one that actually killed adoption. In May the auto desk turned your alerts off. Too many hits. Adjusters were getting flagged on clean files, they stopped looking, and once an adjuster decides an alert is noise you never get them back. That's the second time we've been through this with an analytics vendor.
Rep
That's a threshold and workflow failure, not a model failure, and it's one we should have caught in a QBR instead of letting it die quietly. Here's what I'd put in the plan: Ray sets the alert volume. Not us. He tells us how many referrals a week his unit can genuinely work — if that's forty, we size to forty, and we tune the threshold until precision holds at that volume. Then we retrain the auto desk to it, desk by desk, not one webinar. Would Ray sit in that session and own the number?
Buyer
He would. He'd enjoy it, frankly. But I need to know what happens next hail season, because that's what I'll get asked. Not what happens on a normal Tuesday.
Rep
Fair, and that's the only question that matters. Three things, all dated: Marcus is your named support engineer as of September 1 — your median first response since then is 3.1 hours across fourteen tickets and I'll send you the list. There's a volume trigger on your feed that pages our on-call when your daily FNOL count goes above your normal range, so we know you're in a surge before you tell us. And I want a four-hour P1 response written into the contract with a credit attached, so if we miss it during a storm you get paid without having to chase me. I'd rather sign a credit than give you a discount — the discount just tells you I was overcharging.
Buyer
We're a mutual. Surplus is policyholder money. Signing anything means I'm defending the expense line, and 27% cheaper is a very easy sentence for someone to say in a planning meeting.
Rep
Then let's not renew 160 seats. You've got about 50 real users and the property desk we haven't trained yet. Renew 70, step-up pricing on the rest as adoption proves out, and the credits for April through June come off this term. That gets your expense line down without me pretending the last twelve months were fine. What's the threshold that doesn't need board sign-off for phase one?
Buyer
Under a hundred and fifty and it stays with me and the Chief Claims Officer.
Rep
Then here's what I'd like. I'll have the 90-day plan to you tomorrow with names and dates on both sides, including the two things I can't deliver, written down. I want thirty minutes with you, your Chief Claims Officer, and Ray on the 19th to walk it through — my VP of Customer Success will be on it, because the Q2 failure was his org and he should say that to your CCO himself. Before that: on a scale where 1 is you've already decided to leave — where are we?
Buyer
Four. Maybe a five because of the subro numbers. Get the back-test in front of my actuary and get through the 19th without another surprise and I'll tell you where it goes.
Rep
Four is honest and I'll take it. One more thing — does this route through procurement, and do they already have the other quote?
Buyer
They've had it since October.
Rep
Then I need to be in front of them by the 26th, not in week five. Who's the contact, and can you make the intro after the 19th?
Objections you will hear
What they say, and what you say back.
| Objection | How to answer it |
|---|---|
| “You went dark for six days during our hail event. Why would next storm season be any different?” | Don't argue with the premise and don't lead with the SLA document — they've read SLA documents from vendors who then disappeared. Give them the structural change with a date and their own data: when the support reorg happened, who their named engineer is now, and what their median first response has actually been since. Then give them the surge-specific mechanism: a volume trigger on their FNOL count that pages your on-call before they call you, so being in a cat event is something you detect rather than something they report. Close it commercially: 'I want the four-hour P1 written into the contract with a credit attached. If we miss it during a storm you get paid without chasing me.' A signed credit is worth more than a discount because it proves you expect to be held to it. |
| “We're eighteen months into a core migration. My VP of IT says nothing new touches the extract until after go-live.” | Agree with the IT VP before you answer, because he's right and something has already broken on him — usually a vendor that tried to write back into the PAS and took statutory reporting down with it. Get specific about your integration surface: read-only against the claim file, flat file or message queue, no schema change, nothing touching check issuance or statutory extracts. Then invert the objection: 'Core migrations run 24–36 months and claims service almost always degrades during them. What's your days-to-close done since the project started?' If cycle time has gone the wrong way, dropping the one thing that triages severity mid-migration is the risky move. Offer to have the VP of IT on a fifteen-minute call to confirm the surface himself rather than taking your word for it. |
| “My chief actuary won't book cycle time as savings. If I renew, I need a number he can defend in a reserve review.” | Never hand a carrier an ROI model with your own assumptions in it — it will get taken apart line by line and you'll lose credibility on the arithmetic before you lose it on the product. Bring the model empty and let them fill it with their paid losses, their LAE ratio and their own leakage estimate. Then move the claim to something bookable: cycle time is soft, but ALAE per claim on scored versus unscored files, and documented subrogation recoveries, land in a reserve review. Offer the back-test: run last year's closed files, show what was identified against what was actually recovered, so the number is calculated on real settlements rather than projections. Say out loud: 'If your actuary can't defend that number, I don't want you renewing on it.' |
| “The auto desk turned your alerts off in May. Too many hits on clean files and the adjusters stopped looking.” | Concede it fast — and concede that you should have caught it in a QBR rather than letting it die quietly, because that's the part that's actually your fault. Then diagnose properly: ask what the threshold was set at and what precision looked like at that setting. Most of these failures are threshold and workflow, not model. The fix belongs to their SIU Director: he tells you how many referrals a week his unit can genuinely work, you size the threshold to that capacity, and you retrain the auto desk desk-by-desk rather than with one webinar. Put his name and the target volume in the 90-day plan with a signed-off threshold document as the definition of done. Adoption failed partly because nobody on their side owned it — this is how you fix that without saying it. |
| “I've got a quote that's 27% under your renewal and they'll cover the migration.” | Never disparage, and ask what they liked — genuinely. Then find out how far it's gone: demo with the SIU team or just with the VP, pricing in writing, security and model governance review completed, and critically, has anyone scoped the extract work with the VP of IT mid-migration. A lunch is a different conversation than a signed order form. Reframe from licence delta to total switching cost, plainly: re-tuning referral thresholds from scratch, a new extract build competing with the core migration, six months where the subro queue is worse than today, and inheriting a new vendor's version of the same support problem with no goodwill and no track record. Then the killer question: 'Ask them what their P1 response times were during last year's storm season. Ask for ticket data, not the SLA document.' Do not match on price in this section. |
| “We're a mutual — surplus is policyholder money. I have to defend this expense line and cheaper is an easy sentence in a planning meeting.” | Move it out of tech spend and into combined ratio, where they live. Every dollar of this sits in the expense ratio; every point of leakage, missed subro or unnecessary ALAE sits in the loss ratio — same surplus either way, and slow claims also cost them in complaint handling, agent defection and rate adequacy. Then help them win the internal argument instead of arguing with them: right-size seats to actual usage rather than discounting the whole thing, apply service credits for the months you failed them, and ask directly what the approval threshold is that keeps phase one with them and the Chief Claims Officer instead of going to the board. A smaller, honest renewal you can grow beats a discounted one that makes the shelfware problem permanent. |
| “Send me the plan and I'll route it to our innovation team.” | The innovation team is where claims vendors go to die, and on a save call it's usually a polite way of saying no. Acknowledge it and anchor back to the operational number: 'Happy to — before I do, is cycle time still the pressure this year, or has it moved to severity or retention?' If they name a real pressure, book time with them and whoever owns that number. On a renewal specifically, add the multi-thread ask directly: you want the Chief Claims Officer and the SIU Director in the plan review, because the person who owns the number and the person who uses it daily are the two people whose absence killed this account the first time. And ask whether procurement already has the competitor's quote — the answer is usually yes, and usually since months ago. |
Questions reps ask about this call
- Is this the same as a policy renewal call script for policyholders?
No. This is a vendor-side save script for renewing a software or analytics contract with an insurance carrier — a claims triage, SIU scoring, subro identification or core-adjacent platform. The buyer is a Chief Claims Officer, VP of Claims Operations, SVP Claims, Director of Claims Transformation or SIU Director, not a policyholder. If you're looking for a script to retain a homeowners policy at renewal, this isn't it, though the underlying principle — acknowledge the bad claim experience specifically before you talk price — travels surprisingly well.
- How much of the call should be spent on what went wrong before I talk about the renewal?
Forty to fifty percent. On a thirty-minute call, that's fifteen minutes of listening minimum, and you open the call by naming the failures yourself with ticket numbers and dates rather than asking 'how's it been going?' Claims leaders spend their working lives around parties who minimise damage. The moment you say the 61-hour first response out loud before they do, you stop being a vendor managing a renewal and start being someone they can negotiate with. Earn the transition explicitly: 'If we can agree that plan is real, can I walk you through what renewal looks like?'
- They're mid-migration to a new policy admin system and say there's no IT bandwidth. Should I just ask for a short extension?
Not first. Ask what their average days-to-close from FNOL has done since the migration project started — it has usually gone the wrong way, which is an argument for keeping the thing that triages severity, not dropping it. Then narrow your integration surface out loud: read-only against the claim file, flat file or message queue, no schema change, nothing touching check issuance or statutory extracts, and offer to have their VP of IT confirm it himself. If it genuinely needs IT work, say so and negotiate a right-sized, claims-owned scope for now with a roadmap conversation the quarter after go-live. A short extension is a fallback, not an opening.
- Should I discount to save the account after a bad year?
No — not as an apology. A price cut reframes twelve months of service failure as a pricing dispute, tells them list price was always soft, and resets your floor permanently. If you owe them something, service credits for the specific affected months are the honest instrument: dated, finite, and they don't touch your renewal rate. Better still, right-size seats to actual usage. If they bought 160 licences and 50 people log in, renewing 70 with step-up pricing gets their expense line down, kills the shelfware objection for good, and signals you'd rather be right than big.
- What do I do if the person on the call isn't the one who can approve the renewal?
Assume they might not be, even if they signed last time — claims orgs reorganise, and budget authority moves. Ask directly what happens between now and signature: who signs, whether it routes through procurement, whether legal or compliance need to refresh a model governance review. Then get the Chief Claims Officer and your actual power user — usually the SIU Director or a subro manager — into the plan review together, with your exec sponsor on the line. Most stalled carrier renewals die in procurement in week five, and procurement has usually had the competitor's quote since months before you found out about it.
- What exactly should the 90-day plan contain for a claims platform?
Four fields per line: what, who by name on both sides, by when, and what 'done' looks like. Typically: desk-by-desk training for the adjusters who were never onboarded with a weekly-active target at day 60; a threshold reset where the SIU Director — not you — sets the referral volume his unit can actually work; a named support engineer with a written P1 response time and a cat-season volume trigger; 30/60/90 reviews with a usage report and a date on the first one; and, written down, the one or two things you cannot deliver. Then ask what they'll own. A plan where only the vendor has obligations is a wish list, and that's exactly how adoption failed the first time.