Quick Introductions: Who's in the Room?

• Who are the people we should consider part of the core decision team for this claims initiative? Options: VP of Claims, Head of Auto Physical Damage, Claims Transformation Lead, CIO/CTO, Head of IT Integrations, Head of Compliance/Legal, Procurement, Other (please name)
• Which single role will have ultimate sign-off authority for a pilot and contract? Options: VP of Claims, Head of Transformation, CIO/CTO, Chief Procurement Officer, CEO/Board, Other (please specify)
• Tell us who will be the operational day-to-day point of contact for discovery and what their title is.
• Are there informal influencers (committees, regional leads, union reps, audit teams) who typically shape claims decisions? If so, who are they?
• How do you prefer we coordinate communications with the team (weekly steering, biweekly working sessions, Slack, email)? Options: Weekly steering meeting, Biweekly working session, Ad-hoc as needed, Dedicated Slack/MS Teams channel, Email updates only, Other

Who Really Decides — and How Fast Can They Move?

• If everything aligned politically and technically, how quickly could your organization approve a pilot start date? Options: Immediately (under 2 weeks), Within 1 month, 1–3 months, 3–6 months, Longer than 6 months
• Describe the approval path from pilot approval to production—who signs, who advises, and which committees meet along the way?
• Which calendar constraints are non-negotiable for approvals (budget cycle, regulatory filings, board windows, fiscal year cutoffs)? Options: Budget cycle, Quarterly board reviews, Regulatory filing windows, Vendor procurement windows, No fixed constraints, Other
• When prior projects ran late, what were the top two root causes (e.g., data access, security review, executive reprioritization)? Options: Data access delays, Security/compliance review, Integration complexity, Stakeholder reprioritization, Procurement/legal negotiations, Vendor readiness, Other
• What single action from a vendor or partner has historically accelerated internal approvals the most?

What Would Move the Needle for Leadership?

• Which single metric—if improved—would change the conversation from ‘maybe’ to ‘go’ at the executive level? Options: Cycle time reduction (%), Loss Adjustment Expense (LAE) reduction ($), Estimate accuracy vs. adjuster (%), Supplement rate reduction (%), Straight-through processing rate (%), Customer satisfaction (CSAT/NPS)
• Please share current baseline values for the top 3 metrics you care about (e.g., cycle time = X days; supplement rate = Y%).
• Which of these metrics require auditable trails or regulatory reporting in your jurisdiction? Options: Estimate accuracy, Total loss valuation, Repair cycle time, Payment timing, Customer communications, None of the above, Other
• How confident are you in the accuracy and consistency of your current metric measurements? Options: Very confident, Somewhat confident, Doubtful, Not confident
• If a pilot improves your primary metric but causes a secondary KPI to worsen, which trade-offs would be acceptable and which are unacceptable?

Where the Pain Really Hides in Your Workflow

• Which stage of FNOL-to-settlement creates the most rework, cost, or delay for you today? Options: FNOL triage, Photo/virtual appraisal, Estimate generation, Supplement handling, Shop assignment/scheduling, Parts sourcing/delivery, Payment/finalization
• Walk us through a recent claim that required multiple supplements or escalations—what caused the divergence from the initial estimate?
• How frequently do initial estimates lead to at least one supplement, and what are the top 2 causes of those supplements? Options: <5%, 5–15%, 15–30%, >30%
• Describe the current touchpoints and handoffs with body shops—where do most approvals or disputes occur?
• How do your adjusters describe the emotional toll of rework and bad estimates (frustration, overtime, reduced morale)? Options: High frustration/overtime, Moderate stress, Occasional annoyance, Minimal impact

Assumptions We Should Put on the Table

• What’s one assumption about AI-based estimation or photo-first claims you’d be willing to challenge if given credible evidence otherwise?
• Do you assume AI cannot meet adjuster-level accuracy on complex repairs—how have you tested that belief so far? Options: Yes, not achievable, Possibly achievable with caveats, We believe it can match adjusters, Not tested
• Which integrations or data weaknesses have you labeled ‘too hard’ in previous projects? Options: Legacy claims system APIs, Body shop management systems (BMS), Parts pricing feeds, Telematics/vehicle data, Payment rails, None
• Are there regulatory or audit assumptions you’ve made that might be more flexible than you think (e.g., scope of photo evidence, virtual appraisal acceptance)?
• If we presented evidence that contradicted a key assumption, what internal barriers might still stop you from changing course? Options: Budget constraints, Political resistance, Procurement rules, Technical debt, Compliance concerns, Other

The Practical Limits: Tech, Data, and Security

• Which single system or data gap would be a showstopper for integration (i.e., without it a pilot cannot proceed)? Options: Core claims system access, Body shop management system (BMS) integration, Parts/pricing feed, Payment API, Identity/consumer data, Security/compliance certification (SOC2)
• Please list the claims, billing, shop, and parts systems (with versions) we must integrate with and any access constraints.
• Which data fields are available in real-time versus batched exports (photos, VIN, labor hours, estimates, parts lists)? Options: Real-time, Batched overnight, Manual export, Unavailable
• Do you have security or privacy certifications and controls we must meet before receiving test or production data (SOC2, ISO27001, local privacy regs)? Options: SOC2, ISO27001, Local privacy law compliance (e.g., CCPA, GDPR), Internal security review only, No formal certification required
• Have you tried similar integrations previously—what specific blockers or surprises came up during those projects?

What Would Success Feel Like to Your Team (Beyond the Numbers)?

• Beyond hard metrics, what cultural or operational changes would make leadership feel the project was worth it?
• If adjusters regained X minutes per claim or saw fewer escalations, how should that manifest in their daily work and morale?
• How should body shops signal acceptance—faster scheduling, fewer disputes, or explicit partnership metrics? Which do you trust most? Options: Faster scheduling, Fewer estimate disputes, Higher repair acceptance rate, Positive shop feedback, Other
• What non-quantitative signals (e.g., fewer complaint escalations, improved internal narratives) would you include in a success report?
• Which stakeholder groups will need visible, quick wins to stay engaged, and what would those wins look like for each group? Options: Claims leadership, Adjusters, IT/Integrations, Procurement/Legal, Body shops, Customers/Call centers

Risk, Governance, and the Red Lines

• What single operational, financial, or regulatory risk would cause you to halt the project immediately? Options: Regulatory non-compliance, Material overpayment risk, Data breach, Major shop pushback, Unacceptable customer complaints, Other
• What governance structure and cadence do you require during pilot (steering committee, weekly ops, monthly exec review)? Options: Weekly ops + monthly steering, Biweekly working group + executive monthly, Ad-hoc as issues arise, Formal steering committee only
• What contractual protections or KPIs (caps on overpayment, SLA penalties, indemnities) are essential before you sign?
• What rollback or pause criteria must be defined for the pilot to protect your operations and customers? Options: Error rate threshold, Customer complaint threshold, Financial loss threshold, Security incident, None required
• Who in legal/procurement/compliance should pre-review draft SOWs to avoid late surprises?

Pilot Scope: Small Enough to Win, Big Enough to Prove

• What minimum claim volume and composition (e.g., low-severity photo claims vs. complex desk-reviews) do you believe are required to demonstrate statistical validity? Options: <500 claims, 500–2,000 claims, 2,000–10,000 claims, >10,000 claims
• Which lines, geographies, or repair types should we include or explicitly exclude from the first pilot? Options: Personal auto (comp/PD), Commercial auto, Glass-only, Total loss, High-value fleet, Specific regions/states to exclude
• What acceptance criteria do you require at pilot close (accuracy thresholds, supplement rate cap, cycle time target, user satisfaction)? Options: Estimate accuracy threshold (%), Supplement rate target (%), Cycle time reduction target (days), Adjuster/shop satisfaction score, Other
• Who will be responsible for daily pilot operations, and who signs the formal pilot completion approval?
• What timeline (weeks/months) and sample size would you consider sufficient for a credible go/no-go decision? Options: 4–6 weeks, 6–12 weeks, 3–6 months, 6+ months

Next Steps and How We'll Keep Momentum

• If leadership said yes today, what is the earliest realistic start date for technical onboarding and data sharing? Options: Immediately (within 2 weeks), Within 1 month, 1–2 months, 2–3 months, 3+ months
• What specific internal approvals must we help you prepare (security sign-off, data-sharing agreement, budget request)? Options: Security sign-off, Data-sharing agreement, Budget approval, Procurement contract, Legal review
• What materials would make internal approvals easier (pilot SOW template, ROI model with your baselines, technical integration checklist)? Options: Pilot SOW template, Custom ROI model, Integration checklist, Legal/compliance FAQ, Executive one-pager
• Who should attend the next working session from your side, and what should they bring (data extracts, system access, decision authority)? Options: Claims operations lead, IT/integration lead, Security/compliance, Procurement/legal, Body shop operations contact
• Are there any unstated concerns or political dynamics we should anticipate so we can address them proactively?

Start: How Claims Usually Begin (Quick, real-world snapshot)

• Walk me through the very first hour after FNOL on a recent claim that went well—what happened, who touched it, and how did it move?
• Which channels do customers use to report FNOL most often? Options: Mobile app / photos, Web portal, Call center, Agent/Broker, Third-party shop intake, Other
• How many first-notice claims does your team typically handle on an average business day? Options: Under 100, 100–500, 500–1,000, 1,000–5,000, 5,000+
• How are photos and VIN data collected today (select all that apply)? Options: Customer-submitted photos via app, Adjuster-taken photos, Shop-provided photos, OEM/vehicle data feed, No systematic photo/VIN capture, Other
• Tell us about one recent FNOL that surprised you—what made it stand out?

Where Bottlenecks Hide (Which step quietly eats your time and margin?)

• Which single step in your FNOL→settlement flow costs you the most time or money right now, and why do you think it's being tolerated?
• What is your median end-to-end cycle time today for non-total-loss, repairable claims (FNOL → payment/repair start)? Options: <24 hours, 1–3 days, 4–7 days, 8–14 days, 15+ days
• Where do claims most often stall—triage, appraisal, parts, shop scheduling, rental, or payment? Options: Triage / assignment, Appraisal/estimate, Parts sourcing, Shop scheduling, Supplements/rescoping, Payment/settlement
• How often do you have a claims backlog or surge that requires overtime, temp adjusters, or vendor appraisers? Options: Daily, Weekly, Monthly, Quarterly, Rarely/Never
• Describe a recent bottleneck in enough detail that we could replicate the root cause—what systems, people, and data were involved?

Who's Doing the Heavy Lifting? (Are your adjusters set up to scale?)

• Are your adjusters primarily field, desk, or a mix—and which group is currently most over-capacity? Options: Mostly field, Mostly desk, Balanced mix, Rely on vendor/contract adjusters, Other
• On average, how many closed estimate-capable claims does an adjuster close per day? Options: <5, 5–10, 11–20, 21–35, 35+
• Which tasks consume the bulk of adjuster time today (pick top three)? Options: Photos review, Estimate writing, Phone calls & coordination, Supplements handling, Data entry into CMS, Shop liaising, Other
• How do adjusters feel about automated estimates—skeptical, cautiously optimistic, or eager? Any recent reactions or quotes? Options: Skeptical, Cautiously optimistic, Eager, Divided / mixed
• What training or QA processes support adjuster estimation accuracy today (mentoring, calibration sessions, spot audits)?

Systems and Integration Reality (If the tech layer falters, what breaks first?)

• If your core claims system became unavailable for 24 hours, which downstream capability would you lose first and why?
• Which systems currently integrate with claims workflows (select all that apply)? Options: Core Claims Management System (CMS), Body Shop Management System (BMS), Parts catalogs / vendor pricing, Rental / mobility platform, Payment processor, Telematics/EV data, Third-party estimate engines, None/Manual integrations, Other
• What is the nature of those integrations—real-time APIs, nightly batches, manual CSVs, or spreadsheet handoffs? Options: Real-time APIs, Near-real-time webhooks, Nightly batch files, Manual CSV imports/exports, Screen-scrape/manual entry, Other
• Do we have read/write access to claims records, images, and notes for a pilot? If not, what approvals or contracts stand in the way?
• What data fields or feeds are most fragile or incomplete today (parts list, labor times, VIN decode, prior loss history)? Options: Parts pricing, Labor times, VIN / vehicle spec, Prior claims history, Photo metadata, Shop repair timelines, None / data is robust, Other

Body Shops and Repair Partners: What's Their Truth?

• Do your repair partners see your processes as helpful, neutral, or obstructive—and what is the most common shop complaint? Options: Helpful, Neutral, Obstructive, Varies by shop network
• What percentage of repairs are handled by your preferred network versus arbitrary local shops? Options: 0–25%, 26–50%, 51–75%, 76–90%, 91–100%
• Which shop workflows are most error-prone: estimate acceptance, scheduling, parts ordering, supplement submission, or final invoicing? Options: Estimate acceptance, Scheduling, Parts ordering, Supplement submission, Final invoicing, All of the above
• Do your shops use a BMS—and if so, which vendor(s)? Are they connected to your CMS or are handoffs manual?
• Share a concrete example of a shop-related delay that impacted cycle time or cost—what happened and how was it resolved?

Supplements, Surprises and Escalations (Why do estimates get reopened?)

• When a supplement occurs, what do you blame most: initial estimate error, unseen damage at repair, or shop behavior? Options: Initial estimate error, Hidden damage discovered at repair, Shop upcoding / overbilling, Parts availability delays, Policyholder behavior / late disclosures, Other
• What proportion of repairable claims generate at least one supplement? Options: <5%, 5–10%, 11–20%, 21–35%, 36%+
• Average additional days caused by supplements (from repair start to completion)? Options: <1 day, 1–3 days, 4–7 days, 8–14 days, 15+ days
• Which types of damage or vehicle classes drive the highest supplement rates (e.g., EVs, heavy trucks, late-model cars, structural damage)? Options: EVs / high-voltage, Light commercial / fleet, Late-model collision vehicles, Structural / frame damage, Cosmetic-only, Other
• Describe a memorable supplement dispute—how did it start, who escalated it, and what was the outcome?

Data Quality: If the Input Is Flawed, the Output Lies

• What percentage of claim photo sets would you classify as high-quality (clear angles, good lighting, required views present)? Options: <30%, 30–50%, 51–70%, 71–90%, 91–100%
• Do you provide photo guidance or templates to customers or shops today? If so, what channels and adoption rates? Options: Yes—app guidance (high adoption), Yes—app guidance (low adoption), Yes—call center guidance, Yes—shop guidance, No formal guidance, Other
• Which data elements are consistently missing or inaccurate that would impede automated estimating (VIN, mileage, prior damage, options package)? Options: VIN, Mileage, Prior damage history, Vehicle options / trim, Repair history, None—data is reliable, Other
• Do you have labeled historical data (photos mapped to true repairs/invoice) available to share for model validation or training? Options: Yes—extensive, Yes—limited samples, No, but can collect, No and cannot share
• How do you currently verify estimate accuracy—post-repair audits, desk reviews, shop invoices comparison, or another method? Options: Post-repair audits, Desk reviews, Invoice reconciliation, Third-party audits, No systematic verification, Other

Regulatory, Audit, and Risk Appetite (Could a regulator pause this?)

• Are there regulatory or state-specific restrictions that constrain virtual/photo-only appraisals in your book of business? Options: Yes—multiple states, Yes—a few states, No known restrictions, Unsure / need legal to confirm
• What audit trail and reporting requirements would a new estimation tool need to satisfy for your compliance and audit teams?
• What is your acceptable overpayment tolerance on pilot claims (e.g., $ per claim or % of claims)? Options: Zero tolerance, <$50 per claim, $50–$200, $200–$500, We evaluate on % basis
• Have you had regulatory inquiries or litigation related to photo-based claims in the last 3 years? If yes, summarize. Options: Yes—detailed history, Yes—minor inquiries, No, Prefer not to say
• How important is a fully auditable decision log (images, model scores, adjuster overrides) for rollout approval? Options: Critical, Important, Nice-to-have, Not necessary

What Success Actually Looks Like (Numbers, leadership expectations, and blind spots)

• If we returned to your executive team in 6 months, which three KPIs would convince them this project is successful? Options: Cycle time reduction, LAE / reserve reduction, Supplement rate reduction, Customer NPS / satisfaction, Shop acceptance rate, Straight-through processing rate, Other
• What target reduction in cycle time or LAE would be considered a win for a pilot (give percentage or absolute numbers)?
• What accuracy threshold for AI estimates would you require to let them be used for straight-through payment on low-complexity claims? Options: >95% parity with human, 90–95% parity, 85–90% parity, Lower if accompanied by controls
• Who must sign-off to move from pilot to production (roles, not names)?
• What hidden political or operational risks could sink a pilot even if the numbers look good?

A Tiny, Low-Risk First Step (What would make leadership comfortable to try this?)

• What would a low-friction pilot look like to you—volume, vehicle classes included, geographies, and timeline?
• Which sample size feels reasonable for a pilot to be statistically meaningful but operationally contained? Options: 50–100 claims, 100–500 claims, 500–1,000 claims, 1,000+ claims
• What data and system access would you be comfortable granting for a 90-day pilot (images, writeback rights, only read access)? Options: Read-only to claims & images, Read + write to estimate field (controlled), Read + write + BMS push, Limited sample via CSV exports, Other
• Who on your side would we need in the room weekly during a pilot (roles and their primary decision/approvals)?
• What would be an acceptable go / no-go decision gate at pilot end (single KPI or composite), and what thresholds matter?

Quick Orientation: Who's In the Room?

• Who will be our primary day-to-day deployment contacts from your team (name, role, email)?
• Which executive or committee has final sign‑off authority for pilot go/no-go and production roll‑out? Options: VP of Claims, Head of Auto PD/Physical Damage, CIO, Claims Transformation Lead, Chief Risk Officer, Other
• What is your target window for pilot start and for first production cutover? Options: Within 30 days, 30–60 days, 61–90 days, 3–6 months, 6+ months
• What business outcomes will justify moving from pilot to production for you (select all that apply)? Options: Estimation accuracy delta vs adjuster (%), Reduction in average cycle time, Lower supplement rate, Lower LAE per claim, Shop/claimant NPS or acceptance, Reduced touchpoints per claim, Other
• Are there non‑negotiable constraints or blackout dates we should plan around (e.g., blackout for seasonality, regulatory reporting windows)? Please list.

If Deployment Fails, What's the Real Cost?

• What single outcome would make you escalate the program immediately if it occurred during pilot? Options: Significant overpayment trend, Major integration outage, Regulatory complaint, Shop refusal to participate, Security breach, Other
• Tell us about one past claims technology deployment that didn’t meet expectations—what went wrong and why?
• How often have integrations with vendors or shops caused schedule slippage in previous projects? Options: Almost always, Often, Sometimes, Rarely, Never
• What is the maximum tolerable outage or data loss window during deployment before leadership requires a stop? (hours/days) Options: <1 hour, 1–4 hours, 4–24 hours, 1–3 days, 3+ days
• If the pilot delivered expected accuracy but shops pushed back, which consequence would hurt you most (operational cost, claimant satisfaction, reputational risk, regulatory exposure)? Options: Operational cost, Claimant satisfaction, Reputational risk, Regulatory exposure, Other

Where Does Your Data Live—and Is It Ready?

• How confident are you that your claims records, images, and VIN/parts data are production‑quality for automated estimation without major cleansing? Options: Very confident, Mostly confident with some gaps, Mixed quality across lines/regions, Significant cleansing required, Unknown
• Which data feeds will we need to ingest for pilot (select all that apply)? Options: Claims Management System (CMS) records, Policy data, Policyholder images/photos, Shop/BMS updates, Parts/catalog data, Vehicle valuation feeds, Telematics/connected car data, Other
• What formats and transport methods do those feeds use today (e.g., REST JSON, SOAP XML, SFTP CSV, flat files, Kafka)? Options: REST/JSON, SOAP/XML, SFTP/CSV, Secure FTP with certificate, Message bus (Kafka/RabbitMQ), Direct DB replica, Other
• What is the expected daily volume of claims and images in the pilot scope, and how many sample claims can you provide for pre‑integration testing?
• Are there common data quality issues we should expect (missing VIN, low‑res images, incorrect loss codes)? Please provide examples and frequency.
• Do you have a canonical mapping for parts and labor codes that must be preserved, or will mapping be required? Options: Canonical mapping exists and is stable, Mapping required and maintained internally, Mapping required and will be co‑developed, Unknown

Who Owns the Systems We'll Need to Touch?

• Which internal teams or vendor partners are absolute gatekeepers for integrations and credentials (name teams/contacts)?
• Which claims management systems and versions does your organization use in the pilot population? Options: Guidewire ClaimCenter, Duck Creek, Crawford, In‑house/Custom, Other
• Which body shop management systems (BMS) and shop portals will we need to integrate with for the pilot?
• What integration authentication methods are acceptable for your security team (select all that apply)? Options: OAuth2, Mutual TLS certificates, Static API keys, SFTP with keypair, VPN/IP allowlist, Other
• Who will own change control and approvals for schema or endpoint changes during the pilot? Options: Carrier IT/Integration team, Vendor integration owner, Shared governance, Other
• Are there third‑party middleware or ESBs that requests must route through (and do they add transformation rules we need to know)? Options: Yes — details will be provided, No

Can You Imagine a Smooth Test Run? Tell Me About It.

• What would a credible pilot test plan catch before go‑live that current plans might miss?
• Which test environments are available and how closely do they mirror production? Options: Full staging mirror of prod, Partial staging (subset of data), Sandbox with synthetic data, No pre-prod environment available, Other
• How many real claims (and images) should we run through the pilot to feel statistically comfortable with accuracy and supplement rate signals? Options: <100 claims, 100–500 claims, 500–1,000 claims, 1,000–5,000 claims, 5,000+
• Which acceptance criteria will you apply at pilot closure (choose up to three and specify thresholds in the next question)? Options: Estimation accuracy vs adjuster (%), Supplement rate, Cycle time reduction, Shop acceptance rate, Claimant satisfaction/NPS, Integration uptime
• Please specify the numeric thresholds for the acceptance criteria you selected (e.g., accuracy ≥ 90%, supplement rate ≤ 5%).
• Who from claims operations and shops will participate in UAT and sign off on functional acceptance?

What Keeps Your Security & Compliance Team Up at Night?

• Which compliance or regulatory regimes must we satisfy for data handling and AI estimation in your jurisdiction(s)? Options: State insurance regulations, GDPR, CCPA/CPRA, HIPAA, SOC2/ISO27001 expectations, Other
• Do you require evidence of third‑party certifications (SOC2 Type II, ISO27001) or penetration test results prior to production access? Options: Yes — SOC2 required, Yes — ISO27001 required, Pen test required, We will accept vendor attestation, No formal requirement
• Are there data residency or localization constraints for images, PII, or valuation data? Options: Yes — must remain in specific country/region, No residency constraint, Some feeds constrained; details to follow
• What are your encryption requirements for data at rest and in transit? Options: TLS 1.2+ in transit, AES‑256 at rest, TLS in transit, AES‑128 at rest, Only in transit required, Other / custom requirement
• What is your breach notification SLA and escalation path we must accept contractually? Options: <24 hours, 24–48 hours, 48–72 hours, As soon as reasonably practicable
• Are there mandatory privacy controls (PII redaction, consent logging, data retention windows) we must implement for the pilot?

Plan B and Post‑Launch: Who's Ready to Act?

• If week one of production shows an unacceptably high supplement rate, what immediate action would you expect us to take? Options: Pause automated settlement, Throttle to manual review, Rollback to previous process, Adjust model thresholds, Other
• What monitoring and reporting would you like to see daily during pilot (select all that matter)? Options: Estimate accuracy by vehicle type, Supplement rate by shop, Time‑to‑estimate, Integration error rates, Shop acceptance/decline rate, Claimant satisfaction signals
• What SLA do you require from us for response and resolution during the pilot? Options: 4 hours, 8 hours, 24 hours, 48 hours, Other
• Who will run daily governance—triage calls, backlog prioritization, and incident reviews—during pilot operations? Options: Carrier governance owner, Vendor deployment manager, Joint governance, Other
• What training and enablement do your adjusters and shops need before go‑live (e.g., role‑based sessions, quick reference guides, co‑pilot support)? Options: Live training sessions, On‑demand videos, Quick reference guides/Cheat sheets, Onsite shadowing, Ongoing office hours
• Describe your rollback/runbook expectations—who initiates rollback, what triggers it, and how do we validate recovery?

Getting Comfortable — a quick share

• Which of the following best describes your primary role in the claims organization? Options: VP of Auto Claims, Head of Auto Physical Damage, CIO/CTO, Head of Claims Transformation/Innovation, Compliance/Legal, Other
• Tell us briefly about the parts of your book you oversee (personal vs commercial, vehicle types, annual claim volumes, regions).
• Approximately how many vehicle damage claims does your organization handle per year? Options: <10k, 10k–50k, 50k–200k, 200k–500k, >500k
• Who normally attends vendor or pilot decision meetings at your company (select all that apply)? Options: VP Claims, Head PD/Auto, CIO/IT, Procurement, Compliance/Legal, Shop Network Lead, Finance, Other

Are We Missing Millions in Plain Sight?

• When was the last time your team tried to quantify leakage from estimates, supplements, cycle time or shop rework—and what surprised you most?
• Which of these do you believe creates the largest hidden cost today? Options: Initial overpayment due to estimation inaccuracy, High supplement frequency, Excessive rental days / cycle time, Rework and shop disagreements, Incorrect total loss valuations, Other
• Can you describe a recent claim where an estimate went materially wrong—what happened and what downstream impact did it cause?
• If you had to pick one cost you can’t currently measure but suspect is large, what would you choose and why?
• How does missing that visibility make you feel when you brief your leadership or Board? Options: Anxious about financial exposure, Frustrated at lack of progress, Confident we can solve it, Neutral / It’s expected, Other

Show Me the Hard Numbers

• If someone asked you to prove your estimation process is efficient, which three metrics would you hand them—and why might those still be misleading?
• What is your current baseline for average estimation accuracy (estimate vs final paid) today? Options: >98%, 95–98%, 90–94%, 80–89%, <80%, We don’t measure this
• What is your typical supplement rate (supplements per 100 repairable claims)? Options: <5, 5–10, 11–20, 21–40, >40, Unknown
• What is your average cycle time from FNOL to settlement (days)? Options: <3 days, 3–7 days, 8–14 days, 15–30 days, >30 days, Varies widely
• How confident are you in the integrity and completeness of the data that produces these metrics? Options: Very confident, Somewhat confident, Doubting key gaps, Data is poor/unreliable
• How often do you reconcile estimates against final paid amounts and root-cause the differences? Options: Continuous / monthly, Quarterly, Annually, Rarely, Never

Who’s Driving — and Who’s Nervous?

• Which stakeholder in your org would be most likely to block photo-based straight-through processing—and what’s the core fear behind that stance?
• Which stakeholders should be considered essential decision-makers for a pilot (pick up to three)? Options: VP Claims, Head PD/Auto, CIO/IT, Compliance/Legal, Finance/CFO, Shop Network Lead, Customer Experience/CS
• How does compliance or internal audit typically evaluate new claims automation (biggest red lines)?
• How do your partner body shops typically react to new digital workflows or remote estimates? Options: Quick adopters, Cautious but cooperative, Resistant without incentives, Split by shop size/chain, Unknown
• Who would own day-to-day pilot success (name/role) and who controls the budget for scaling?

When Tech Breaks Trust

• Tell me about the last time a technology pilot eroded trust—what triggered the loss of confidence and how quickly did it spread?
• Which potential failure modes keep you up at night with AI estimation? Options: Systematic overpayment, Missed safety-critical damage, Poor integration with claims system, Data security/privacy exposure, Shop/customer dissatisfaction, Regulatory non-compliance
• How do you currently resolve disputes between shop repair estimates and carrier estimates? Options: Adjuster adjudication, Arbitration with shop, Re-inspection, Accept supplement, Other
• What acceptance controls (audit sampling, human-in-loop thresholds, financial recourse) would you require before allowing STH on a segment of claims?
• If a high-profile claimant complaint arose during a pilot, who would be pulled into the response and what outcomes would be unacceptable?

What Would 10% Better Actually Feel Like?

• If estimation accuracy improved by 10% and cycle time fell by 20%, what would be different in your daily operations, team morale, and P&L?
• Which KPIs would you celebrate first as proof the solution is working? Options: Average cycle days, LAE per claim, Supplement rate, Customer satisfaction (CSAT/NPS), Straight-through processing rate, Shop turnaround time
• What are realistic numeric targets you’d accept in a pilot for: estimation accuracy improvement, supplement reduction, and cycle time improvement?
• What hard constraints (regulatory limits, SLAs, audit windows) would prevent you from acting on those gains immediately?
• If the pilot met targets, which rollout path would you prefer? Options: Immediate enterprise-scale, Phased by region/product, Pilot-extension for more evidence, Limited feature rollout, Undecided

Trial Without Theater — Designing a Pilot That Actually Proves Value

• What would a pilot need to look like to survive executive scrutiny and a regulator audit—be specific about controls and artifacts?
• Which claim segments would make the most defensible pilot (select all that apply)? Options: Simple repairable FNOL with photos, Glass-only claims, Commercial fleet small dents, High-frequency low-severity, Total loss candidates, Complex/third-party claims
• What minimum sample size and duration would your leadership accept as credible evidence? Options: <500 claims / 1 month, 500–2,000 claims / 2–3 months, 2k–10k claims / 3–6 months, >10k claims / >6 months, Undecided
• Which system integrations are non-negotiable for the pilot to be believable? Options: Core claims system (CDR/CMS), Body Shop Management System (BMS), Parts catalog / labor guide, Payment system, Rental booking, Telematics / OEM data
• What concrete acceptance criteria (numeric thresholds, audit rules, escalation paths) would end the pilot as a success?
• Who must sign the pilot off (name/role) and what governance cadence would you require during the pilot? Options: Weekly, Bi-weekly, Monthly, Ad-hoc as needed

People, Training, and Shop Adoption

• What would have to be true for your adjusters and partner shops to welcome — rather than resist — this change?
• Which training formats produce the fastest and most durable behavior change for your adjusters? Options: Live instructor-led workshops, Hands-on shadowing, Micro e-learning modules, Scenario-based assessments, Job aids/cheat-sheets
• How do you typically incentivize shops to adopt new workflows or systems? Options: Financial incentives, Preferred shop placement, Operational support/training, Service-level commitments, We don’t incentivize
• Which adoption signals from shops and claimants would prove behavior change (pick top 3)? Options: % shops using tool, Reduction in shop disputes, Claimant NPS improvement, Decrease in rework/supplements, Faster repair lead times
• What ramp timeline feels realistic for training and shop adoption to reach meaningful usage? Options: 2–4 weeks, 1–3 months, 3–6 months, 6–12 months, Longer/uncertain

Data & Security — the Quiet Gatekeepers

• If your data feeds were evaluated tomorrow for a pilot, what single gap would disqualify you?
• Which of these data feeds can you realistically provide within 30 days? Options: Policy/claim metadata, Photos (policyholder / adjuster) with timestamps, Repair orders/ROs, Parts & labor detail, Payment history, Telematics/OEM data, None of the above
• Are any of your feeds subject to special consent or regulatory restrictions we need to plan for? Options: Yes — requires explicit claimant consent, Yes — restricted by regulator, No special restrictions, Unsure
• Which security or privacy certifications do you require for vendors? Options: SOC2 Type II, ISO27001, PCI-DSS, GDPR-compliance, State-level insurance/data controls, None required/undecided
• What is your preferred approach to test environments and synthetic data during pilot validation? Options: Sandbox with anonymized production data, Fully synthetic dataset, Controlled redaction of live data, On-premise isolated environment, Other

Deciding Together — Next Steps & Hidden Risks

• What’s the single question your CFO or legal will ask after this conversation that we should answer first?
• What remaining objections—internal or external—could realistically derail a pilot in the next 30 days?
• Which artifacts would make the decision easy for you (pick all that apply)? Options: Tailored ROI model with carrier data, Detailed pilot plan and timeline, Security & compliance documentation, Integration technical spec, Customer references and case studies, Sample audit reports
• Who should be included in the next meeting to clear those objections (name/role) and what is their availability in the next 2 weeks?
• What would make you feel confident committing to a pilot within 30 days (concrete must-haves)?