A quick orientation: where we begin together

• To get us started, tell me briefly about your current telematics experience and which data sources you actively use (smartphone app, OBD, OEM/embedded, other). Options: Smartphone app, OBD-II device, Embedded/OEM feed, Patchwork of multiple sources, No production telematics yet, Other
• In a typical month, which internal teams touch telematics data or its outputs in your organization? Options: Actuarial, Underwriting, Product, IT/Platform, Legal/Compliance, Distribution/Marketing, Customer Service, Other
• Who is the ultimate decision owner for pricing model changes tied to telematics in your organization? Options: Chief Actuary, Head of Personal Auto/Product, Chief Underwriting Officer, CIO/CTO, Chief Data Officer, Other
• Give a recent example of a pricing or discount change you made because of a new data signal; what happened to enrollments, opt-in, or loss ratio?
• Estimate how many active retail personal auto policies you would expect to include telematics within 12 months if adoption went according to plan. Options: <1,000, 1,000–10,000, 10,000–50,000, 50,000–200,000, >200,000, Unsure

Where the current model trips up

• If your existing rating factors truly separated risk, why do your highest-value drivers still leave for competitors or switch carriers?
• Which traditional rating variables cause the most mispricing in your book right now, and why do they fail for specific driver segments? Options: Age/experience, Territory, Credit-related factors, Vehicle class, Mileage proxies, Other
• How often do predictive lifts from new variables degrade after deployment, and how do you detect that degradation operationally? Options: Less than 6 months, 6–12 months, 12–24 months, Rarely detected, We do not monitor routinely
• Describe a recent pilot or model change where results surprised you, what went wrong or right, and what you learned.
• What single operational failure or data gap would make you halt a telematics pilot immediately? Options: No trip-level linkage to claims, Insufficient opt-in rates, Data quality/accuracy concerns, Regulatory objection, Security/privacy breach, Other

Regulation and privacy, not the background noise

• Are there any regulators, state examiners, or precedent filings that have already signaled concern about telematics-based rating for your company? Options: Yes, active concerns, Yes, informal guidance, No, but monitoring, No
• Who in your legal or compliance function would need to sign off, and what specific artifacts do they typically require? Options: General Counsel, Head of Compliance, Privacy Officer, Regulatory Affairs, Actuarial sign-off, Other
• Tell me about your policyholder consent and opt-in model, how you measure opt-in rates today, and what percentage you view as minimally viable.
• On a scale from 1 to 5, how confident are you that your current privacy notices, data retention, and opt-in language would satisfy the most conservative state regulators? Options: 1, 2, 3, 4, 5
• If a regulator required a time-boxed, auditable pilot with consumer-facing disclosures, would that requirement stop your program, slow it, or make you more cautious? Options: Stop it, Slow it substantially, Proceed with adjustments, Not an issue

The alternatives you're weighing

• Name the vendors, incumbent partners, or internal projects you currently view as viable alternatives for telematics-based pricing.
• Which features or outcomes from those alternatives matter most when you compare them to bringing in a new partner? Options: Cost, Speed to production, Data coverage, Model accuracy, Regulatory defensibility, Integration effort, Other
• What measurable outcomes would your incumbent or current approach need to produce for you to stay with it instead of switching?
• Does anyone internally argue for building this capability in-house rather than buying it, and if so, who is the sponsor? Options: Yes, sponsored by IT, Yes, sponsored by Actuarial, Yes, sponsored by Product, No internal build proposal, Unsure
• How quickly would an improvement from a vendor need to appear for you to justify switching vendors within three months? Options: Immediate (within 1 month), Short term (1–3 months), Medium (3–6 months), Not a fast switch

Integration, data, and the people who make it work

• Assuming your engineering team remains at current capacity, who would own API integrations and how many full-time equivalent engineers could they commit to a pilot?
• List the core systems that must be integrated for pricing and policy administration (policy admin, rating engine, billing, CRM, telematics back-end).
• Do you currently expose APIs or secure SFTP endpoints for trip-level data and are they documented and vendor-accessible? Options: Yes, documented APIs, Yes, SFTP only, Limited internal APIs, No external endpoints available, Unsure
• On a percentage basis, what proportion of your historical policies have usable telematics-linked claims history for retrospective validation? Options: 0–5%, 6–20%, 21–50%, 51–80%, 81–100%, Unknown
• Identify the single person authorized to pause or greenlight access to production datasets, and can they approve within your target timeline?

What winning looks like for your team

• Imagine your loss ratio improved because of telematics, what would have to be true in the first 12 months for you to call the program a success?
• Rate which outcomes matter most to your leadership: claim frequency reduction, claim severity reduction, retention lift, new-business attraction, or regulatory defensibility. Options: Claim frequency reduction, Claim severity reduction, Retention lift, New-business attraction, Regulatory defensibility
• Tell me which KPI thresholds would force a redesign of the model or discount structure (provide examples and numeric targets if you have them).
• List the approvals that would still be needed after a successful pilot to move to production (actuarial, filing, underwriting, IT, legal). Options: Actuarial, Underwriting, IT/Platform, Legal/Compliance, Regulatory filing, Executive sponsor
• Name the one metric that, if achieved by the pilot, would make you prepared to sign a mutual commitment within 30 days.

Timeline and decision triggers

• Point to the internal deadline, external event, or competitor move that would force you to accelerate a rollout.
• When do you need first-rateable telematics data into pricing to affect the next set of rate filings or product launches? Options: Within 1 month, 1–3 months, 3–6 months, 6–12 months, No firm timeline
• Would committing to a three-month pilot be politically acceptable to your board and to your regulators? Options: Yes, acceptable, Acceptable with conditions, Not acceptable, Unsure
• Could procurement and legal deliver a signed contract within 45 days if the pilot demonstrated the expected lift and all artifacts were provided? Options: Yes, Yes with minor redlines, No, would take longer, Unsure
• Estimate the budget range you could allocate for a pilot and initial integration work. Options: <$50k, $50k–$150k, $150k–$500k, $500k–$1M, >$1M, Undetermined

Hard stops and deal killers

• Identify any regulatory or organizational redlines that would stop this project instantly.
• Are there contract terms your legal team will not accept under any circumstances (data resale, cross-border data flows, indemnity levels, liability caps)? Options: Data resale, Cross-border data flows, Unlimited indemnity, No exclusions for regulatory risk, None absolute, Other
• Describe any legacy vendor agreements, OEM restrictions, or third-party terms that would block access to embedded vehicle signals or prevent integration.
• Would inability to access trip-level data for 70% of your renewal population be fatal to the program? Options: Yes, fatal, Manageable with adjustments, No, still viable, Depends on which segments are missing
• Select any of the following present today: ongoing regulator inquiries, pending privacy litigation, data residency constraints, none of the above, other. Options: Ongoing regulator inquiries, Pending privacy litigation, Data residency constraints, None of the above, Other

Acceptance criteria and next steps

• Provide the concrete acceptance criteria you would require to move from pilot to production, including numeric targets, sample size, and operational SLAs.
• Please list the stakeholders who must approve the acceptance criteria and the approval order.
• Select which artifacts you require for regulatory filing support: model documentation, validation reports, privacy impact assessment, opt-in audit logs, other. Options: Model documentation, Validation reports, Privacy impact assessment, Opt-in audit logs, Consumer communications, Other
• By when do you expect a production rollout decision after a successful pilot? Options: Within 30 days, 30–60 days, 60–90 days, More than 90 days, Depends on filing cycles
• When faced with acceptance criteria that fail on one metric but exceed on others, what is your default decision rule? Options: Fail the pilot, Iterate model and re-test, Accept with mitigations, Escalate to executive committee, Case-by-case
• Please identify the person enabled to sign a mutual commitment once criteria are met, and indicate whether remote signing is acceptable. Options: Yes, remote signing acceptable, No, requires in-person, Depends on contract value, Unsure
• Choose from the following blockers that would require executive escalation: legal redlines, budget shortfall, data access issues, integration complexity, regulatory objections, none. Options: Legal redlines, Budget shortfall, Data access issues, Integration complexity, Regulatory objections, None

• Solution Experience Session

• Confirm the current state and its cost
• You confirm the stated current state and acknowledge the documented cost to underwriting, actuarial, and product timelines.
• Provide a de-identified sample portfolio or loss pick to run the score-to-loss translation before the pilot scoping session.
• Map telematics signals to your loss drivers
• You validate that the score-to-price translation and sample cohort outcomes demonstrate a plausible improvement in loss ratio and retention for your use case.
• Deliver a scored sample cohort and score-to-price comparison using the buyer-provided sample, and summarize expected loss-ratio and retention deltas prior to the pilot scoping session.
• Prove score-to-price translation on a sample scenario
• Identify the states and regulatory owners whose approval will be required for a pilot and gather any existing guidance or prior filings.
• You agree on concrete pilot acceptance criteria and the remaining evidence needed for regulatory and legal sign-off.
• Confirm the internal stakeholders who must approve pilot acceptance criteria and schedule a follow-up validation meeting with them present.
• Show opt-in, privacy, and regulatory controls
• Agree pilot acceptance criteria and measurement plan
• Forced validation, confirm alignment

• Solution Experience Session
• Solution Experience Deck
• Solution Brief — Telematics Pricing & Retention

• meeting
• slides
• document

• current_state
• desired_state
• stakeholders
• gaps
• success_criteria
• decision_readiness
• desired_state
• decision_readiness
• current_state
• success_criteria
• gaps
• stakeholders
• desired_state
• success_criteria
• stakeholders
• gaps
• current_state
• decision_readiness
• decision_readiness
• decision_readiness
• decision_readiness
• decision_readiness

• Deploy Smartphone Data Collection App
• Provision OBD-II Device Program
• Integrate OEM Connected-Vehicle Feed
• Implement Data Ingestion and Trip Processing
• Build Driver Behavior Scoring Model
• Calibrate Score to Insurer Loss Experience
• Integrate Score with Rating and Underwriting
• Configure Pay-Per-Mile Billing Integration
• Implement Consent and Privacy Controls
• Activate Policyholder Coaching and Rewards
• Produce Actuarial Rate Filing Documentation
• Run Ongoing Model Monitoring and Recalibration
• Migrate Historical Telematics and Policy Data

Deploy Smartphone Data Collection App

• Which mobile operating system versions must the smartphone SDK or app support for policyholder devices? Options: iOS 14+, iOS 15+, Android 10+, Android 11+, Other
• Do you require a standalone white-labeled mobile app or an SDK to embed in your existing insurer app? Options: White-labeled standalone app, Embed SDK into insurer app, Both, Undecided
• Provide your expected monthly active app user count to size ingestion and storage
• List required consent disclosures or state DOI language that must appear on the app consent screen

Provision OBD-II Device Program

• Specify your preferred OBD-II device distribution model Options: Central procurement and mail, Drop-ship to policyholder by insurer, Policyholder purchases device (bring-your-own), Device distribution partner, Undecided
• Who will own device inventory and RMA (return material authorization) operations?
• How many OBD-II units do you expect to provision in the first 12 months? Options: Less than 1,000, 1,000-10,000, 10,000-50,000, 50,000+
• When do you require tamper-detection, serial-number tracking, and firmware update capability in the devices? Options: At launch, Within first 3 months, Before pilot end, Not required

Integrate OEM Connected-Vehicle Feed

• Where will OEM connected-vehicle feeds need to be ingested (for example SFTP landing zone, API webhook endpoint)? Options: REST API endpoint, SFTP landing zone, Streaming webhook, Message queue, Other
• Is there a target VIN coverage percentage of your in-force fleet that must be reached by pilot or go-live? Options: <10%, 10-30%, 30-60%, 60%+
• Confirm the OEM event types required from the feed such as trip summaries, odometer snapshots, CAN brake events, or DTC (diagnostic trouble codes) Options: Trip summary, Odometer snapshot, CAN brake events, Diagnostic trouble codes, Other
• Describe any OEM authentication or legal attestation the integration endpoint requires (for example mutual TLS or signed data contract)

Implement Data Ingestion and Trip Processing

• Identify which ingestion protocols your platform will accept for telematics data Options: REST API, Streaming webhook, SFTP batch, Message queue
• Estimate the maximum acceptable end-to-end trip processing latency from ingestion to scored output (seconds/minutes/hours) Options: <30 seconds, <5 minutes, <1 hour, Next-day batch
• Select the required trip schema fields you need mapped into the platform (for example start_time,end_time,latitude,longitude,speed,odometer)
• Indicate minimum data quality thresholds required such as GPS accuracy in meters or percent of trips with complete start/end timestamps

Build Driver Behavior Scoring Model

• State the initial driver behavior features you want included in the scoring model (for example hard braking per 1,000 miles, night driving fraction, cell-phone distraction events)
• Would you like the model to prioritize interpretability for underwriting or predictive lift for claims? Options: Interpretability (explainable score), Predictive lift (black-box allowed), Balance of both, Undecided
• Will you require separate models or score segments for primary driver, secondary driver, and household aggregation? Options: Separate models, Single household-level score, Both, Undecided
• Name the measurable acceptance criteria for the driver behavior model to be considered production-ready against your labeled loss dataset

Calibrate Score to Insurer Loss Experience

• Which API or data dump will provide your historical policy and claims data for calibration (for example SFTP path or REST endpoint)?
• What operating environment do your actuarial tools run in for calibration tasks (for example internal analytics cluster, SAS environment)?
• What acceptance criteria will your actuarial team use to sign off calibration such as stability of relativity within X% or required p-values on holdout tests
• How will you handle non-overlapping coverage between telematics population and historical loss population (for example weighting, reweighting, or exclusion)?

Integrate Score with Rating and Underwriting

• Are there specific field names in your rating engine or policy administration system that must store the telematics score (for example TELEMATICS_SCORE, MPG, DISTANCE_BIN)?
• Outline whether you require raw score, percentile, or bucketed banding for underwriters to reference Options: Raw numeric score, Percentile rank, Bucketed tiers (e.g., A/B/C), Multiple representations
• Which measurable criteria will define done for score integration such as score written to policy record within X minutes, end-to-end test on N policies, and premium delta matching expected values
• Do you need role-based visibility controls for score display in underwriting tools (for example underwriter view, actuary view, customer-facing summary)? Options: Yes, No

Configure Pay-Per-Mile Billing Integration

• Provide the billing or general ledger system that will receive mileage and per-mile charges
• List your preferred billing cadence for pay-per-mile charges Options: Daily, Weekly, Monthly, Per-billing-event, Custom
• Specify pricing rules such as per-mile rate, rounding rules, minimum monthly charge, or mileage bands
• Who on your finance team will own reconciliation and disputes for telematics-billed transactions?

Implement Consent and Privacy Controls

• How many states' regulatory disclosures must the consent flow cover at launch Options: 1, 2-5, 6-15, 16+
• Is a privacy impact assessment or data protection impact assessment required for any state DOI submissions? Options: Yes, No
• Provide the required raw data retention period for trip-level telemetry in months Options: 3, 6, 12, 24, 36, Custom
• Are there specific PII fields you require to be pseudonymized or tokenized before storage (for example driver name, phone number, VIN)? Options: Driver name, Phone number, VIN, All of the above, None

Activate Policyholder Coaching and Rewards

• Select the engagement channels you plan to use for coaching and rewards Options: In-app push, Email, SMS, Physical mail, Other
• Would you like templated coaching messages and compliant consent copy created for state DOI review? Options: Yes, No
• Indicate if rewards fulfillment requires integration with third-party vendors and the expected format (CSV, API) Options: CSV export, API integration, Both, None
• Estimate a target policyholder opt-in rate for coaching/rewards that you consider successful in pilot Options: <10%, 10-25%, 25-50%, 50%+

Produce Actuarial Rate Filing Documentation

• Name the states where you intend to file telematics rating components in the first wave
• Would you like us to prepare an actuarial memorandum and example rate pages for each filing? Options: Actuarial memorandum only, Memorandum and rate pages, Full filing packet including forms, Undecided
• Estimate the lead time your regulatory team requires to review actuarial deliverables before submission (weeks) Options: 2, 4, 6, 8, Custom
• Identify any state-specific tests or comparators you need included such as decile relativity tables or experience period adjustments

Run Ongoing Model Monitoring and Recalibration

• Will you require automated monitoring alerts for model drift and data feed failures? Options: Yes, No
• Name the alert thresholds that should trigger investigation for model drift such as AUC drop percentage or Population Stability Index (PSI) threshold
• Who will be the primary point of contact on your team for incident response and model change approval?
• Select the regular deliverables you need from ongoing monitoring (dashboard, monthly summary, recalibration plan, audit log) Options: Dashboard, Monthly summary, Recalibration plan, Audit log, All

Migrate Historical Telematics and Policy Data

• Provide counts for historical trip records and policy rows that must be migrated
• List the source systems and their export formats for historical telematics and policy data (for example SFTP CSV, SQL dump, vendor portal)
• Specify the migration completeness threshold you require by percent and accepted error rate Options: 95%, 98%, 99%, Custom
• Are reconciliation reports and a sampling verification plan required post-migration? Options: Yes, No

• Master Services Agreement (MSA)
• Statement of Work (SOW)
• Subscription & Order Form
• Data Processing Agreement (DPA)
• Data Sharing and Use Agreement
• Regulatory Support Addendum
• Security and Compliance Addendum
• Service Level Agreement (SLA)
• Implementation Acceptance and Handover Agreement
• Change Order and Scope Modification Agreement
• Termination and Data Retention Schedule

• Go-live Health Check (weeks 1-4)
• First Measurement Review (weeks 4-10)
• Acceptance Gate Review (around day 90)
• Quarterly Outcomes Review

• Publish the quarterly metrics pack and updated runbook for the next review.
• Implement agreed score recalibration or data-quality fixes and report validation results within the agreed window.
• Schedule the Acceptance Gate Review once corrective actions have verification evidence.
• Restate acceptance criteria from Pilot Evaluation
• Each acceptance criterion recorded in Pilot Evaluation is evaluated and a pass/fail result is documented.
• A formal acceptance decision is captured and any required remediations have clear actions and deadlines.
• Legacy system decommissioning status is confirmed or a retention/read-only plan is documented to prevent dual-running.
• Publish the acceptance results table with pass/fail status and remediation plans.
• Initiate archive or migration of legacy system data per the agreed retention plan.
• Create the 30-day post-acceptance monitoring runbook and handover to operations.
• Trend review of primary outcome metrics
• Confirm whether core outcome metrics are meeting or moving toward the targets recorded in Pilot Evaluation.
• Agree a prioritized list of enhancements tied to expected metric impact and a plan to deliver the top items.
• Ensure any regulatory or privacy issues are identified and have clear mitigation steps.
• Produce a prioritized enhancement backlog with estimated metric impact and targeted delivery quarter.
• Close or re-scope high-severity operational defects that affect scoring or data integrity.
• Reconfirm success criteria and owners
• All deployment validation checks are either closed or have named owners and target dates.
• Early adoption signals (device onboardings, app installs, initial opt-ins) are documented and flagged if below expected ranges.
• Immediate remediation actions and verification steps are agreed and published within 48 hours.
• Publish the deployment validation checklist with pass/fail results and owners.
• Open tickets for each operational blocker with target resolution dates.
• Provide a short report of onboarding numbers and initial opt-in counts for the first measurement meeting.
• Present first measurement data vs Pilot Evaluation targets
• Clear view of where current opt-in rate and loss ratio sit relative to Pilot Evaluation targets.
• Root causes for any KPI gaps are identified and a prioritized remediation plan is agreed.
• Timeline and required closures to reach the acceptance gate are confirmed.
• Deliver a data pack that includes opt-in funnel metrics and loss-ratio calculations used in the analysis.
• Deployment and integration validation
• Operational issues and incident burn-down
• Present measured outcomes against each criterion
• Diagnose root causes for gaps
• Early adoption signals and usage patterns
• Enhancement prioritization by measured impact
• Document pass/fail per criterion and capture formal acceptance decision
• Agree corrective actions and timelines
• Regulatory and privacy compliance check
• Open operational blockers
• Confirm timeline to the acceptance gate
• Remediation plan for any failed criteria
• Agree immediate remediation actions
• Agree actions for the upcoming quarter
• Incumbent system wind-down confirmation
• Next steps to production monitoring