Start With Your One-Line: What’s the real urgency?

• In one sentence, why are you exploring a new eligibility and enrollment solution right now?
• How fast does leadership expect measurable improvement (e.g., weeks, months)? Options: Immediate (days–2 weeks), 30 days, 60–90 days, 3–6 months, 6–12 months, Unsure
• Who is ultimately accountable to the governor for meeting these goals? Options: Commissioner / Secretary, Deputy Director, State CIO, Program Director, Federal liaison, Other
• Which trigger best describes your situation right now? Options: Federal corrective action, Legacy COBOL failure/obsolescence, Disaster-driven surge, Policy change, Proactive modernization, Other
• What outcome would make you say, in three months, 'this decision was the right call'?

Are You Comfortable Leaving Payments to Chance?

• If you had to name the single biggest risk to uninterrupted benefit payments during transition, what would it be?
• How often in the past 12 months have timeliness or accuracy failures resulted in corrective action or formal findings? Options: Multiple times, Once, Close call but no finding, Never, Unsure
• Describe a recent incident where a system or process change nearly caused payment disruption—what happened and who felt the impact?
• What level of short-term risk is acceptable during transition, if it accelerates long-term compliance? (Select one.) Options: No increase in risk allowed, Small, tightly controlled risk, Moderate, with safeguards, Willing to tolerate larger risk for faster gains, Unsure
• Which stakeholders would elevate any sign of payment disruption immediately? (Select all that apply.) Options: Governor's office, Federal program office, State auditor, Legislature, Media/press, Client advocacy groups, Other
• How would missed payments or audit findings feel politically and operationally for your team?

What’s Hidden in the COBOL Black Box?

• Which legacy behaviors or business rules do you suspect exist only in COBOL code and nowhere in documentation?
• Which of these legacy artifacts currently drive decisions in your day-to-day operations? Options: Untested COBOL routines, Manual spreadsheets/workarounds, Ad-hoc supervisor notes, Paper case files, Legacy batch reports, Other
• How frequently do caseworkers apply manual overrides or exceptions because the system doesn’t match policy? Options: Daily, Weekly, Monthly, Rarely, Never, Unsure
• List two edge-case scenarios you worry a new rules engine might not replicate correctly.
• Have data conversions or feeds failed in past modernizations? If yes, what patterns repeated? Options: Yes—mapping/format issues, Yes—missing historical records, Yes—timing/latency failures, No major failures, Unsure
• How would you prioritize undocumented rules, data cleanup, and infrastructure stabilization (rank short answer)?

Who Really Pulls the Levers and Signs Off?

• Who will say 'go' or 'stop' for each of these areas: policy interpretation, technical acceptance, federal reporting, and public communications?
• Which groups must be included in approval gates for design or cutover? (Select all that apply.) Options: Policy/Program Office, IT/Architecture, Legal/Compliance, Federal Program Managers, County/Local partners, Vendor PMO, Other
• How do you currently document who is accountable for an eligibility decision when multiple parties are involved? Options: Case notes, Audit log in system, Manual sign-off forms, Not consistently tracked, Other
• Describe political or federal sensitivities we should know about up front (e.g., upcoming audits, elections, new federal guidance).
• Who in your team would be the most resistant to outsourcing parts of eligibility operations, and why?

If Washington Said 'Fixed', What Would That Actually Be?

• What specific accuracy and timeliness thresholds would satisfy federal reviewers in your view? Options: >= 98% accuracy; 100% timeliness, >= 98% accuracy; federal timeliness targets, 95–97% accuracy; federal timeliness targets, <95% accuracy acceptable short-term, Unsure—need guidance
• Which citizen experience signals matter most to you—what should improve for people calling or applying? Options: Faster resolution, Clearer explanations, Fewer recontacts, Reduced paperwork, Better multi-channel access, Other
• What audit artifacts or outputs must be produced end-to-end to satisfy the federal team? Options: Decision-level audit trail, Caseworker activity logs, Batch processing records, Federal report reconciliation, All of the above, Unsure
• If the governor asked you to point to one metric in 6 months to prove success, which metric would you choose and why?
• What trade-offs would you accept between speed, accuracy, and scope during transition? Options: Accuracy prioritized over speed, Speed prioritized over accuracy, Balance equally, Scope reduced to protect accuracy, Unsure

How Will Your People Live Through This Change?

• If we asked your front-line staff how they feel about modernization, what three words would they use?
• What is your current caseworker-to-workload ratio and how flexible is it during surge? Options: <1:50 (light caseload), 1:50–1:200, 1:200–1:500, >1:500 (high caseload), Unsure
• How do new policies, exceptions, or system quirks get taught to staff today? Options: Formal training sessions, On-the-job mentoring, Email/announcements, Shadowing/supervisor guidance, Not consistently taught
• What would make your staff trust a vendor’s contact center or BPO agents to handle sensitive eligibility conversations?
• How high is turnover among experienced eligibility staff, and what typically causes departures? Options: Low turnover, Moderate—compensation, High—burnout/caseload, High—unclear career path, Unsure
• What training cadence and competency checks would you require before we route live cases to external agents?

Can a New Rules Engine Recreate Your Worst Edge Cases?

• Which three legacy scenarios keep you awake because they’re hard to model in a rules engine?
• How important is transparency into decision logic (explainability) for auditors and supervisors? Options: Critical—must be fully explainable, Important—mostly explainable, Somewhat—exceptions okay, Not a priority
• What data sources must the new system ingest and reconcile on day one? Options: State payroll/earnings, Federal data hubs, Local records, Legacy batch files, Provider feeds, Other
• Have you previously tested rule parity between legacy and modern systems? What were the failure modes? Options: Yes—parity achieved, Yes—parity gaps in edge cases, No formal parity testing, Unsure
• What level of manual review do you expect to keep after cutover for high-risk cases? Options: None—fully automated, Small % flagged for review, Significant manual review for first year, Depends on case type

What Would a Safe Cutover Plan Have to Include?

• What concrete rollback triggers would immediately stop a cutover and revert to legacy processing?
• Which of the following must remain uninterrupted during cutover? Options: Federal reporting, Benefit payments, Critical public hotlines, Data transfers to other agencies, All of the above
• What surge scenarios should we rehearse in parallel runs (select all that apply)? Options: Natural disaster influx, Policy-driven spike, System outage elsewhere, Pandemic-like surge, Other
• Who will own daily cutover coordination across agencies and vendors (name/role)?
• What communications cadence and audience segmentation would you require during cutover? Options: Daily executive brief, Twice-daily ops check, Weekly stakeholder summary, Real-time incident channel, Other

Money, Contracts, and Guarantees — What Really Matters?

• Which contract levers are non-negotiable for you (select all that apply)? Options: Accuracy SLAs, Timeliness SLAs, Staffing FTE guarantees, Federal reporting compliance clause, Liquidated damages, Performance incentives
• Do you prefer penalties for missing SLAs, incentives for exceeding them, or a blend? Options: Penalties, Incentives, Blend, Case-by-case
• How important is vendor staffing continuity (same supervisors/SMEs) during the first 12 months? Options: Critical, Important, Nice to have, Not necessary
• Are there procurement or state law constraints we must design around (e.g., pay structure, background checks, residency)?
• What would constitute an unambiguous mutual commit from your side (e.g., funding authorization, data access, program approval)?

What Evidence Would Let You Say ‘Go’ Next Week?

• If we provided three artifacts next week, which would matter most: parity test run, staffed surge plan, or a draft SLA—rank your top pick? Options: Parity test evidence, Staffing & training roster, Draft contract & SLAs, Detailed cutover Gantt, Other
• Who must be at the table for a 60–90 minute decision workshop to move forward (roles/emails if possible)?
• What calendar windows are off-limits due to audits, elections, or fiscal deadlines in the next 6 months?
• What blockers would prevent quick engagement (e.g., pending RFP, budget hold, legal review)? Options: Pending RFP/procurement, No budget authority yet, Legal/compliance review, Political sensitivity, None
• How would you like us to demonstrate respect for your constraints while rapidly validating feasibility—what’s a reasonable first milestone? Options: 72-hour discovery findings, 2-week parity prototype, 30-day feasibility report, Other

Tell Us About the World You're Running Today

• In one or two sentences, how would you describe today’s eligibility platform and operations?
• Which public assistance programs does this system currently serve? (select all that apply) Options: SNAP, TANF, Medicaid, CHIP, Unemployment Insurance, Child Care Subsidies, Other
• How long has your COBOL-based system been the authoritative source for eligibility? Options: < 5 years, 5–10 years, 10–20 years, > 20 years
• Which operational outcomes are causing the most immediate pain right now? (select up to three) Options: Late payments, Incorrect denials, Backlogs/aging cases, Surge failures, Poor citizen satisfaction, Federal audit findings
• Who is the single point owner for payment timeliness and system uptime inside your agency? Options: Commissioner/Director, Deputy Director, Program Manager, State CIO, Shared ownership, Other
• When you picture this environment succeeding for the governor and citizens, what three signals would you expect to see?

What Legacy Rules Keep Breaking Assumptions?

• Which legacy rule behaviors are you quietly relying on that no one can fully explain?
• Can you name specific legacy rules or hard-coded exceptions that regularly produce unexpected outcomes?
• How often do these legacy behaviors create exceptions that require manual overrides? Options: Daily, Weekly, Monthly, Rarely
• Do you have up-to-date documentation (decision tables, flow diagrams, business rules) for those behaviors? Options: Complete and current, Partial/fragmented, Outdated, None
• When legacy logic diverges from written policy, which wins today—system behavior or policy guidance? Options: System behavior, Policy guidance, Depends on caseworker, Unclear
• If we had to reproduce every legacy exception exactly, which three would you rank highest priority and why?

Where Do Edge Cases Hide and How Often Do They Bite?

• Which rare scenarios take disproportionate staff time or trigger most escalations?
• Which of these edge-case categories apply in your caseload? (select all that apply) Options: Interstate cases, Mixed household types, Erroneous income records, Identity/documentation gaps, Retroactive eligibility, Emergency/disaster declarations, Other
• Approximately what percent of your active cases fall into those edge-case buckets? Options: <1%, 1–5%, 5–15%, 15–30%, >30%
• Tell a recent story of a single edge case that escalated to leadership or the governor’s office—what happened and why?
• How does the current system surface or flag these cases for specialized handling? Options: Automated flags/rules, Manual triage by staff, Separate queueing system, Not effectively surfaced
• What workarounds do staff use today to resolve these exceptions, and how fragile are those workarounds?

Who's Holding the Work — People, Shifts, and Burnout

• Are you unintentionally relying on heroic staff to keep payments on time?
• How many full-time equivalent caseworkers are processing eligibility actions today (give best estimate)? Options: <50, 50–199, 200–499, 500–999, 1000+
• What is your annual frontline turnover rate for eligibility staff? Options: <10%, 10–25%, 25–50%, >50%
• How do you currently onboard and certify staff on complex policy and system navigation? Options: Formal classroom training, On-the-job apprenticeship, Shadowing/mentoring, Ad-hoc/tribal knowledge, Vendor-run training
• When staffing gaps appear, what short-term fixes are used and how do those affect accuracy or timeliness?
• How do your caseworkers describe their day—overwhelmed, empowered, or something else? Options: Overwhelmed, Mostly managing, Efficient and empowered, Mixed

If the System Stutters, Who Pays the Price?

• How close has a system slowdown or outage come to triggering a missed federal timeliness standard in the last 24 months? Options: We missed standards, We narrowly avoided missing, No close calls, Not tracked
• List recent incidents that impacted payments or reporting in the last two years—what failed and for how long?
• What are your recovery time (RTO) and recovery point (RPO) targets for eligibility services? Options: RTO <1 hour / RPO <5 minutes, RTO 1–4 hours / RPO 5–60 minutes, RTO 4–24 hours / RPO 1–24 hours, No formal targets
• How do you handle peak-volume events today (e.g., disasters, mass layoffs)? Options: Auto-scale infrastructure, Surge staffing contracts, Manual triage and delays, We struggle to respond
• Which monitoring and alerting tools are in place to detect failures that affect payments? Options: Real-time dashboards, Threshold alerts, Manual health checks, No systematic monitoring
• Who is notified when a payment-impacting failure occurs and what is the escalation path? Options: Ops team only, Ops + program leadership, Executive/commissioner alerted, No clear escalation

Paper Trails, Data Gaps, and the Secrets They Hide

• Which data elements do you not trust, and what happens operationally when they’re incorrect?
• Rate the reliability of these core data domains in your system Options: Identity (SSN/DOB), Income/employment, Household composition, Address/contact information, Benefit history
• How are external interfaces and batch feeds (TANF, unemployment, federal sources) managed and reconciled? Options: Near real-time APIs, Daily batch jobs, Weekly/monthly transfers, Manual file exchanges
• Have you attempted a data conversion or migration before? What failed or surprised you about that effort?
• How frequently do you reconcile operational case outcomes to federal reporting extracts? Options: Daily, Weekly, Monthly, Quarterly, Ad-hoc/never
• What percentage of cases require manual data correction before being reportable? Options: <1%, 1–5%, 5–15%, >15%

What Would Put Federal Auditors and the Governor at Ease?

• If federal auditors came tomorrow, what single evidence item would you struggle to produce?
• Which federal reporting outputs give you the most anxiety (select all that apply)? Options: Timeliness reports, Accuracy/error rate reports, Financial reconciliations, Eligibility audits, Other
• What controls and attestations are currently in place to prove accuracy and timeliness? Options: Automated reconciliation, Manual attestations, Third-party audits, Limited controls
• How would missing a key federal metric affect your political leadership or corrective action status? Options: Immediate corrective action, Increased oversight, Public reporting but limited sanctions, Uncertain
• What evidence or capability would make you comfortable committing to a modernization timeline? Options: Parallel run results, Third-party validation, Staffing guarantees, Escrow/rollback plans, Other
• Which audit trails or artifacts are easiest for you to produce today, and which are the hardest?

What Would Break This Project in the First 12 Months?

• If you had to name one early failure that would be politically catastrophic, what would it be?
• Which of these risks feel most likely to occur during transition? (select top three) Options: Rules divergence from legacy, Insufficient surge staffing, Data conversion errors, Outages during cutover, Federal policy change, Stakeholder misalignment
• What explicit rollback triggers or acceptance criteria would you insist on for a full cutover? Options: Accuracy threshold, Timeliness threshold, QA pass rate, Parallel run concordance, Stakeholder sign-off
• Who are the internal and external stakeholders who must be engaged during month-by-month progress (list roles)? Options: Governor's office, Commissioner/Deputy, State CIO, Federal program office, County/field leadership, Union/HR, Other
• What three mitigations would make you say 'go' on an aggressive 12-month improvement mandate?
• Which early wins could we deliver in the first 90 days to reduce political and operational risk? Options: Stabilize payment engine, Resolve top edge-case classes, Deliver federal report concordance, Staff augmentation for surge

Quick Check — Who’s in the Room?

• Which role best describes your participation in this readiness planning? Options: State Commissioner/Director, Deputy Director / Program Lead, State CIO / IT Lead, Program Manager / Operations, Federal Liaison, Vendor/Integrator, Other
• Which stakeholders must explicitly approve go/no-go for deployment readiness? Options: Governor’s office, Federal program officer, State procurement/contract, IT security office, Program operations leadership, Quality assurance team, Other
• What firm dates or windows constrain your cutover planning (e.g., federal deadlines, fiscal milestones)? Options: Specific calendar date (e.g., regulatory deadline), End of corrective action period, Avoid month-end/benefit cycle, Rolling waves by region, No firm constraint, Other
• Who will serve as the day-to-day readiness lead we coordinate with during parallel operations?
• What is your single biggest fear about the deployment that you’d want us to address first?

If We Cutover Tomorrow, What Would Break?

• What hidden dependency — a system, report, or manual step — would most likely interrupt benefit payments if it failed at cutover?
• Which legacy COBOL behaviors or business rules do you currently rely on that are hard to find in documentation?
• Which edge-case categories cause the most variance in decisions today? Options: Interstate transfers/portability, Retroactive eligibility, Complex household/blended benefits, Nonstandard income sources, Manual overrides/waivers, Other
• How often do these edge cases occur and what percent of throughput do they represent? Options: Daily / >5% of cases, Weekly / 1–5%, Monthly / <1%, Rarely / anecdotal
• Can you share a recent example where a failure mode led to a missed federal timeliness or accuracy target? What happened?

How Confident Are You in Your Data?

• What proportion of your case and eligibility data do you expect will require cleansing or transformation before the cutover? Options: None — data is ready, Small fraction (under 10%), Moderate (10–30%), Large (30–60%), Very large (over 60%), Unsure
• Which of these data issues are present today that could derail conversions? Options: Missing/mismatched identifiers, Inconsistent household composition, Old lookup tables/rules, Sensitive PII protection concerns, Incomplete historical payments, Other
• What sample datasets can you provide for mapping and validation (scope, formats, frequency)? Options: Full anonymized extracts, Subset extracts by region, API access to live test data, No samples available yet, Other
• Who owns the authoritative source for each critical data domain (eligibility, payments, case notes, audit logs)?
• How would you prioritize data correctness vs. expediency if we had to choose during conversion (e.g., delay cutover to fix data or proceed and remediate post-cutover)? Options: Prioritize correctness even if delayed, Balance both equally, Prioritize meeting cutover timeline and remediate after

If Your SMEs Were Gone Tomorrow, Could You Survive?

• If your three most experienced policy SMEs were unavailable during launch, how would you expect complex adjudications to be handled? Options: We have documented escalation paths, We would rely on vendor SMEs, We would pause complex cases, Unsure
• Which knowledge-transfer assets currently exist and which are missing (e.g., runbooks, decision trees, recorded shadow sessions)? Options: Runbooks, Decision matrices, Recorded training sessions, Annotated legacy code explanations, None of the above, Other
• What proportion of your frontline caseworkers regularly handle edge cases versus scripted routinized work? Options: Most (over 60%), Some (20–60%), Few (under 20%), Unsure
• How do you currently measure caseworker competency and how often are performance assessments performed? Options: Monthly QA reviews, Quarterly assessments, Annual training checks, Ad-hoc / as needed, No formal measurement
• What retention or staffing risks should we plan for during training and the parallel period (e.g., seasonal turnover, collective bargaining restrictions)?

What Would a Surge Really Look Like — and Who Breaks First?

• At what multiplier of baseline volume (e.g., 2x, 3x) have you historically seen critical failures or missed timeliness? Options: 1.5x, 2x, 3x, 4x+, No historical surge
• Which components do you believe are most fragile under surge: intake, rules processing, payments, QA, or contact center? Options: Intake, Rules engine, Payments/reconciliation, Quality assurance, Contact center, Other
• What surge triggers are realistic for your program (natural disaster, policy change, eligibility expansion, corrective action deadline)? Options: Natural disaster, Policy change, Economic shock/unemployment spike, Federal corrective action, Other
• Do you have existing surge playbooks and have they been exercised in the last 24 months? Options: Yes — exercised recently, Yes — not exercised recently, No playbook, Partial playbook
• What maximum acceptable response times and SLA targets must hold during a 3x surge to avoid federal penalties? Options: Meet baseline SLAs at all times, Allow temporary SLA degradation with reporting, Unclear / need to define

Parallel Run — Will Two Systems Create Reconciliation Chaos?

• What reconciliation cadence would you prefer during parallel operations: daily, weekly, or event-driven? Options: Daily, Weekly, Bi-weekly, Event-driven / exceptions only, Other
• Which acceptance gates should determine when a case moves from legacy to new platform (e.g., accuracy threshold, audit signoff, user competency)? Options: Accuracy threshold, Audit acceptance, Caseworker certification, Federal signoff, Other
• What reconciliation methods do you currently use or prefer (automated diff tools, manual audits, sampling)? Options: Automated diffs, Manual reconciliation, Statistical sampling/audits, Hybrid approach, Other
• How many parallel run iterations do you expect before full cutover, and over what timeframe? Options: 1–2 iterations, 3–6 iterations, Continuous / rolling for 12 months, Unsure
• What team or role will own day-to-day reconciliation and exception resolution during parallel?

If We Ever Had to Roll Back, How Fast Must We Be?

• What is the maximum acceptable exposure window (hours/days) before federal timeliness or payment integrity consequences occur? Options: Under 4 hours, 4–24 hours, 1–3 days, More than 3 days, Unsure
• What concrete rollback triggers would you require (e.g., payment errors above X%, outage affecting Y% of users)?
• Who has final authority to approve an emergency rollback and what communication chain should be triggered? Options: Governor’s office, Commissioner/Director, Federal program officer, Joint state-vendor committee, Other
• How many rollback rehearsals would you expect before go-live (none, 1, multiple) and in what environments should they run? Options: None, One rehearsal, Multiple rehearsals, Continuous tabletop exercises
• What data synchronization or reprocessing steps worry you most during a rollback (payments, appeals, audit logs)? Options: Payments, Appeals/case notes, Audit logs, Benefit history, Other

What Metric Will Make the Governor Sleep at Night?

• Which single outcome would you use to declare deployment success publicly (e.g., 98% accuracy, 95% on-time payments, zero material audit findings)? Options: First-pass accuracy %, On-time payments %, Reduction in case backlog, Citizen satisfaction score, Zero critical audit findings, Other
• Which federal reporting outputs must exactly match legacy reports, and are there any tolerances for format or timing differences?
• What level of evidence (test results, sample audits, federal pre-acceptance) do you require before signing full cutover acceptance? Options: Automated test pass rates, Third-party sample audits, Federal pre-acceptance, Executive sign-off via committee, Other
• How should we handle continuous improvement post-cutover—regular retrospective cadence, shared issue tracker, or an ongoing governance board? Options: Weekly retrospectives, Monthly governance board, Shared issue tracker with SLAs, Ad-hoc as issues arise, Other
• Who will be accountable for documenting lessons learned and maintaining the shared enhancement backlog after deployment?

Start With the Big Picture — Who You Serve and Why It Matters

• Which benefit programs are you most urgently focused on modernizing right now? Options: SNAP, TANF, Medicaid eligibility, Unemployment insurance, Child care subsidies, Multiple of the above, Other
• How would you describe, in one sentence, the specific problem that keeps you up at night about this program?
• Who in your agency is ultimately accountable to the governor for program performance and federal timeliness? Options: Commissioner, Deputy Commissioner, Program Director, State CIO, Combined/Shared accountability, Other
• What recent trigger prompted you to consider a modernization partner now (select all that apply)? Options: Federal corrective action plan, Audit findings, COBOL system failure, Natural disaster / surge, Staffing shortages/caseworker turnover, Cost pressures, Other
• When you think about stakeholders—Governor’s office, federal partners, caseworkers—what single concern do you hear most often from them?
• How soon does leadership expect measurable improvements (e.g., accuracy, timeliness) after a partner is selected? Options: 3 months, 6 months, 9-12 months, 18+ months, No firm deadline

Are You Comfortable Letting 'Good Enough' Keep You Vulnerable?

• If you had to name one assumption your team keeps making about current performance that might be wrong, what would it be?
• How frequently have you experienced federal findings or corrective actions related to eligibility accuracy or timeliness in the last three years? Options: None, Once, Twice, Three or more
• What’s the emotional climate when corrective action is discussed—defensive, resigned, urgent, crisis-driven, or something else? Options: Defensive, Resigned, Urgent, Crisis-driven, Hopeful/optimistic, Other
• How does a missed timeliness target actually impact real people in your state—describe a recent example if available.
• What accuracy and timeliness thresholds would make federal auditors pause or withdraw an escalation? Options: >98% accuracy, 95–97% accuracy, 90–94% accuracy, Timeliness within federal standard consistently, Other

Where Do Your Legacy Systems Secretly Break Trust?

• How often do legacy COBOL rules produce incorrect determinations that only show up after federal reporting—daily, weekly, monthly, or less? Options: Daily, Weekly, Monthly, Quarterly, Rarely/Unknown
• Which failure modes from the legacy stack have the biggest downstream impact (select up to three)? Options: Incorrect eligibility calculations, Missed renewal triggers, Data loss in conversions, Batch processing delays, Integration timeouts with payment systems, Reporting mismatches with federal schema, Other
• Tell us about an edge case that the current system gets wrong or that your team still treats as a manual workaround.
• Do you have historical logs, test cases, and known edge-case datasets available for vendor validation, and where do they live? Options: Yes—comprehensive and accessible, Partial datasets available, Limited/no documented cases, We don't know yet
• How do current infrastructure limits (CPU, batch windows, network) shape when and how determinations run?

Who Really Calls the Shots When Decisions Matter?

• If we had to deliver a change that affects eligibility rules, who signs off at the state level—and who needs to be informed at the federal level?
• Where do decision bottlenecks usually occur—policy interpretation, legal review, procurement, IT change control, or stakeholder alignment? Options: Policy interpretation, Legal review, Procurement, IT change control, Inter-agency coordination, Other
• What is your internal decision timeline for contracting and implementation approval (weeks/months)? Options: <4 weeks, 4–8 weeks, 2–4 months, 4–6 months, Longer/uncertain
• How comfortable are your subject-matter experts explaining complex policy decisions to non-technical stakeholders (e.g., governor’s office or auditors)? Options: Very comfortable, Somewhat comfortable, Not comfortable, We lack SMEs
• Who would be our primary day-to-day contacts for rules, QA, reporting, and escalation during transition?

When the System Is Overwhelmed, Who Suffers—and How Fast?

• When a sudden surge occurs (e.g., disaster, economic shock), how many times normal volume have you experienced? Options: 1.5–2x, 2–3x, 3–5x, >5x
• What are the first three operational signs you see that indicate you are hitting capacity limits?
• How does backlog manifest—longer call wait times, increased manual reviews, missed payments, or other failures? Options: Longer call waits, Increased manual reviews, Late or missed payments, Federal reporting shortfalls, Other
• Describe the last time you executed an emergency surge play—what worked, what failed, and how long did recovery take?
• Which surge mitigations are available today (select all that apply)? Options: Cross-trained state staff, Temporary contractors, Automated triage rules, Extended batch windows, Alternate data centers, None of the above

If the Governor Asked You What 'Good' Looks Like, What Would You Say?

• What three measurable outcomes would satisfy the governor’s office (rank or list in order of importance)?
• Which federal metrics are non-negotiable for you—accuracy, timeliness, audit acceptance, or something else? Options: Accuracy, Timeliness, Audit acceptance, Consistent federal reporting, Citizen satisfaction
• What’s an acceptable tradeoff, if any, between reaching perfect accuracy immediately versus guaranteeing uninterrupted payments during transition? Options: Prioritize uninterrupted payments, Prioritize accuracy improvements, Balance with parallel processing, Undecided
• How would you like citizens’ experience to change—faster decisions, clearer communications, fewer manual calls, or other improvements? Options: Faster decisions, Clearer communications, Fewer manual calls, More self-service options, Other
• What sample success signal would you want to see within the first 90 days of a parallel run?

What Would a Vendor Absolutely Have to Promise to Make You Sleep at Night?

• Which contractual guarantees matter most to you—accuracy SLAs, staffing levels, reporting fidelity, or financial penalties? Options: Accuracy SLAs, Staffing guarantees, Federal reporting fidelity, Penalties/credits, Escalation pathways
• What minimum accuracy and timeliness SLA thresholds would you require in a contract? Options: ≥98% accuracy, 95–97% accuracy, 90–94% accuracy, Timeliness within federal standard, Other
• How important is vendor transparency—daily dashboards, audit logs, and open-book QA—to your decision? Options: Critical, Very important, Somewhat important, Nice to have, Not important
• If a vendor proposed a financial penalty for missed SLA, what type of remedy would feel proportionate to you? Options: Service credits, Termination rights, Remediation plans with third-party oversight, No penalties—collaborative approach
• Which federal reporting obligations must be explicitly covered in the vendor SOW (select all that apply)? Options: FNS/SNAP reporting, CMS/Medicaid reporting, DOL/unemployment reporting, Child care/federal grant reporting, Other

Can We Recreate The Weirdest, Most Painful Cases Before Cutover?

• What are the top five edge-case scenarios that, if missed, would trigger federal scrutiny or a public outcry?
• Do you have anonymized case files or synthetic datasets for each edge case, and can those be used by a vendor in testing? Options: Yes—full dataset, Partial/anonymized only, We can produce on request, No
• How do you currently validate rule changes—peer review, SME sign-off, parallel-processing comparison, or other? Options: Peer review, SME sign-off, Parallel-processing, Automated regression tests, Other
• What acceptance criteria would you require before a vendor’s rules engine is allowed to handle live cases (examples: match rate, false positive rate)?
• Who in your organization would be responsible for certifying that edge cases are handled correctly during transition?

What Would a Safe Transition Actually Look Like Day-to-Day?

• How do you feel about a year-long parallel run—comforting safeguard or unacceptable delay? Options: Comforting safeguard, Necessary but costly, Unacceptable—need faster cutover, Unsure
• Which day-one operational items must be in place before you allow any vendor-handled determinations to touch live payments?
• What rollback triggers would force you to revert to legacy processing during or after cutover? Options: Payment delays beyond X days, Accuracy below SLA, Federal reporting mismatch, Caseworker competency issues, Other
• How would you prefer escalation to work during cutover—daily war room, weekly steering meetings, or automated alerting with named contacts? Options: Daily war room, Twice-weekly steering, Weekly check-ins + alerts, Escalation only on incidents
• What training, shadowing, or certification would you require for vendor staff who will touch complex eligibility decisions? Options: Policy certification exam, Shadowing state SMEs, Periodic QA audits, Ongoing refresher trainings, Other

What Would Make You Confident To Move Forward — Right Now?

• What are the top three deal-breakers that would stop you from selecting a vendor today?
• Which evidence would most convince you—live throughput demos, audited performance reports, references from similar states, or a short pilot? Options: Live throughput demo, Audited performance reports, Peer references, Short pilot/POC, Other
• If we proposed a 90-day pilot with defined acceptance criteria, what would you require to feel that pilot was meaningful?
• Realistically, who needs to be on the approval chain for a contracting decision, and what would make them say yes?
• Finally, what immediate next step would you prefer—detailed SOW draft, technical deep dive with your SMEs, or a small-scale pilot design? Options: Draft SOW, Technical deep dive, Pilot design, Stakeholder alignment workshop, Other