Starting Together: What's at Stake for Your Team

• Briefly: what single event or trigger brought you to consider a master data pilot today? Options: Merger/Acquisition, Revenue reporting discrepancy, Regulatory requirement, Operational pain (duplicates), New executive mandate, Other
• Tell us in one short sentence: what outcome would make this conversation worth your time?
• How severe is the business impact today—are we looking at lost revenue, regulatory risk, operational cost, or reputational exposure (choose the best fit)? Options: Material revenue impact, Regulatory/compliance risk, High operational cost / inefficiency, Reputational risk, Multiple of the above, Unclear
• Who in your org feels the pain most acutely today (names/titles are fine)?
• Have you tried to fix this before (e.g., spreadsheets, scripts, existing MDM, ERP/CRM native tools)? If yes, what stopped it from scaling? Options: Spreadsheets/manual work, Homegrown scripts/tools, ERP/CRM native features, Legacy MDM vendor, Governance/political resistance, Never tried before, Other
• How quickly do you need proof that a phased MDM approach can deliver value? Options: Within 2 weeks, 2–6 weeks, 6–12 weeks, 3+ months, Unsure

If This Keeps Going, What Breaks First?

• If duplicate records and inconsistent IDs continue unchecked, what is the first measurable thing that will fail—reports, billing, regulatory filing, or executive trust? Options: Financial reporting, Customer billing/collection, Regulatory submissions, Sales/forecast accuracy, Customer experience, Executive confidence
• How often do you discover material errors (e.g., overstated revenue, duplicate accounts) in your regular reports? Options: Weekly, Monthly, Quarterly, Rarely, Don't know
• Can you share a recent example where master data issues caused a costly or embarrassing outcome? What happened and who was impacted?
• When data-driven teams distrust reports, what workaround do they use (manual reconciliation, ad-hoc queries, avoid using system)? Options: Manual spreadsheets, Ad-hoc SQL queries, Avoiding the system / siloed work, Rely on single source app (CRM/ERP), Other
• How would continued distrust in golden records affect upcoming strategic initiatives (M&A, new product launch, regulatory audit)?
• On an emotional level, how confident do executives feel about the numbers they make decisions from today? Options: Very confident, Somewhat confident, Wary / anxious, Not confident at all, They avoid making decisions

Who's Actually Driving the Decision—and Who Can Stop It?

• If governance were a tug-of-war, who are the teammates pulling you forward—and who could pull the rope the other way? Options: CDAO/Chief Data Officer, VP Data Management, CFO/Finance, Head of Sales/Revenue Ops, Legal/Compliance, IT/Platform, Line-of-business owners, Other
• Which role must sign off on a pilot (technical, business, legal, procurement)? Please list titles and decision criteria if known.
• How tolerant is your executive team of risk and experimentation—are pilots encouraged to fail fast, or do they require near-perfect results from day one? Options: High tolerance (experiment freely), Moderate (controlled experiments), Low tolerance (need predictable outcomes), Zero tolerance
• Who is likely to resist a neutral MDM hub governing records, and what are their main concerns? Options: Application owners (CRM/ERP), Regional data stewards, Legal/Privacy, Business unit leaders, Other
• If political resistance arises, what escalation path or sponsor exists to resolve it? Options: CDAO/CDO sponsor, CFO sponsor, CEO/Board escalation, No clear sponsor, Other
• What would convince your skeptical stakeholders that a single-domain pilot won't become a three-year enterprise rewrite?

Where Does the Data Live—and Where Is It Broken?

• If you had to hand us the single worst system for customer data right now, which system would that be and why? Options: Legacy CRM, ERP/Finance, Billing system, E-commerce platform, Custom application, Data warehouse/BI, Other
• Which systems contain the same entity (customer/product/supplier) that you want to reconcile for the pilot? Options: CRM, ERP, Billing/Payments, E-commerce, Marketing platforms, Data warehouse/BI, Third-party lists, Other
• Do you have a representative sample available for the pilot (e.g., 100K customer records across two systems)? If yes, where is it stored and who owns access? Options: Ready in DB/CSV, Ready but requires masking, Needs extraction, Not available yet, Unknown
• Describe the most common data quality problems you see (missing IDs, name variations, address inconsistencies, outdated contacts, duplicate account hierarchies). Options: Missing IDs, Inconsistent names/spellings, Address variations, Stale contact info, Duplicate hierarchies/parent-child issues, Incorrect account types, Other
• Approximately what portion of records do you suspect are duplicates or near-duplicates today? Options: <5%, 5–15%, 15–30%, 30–60%, >60%, Don't know
• Are there regulatory or contractual restrictions (PII handling, residency, masking) that would affect data extraction or processing for a pilot? Options: Yes—strict, Yes—manageable, No restrictions, Unsure
• Who currently owns the canonical identifiers or 'source of truth' for this entity (team/title)?

What Would a Trusted Golden Record Actually Enable?

• Imagine analysts stopped manually verifying merged records—what business activity would change first and most meaningfully? Options: Revenue reporting, Regulatory reporting, Customer service accuracy, Sales targeting/upsell, Billing reconciliation, Other
• What acceptance thresholds would convince business users to trust golden records without manual checks (e.g., match precision, recall, golden-record completeness)? Options: >95% precision, >90% recall, X% field completeness, Business-rule validation pass rate, User sign-off/trial period, Other
• Which KPIs will you use to judge pilot success (pick up to three)? Options: Reduction in duplicate accounts, Improved revenue accuracy, Time saved in reconciliations, Fewer regulatory exceptions, User trust / adoption rate, Reduced support tickets, Other
• What level of automation vs. human review is acceptable during pilot validation (fully automated merge, rule-based with human audit, human-in-the-loop for edge cases)? Options: Fully automated, Rule-based with sample audits, Human-in-the-loop for merges, Hybrid depending on confidence score, Prefer manual initially
• If we deliver golden records for one entity/domain, which downstream system should receive synchronized results first to demonstrate value? Options: CRM, Billing/Finance, Data warehouse/BI, Customer service platform, Marketing system, Other
• What would be the smallest, most convincing success signal you'd accept to greenlight scaling (example: 100K records matched with X% accuracy and stakeholder sign-off)?

What Could Scuttle a Pilot Before It Starts?

• If this pilot fails internally, what's the single most likely reason it would be killed? Options: Data access delays, Political/blocking by app owners, Poor match quality, Scope creep / shifting goals, Budget cuts, Legal/privacy issues, Other
• Which of the following is your team's biggest constraint right now (pick one)? Options: Limited engineering/ETL time, Legal/privacy review backlog, No clear sponsor, Unclear scope, Insufficient sample data, Competing priorities
• What mitigation already exists or could be mobilized quickly if we hit resistance (executive sponsor, technical champion, legal signoff template)? Options: Executive sponsor identified, Technical champion committed, Pre-approved data handling, Fast-track procurement, No mitigation currently, Other
• How comfortable are you with a phased pilot that intentionally limits scope to one entity/domain and one target system? Options: Very comfortable, Somewhat comfortable, Hesitant—prefer broader scope, Opposed
• What absolute non-starter constraints must we know now (e.g., cannot move PII off-prem, budget ceiling, legal non-negotiables)?
• Who needs to be looped in immediately to avoid delays (names/titles/email preference)?

Next Steps: Small Commitments That Prove Big Value

• Which pilot boundary feels most realistic to you for a first proof (pick one)? Options: Customer domain across 2 systems (100K records), Product domain in one system (50K SKUs), Supplier domain for a single region, Account hierarchies for one business unit, Other
• What sample size do you feel comfortable committing to load for initial evaluation? Options: 10–25K, 25–100K, 100–250K, 250K+, Unsure / want guidance
• How quickly could your team provide an extract or read-only access to the chosen systems? Options: Within 1 week, 1–3 weeks, 3–6 weeks, Longer than 6 weeks, Need approval
• Who would be the day-to-day owner for the pilot on your side (name/title), and who is the executive sponsor we should reference in our plan?
• Which acceptance milestone would trigger a commercial/phased expansion conversation (e.g., user sign-off, KPI hit, reconciliation reduction)? Options: Business user sign-off, KPI threshold met, Successful sync to target system, Risk/regulatory acceptance, Other
• What outstanding questions or hesitations should we address in the first week to keep momentum?
• Finally: what would you like to see in a one-page pilot plan from us by the end of next week? Options: Scope and timeline, Data access checklist, Acceptance criteria and KPIs, Roles and responsibilities, Commercial terms / phased pricing, All of the above

Where Does This Live Today?

• Which systems do you believe currently contain customer records we should consider for the pilot? (select all that apply) Options: CRM (Sales/Service), Billing/Finance system, ERP, Data warehouse / lake, Marketing automation, Order management, Support/Case system, Analytics/BI, Proprietary/line-of-business app, Other (list below)
• Which two source systems would you prioritize for sampling if we had to pick a focused pilot pair? Options: CRM, Billing/Finance, ERP, Data warehouse, Marketing, Support/Case, Custom app, Other (specify)
• Roughly how many customer records exist in each prioritized source (estimate per system)? Options: <10k, 10k–50k, 50k–100k, 100k–500k, 500k–1M, >1M, Unknown / need to investigate
• Who are the technical contacts or teams that own extracts for these systems? Please list names and roles.
• Are automated extracts permitted from these systems, or do we need manual exports/approvals? Options: Full automated extracts permitted, Extracts allowed with approvals, Only manual exports allowed, Extracts restricted for certain fields (PII), Unsure — need to confirm
• How often do those systems currently synchronize customer information with each other, if at all? Options: Real-time sync, Daily batch, Weekly, Ad-hoc/manual, No synchronization today, Unsure

Are We Counting Duplicates as Real Customers?

• If a month-end revenue report overstated sales because duplicates were treated as separate customers, how would that change how your leadership prioritizes this work? Options: Immediate top priority, High priority but needs proof, Back-burner until compliance forces it, Not a priority / skeptical, Unsure
• Can you describe a recent incident where duplicates or inconsistent customer records caused a measurable business problem (revenue, regulatory filing, customer experience)? Tell us what happened and who noticed it.
• How frequently do downstream teams request manual reconciliations or spreadsheets to resolve customer identity issues? Options: Daily, Weekly, Monthly, Quarterly, Rarely, Never
• Which business processes suffer most when customer identity is unclear (select top three)? Options: Revenue recognition / billing, Regulatory reporting / compliance, Customer service / case routing, Sales pipeline accuracy, Fraud detection, Marketing segmentation, Credit/limits/collections, Other
• When duplicates are found today, how are they resolved and how long does resolution typically take? Options: Automated merge rules, Manual analyst reconciliation, Business owner adjudication, Left unresolved, Other — describe
• How does the existence of duplicates make your stakeholders feel—frustrated, distrustful of reports, anxious about audits, or something else? Options: Frustrated, Distrustful of data, Anxious about audits, Indifferent, Other — describe

What's Hidden in the Data?

• If we pulled a random sample of 10,000 records from your prioritized systems today, roughly how many would you expect to be missing critical matching attributes (email, phone, postal address, tax ID)? Options: <10%, 10–25%, 25–50%, 50–75%, >75%, Don't know
• Which fields do you view as essential for reliable matching in your environment (pick all that apply)? Options: Full name, Email, Phone, Postal address, Company registration / tax ID, Account number, External identifiers (CRM ID), Other — specify
• Are global or authoritative identifiers available and consistent (tax ID, national ID, vendor codes), and for what proportion of records? Options: Yes, for most records, Partial coverage (20–50%), Rarely available, Never available, Unsure
• Describe the most common data quality patterns you've noticed (formatting issues, transposed fields, legacy codes, merged accounts). How often do these patterns repeat?
• Do you have PII, residency, or regulatory rules that limit which fields we can copy or process for a pilot? If so, which ones? Options: No restrictions, PII allowed with anonymization, PII restricted (cannot be moved), Cross-border residency restrictions, Need legal sign-off, Unsure
• How long would your team need to assemble and secure a 100k-record sample (including approvals and anonymization if needed)? Options: <1 week, 1–2 weeks, 2–4 weeks, >4 weeks, Depends on approvals

Who Holds the Keys—and Who Says No?

• Who in your organization could effectively block a pilot because they fear losing control or ownership of their system’s data—and what is their primary concern?
• Which stakeholders must be engaged or sign off before we proceed (pick all that apply)? Options: VP Data Management, CDO / Head of Data, IT Ops / Platform team, Application owners (CRM/ERP), Legal / Privacy, Finance, Business unit leaders, Other — specify
• How have system owners responded to past centralization or shared-governance efforts—were they cooperative, resistant, or conditionally supportive? Options: Cooperative, Resistant, Conditionally supportive, Indifferent, No past interactions
• What governance model is currently in place for master data (centralized, federated, business-unit owned, none)? Options: Centralized, Federated, Business-unit owned with informal rules, Ad hoc / no formal governance, Hybrid
• When there’s disagreement about a golden record, what decision mechanism would your organization trust most—technical rules, business owner adjudication, or an executive sponsor? Options: Technical deterministic rules, Business owner adjudication, Executive sponsor decision, Third-party arbitration, Combination
• How long has governance been an active agenda item, and what internal politics should we be mindful of during a pilot?

If We Ran a Pilot Tomorrow—Would You Trust the Results?

• What minimum golden record accuracy or trust threshold would let your analysts stop manual verification (select one)? Options: ~90%, ~92–94%, ~95–96%, ~97–98%, ≥99%, Need different metric / unsure
• Which acceptance KPIs matter most to your team for pilot success (pick up to three)? Options: Precision / false-positive rate, Recall / false-negative rate, F1 or composite match score, Reduction in manual reconciliations, Time-to-resolution for duplicates, Downstream reconciliation errors, Regulatory accuracy metrics, Other — specify
• What sample size do you feel is persuasive for evaluation: 10k, 50k, 100k, 500k, or something else? Options: 10k, 50k, 100k, 500k, Other — specify
• Are there acceptance constraints tied to external auditors or regulators that would affect whether a pilot's results are deemed valid? Options: Yes — specific auditor/regulator rules apply, Possibly — needs legal review, No external constraints, Unsure
• Who in your organization will formally accept pilot results and what practical criteria will they use (names/roles and criteria)?
• If we demonstrate golden records that meet your thresholds, what would you consider an appropriate next step to scale (expand systems, add entity types, governance ramp)? Options: Expand to second system pair, Add one new entity domain, Move to phased roll-out, Create center-of-excellence, Other — describe

What Could Break This Before It Starts?

• What's the one failure mode that would make this pilot feel like 'another failed MDM project' in your organization?
• Do you face legal, privacy, or data residency constraints that could prevent us from copying or processing data for matching? Options: Strict constraints — additional approvals required, Allowed with anonymization, No constraints for pilot data, Unsure — need legal review
• What internal resourcing limitations (people, time, connectivity) are most likely to delay a sample extract or the pilot itself? Options: Lack of engineering time, Security/legal approval delays, Business owner bandwidth, Network / access issues, Other — specify
• Are there competing initiatives or change fatigue that could cause stakeholders to deprioritize this work? If yes, which projects?
• What budget or procurement risks could cause the pilot to be paused, and how have you mitigated similar risks before?
• If issues appear during the pilot, what escalation path would keep the work alive (who do we call and what authority do they have)?

Fast Wins or Long Wars—Which Path Do You Prefer?

• If nothing changes in six months, which metric or process will be noticeably worse—and why should we act now?
• Which quick win would deliver visible, business-facing value in weeks rather than months (pick one)? Options: De-duping billing lists to fix revenue reporting, Matching CRM and billing records for single-customer view, Cleaning top revenue-generating customer records, Synchronizing golden records to one target system, Other — specify
• Are you open to a phased delivery that starts with one entity domain and expands only after business trust is established? Options: Yes — prefer phased approach, Maybe — depends on scope, No — prefer broader scope, Unsure
• How would you like progress and confidence reported during the pilot (select all that would be useful)? Options: Interactive dashboard, Weekly executive summary, Working sessions with analysts, Automated quality metrics, Milestone sign-offs, Ad-hoc workshops
• What timeline feels realistic to produce trusted golden records for a pilot pair (from sample receipt to validated results)? Options: <2 weeks, 2–4 weeks, 4–8 weeks, 2–3 months, >3 months
• What is the next concrete decision or meeting we should schedule to move this pilot forward, and who needs to attend?

Quick Win Snapshot

• In one sentence, what single outcome must this pilot prove for your organization to call it a success?
• Which primary business metric will be used to judge pilot success? Options: Duplicate rate reduction, Revenue reconciliation improvement, Ability to produce single-customer regulatory report, Reduction in manual reconciliation effort, Time-to-identity (median), Other
• Which sample size and system pair would you consider persuasive for a vote to scale? Options: 10,000 records from two systems, 50,000 records from two systems, 100,000 records from two systems, A specific cohort or geography (specify below), Other
• Who must sign off on pilot success before funding or scaling is approved (list roles/titles)? Options: CDAO / Head of Data, VP Data Management, Head of Finance, Head of Risk & Compliance, Business Unit Lead, Other
• Ideally, how soon would you like the pilot to start once data access is confirmed? Options: Immediately (within 2 weeks), 2–4 weeks, 1–2 months, 3+ months, Unsure

If This Fails, Who Will Notice?

• If the pilot doesn't deliver as expected, whose credibility or KPIs are most at risk—what makes that person vulnerable?
• Which teams will feel operational pain first if we cannot produce a reliable golden record? Options: Finance/Revenue Reporting, Risk & Compliance, Sales / RevOps, Customer Support, M&A / Integration Team, Other
• Do you have a formal escalation path or rapid decision forum that can resolve pilot blockers in days rather than weeks? Options: Executive steering committee, Program manager + SMEs, Ad-hoc escalation via sponsor, No formal escalation path
• What political or ownership conflicts have blocked previous cross-system data efforts in your organization?
• How would a failed pilot change sponsorship, budget, or the appetite for phased MDM approaches? Options: Pause further funding, Require larger proof before next phase, Shift to a different vendor/approach, No change, Unsure

Where Are Your Best — and Worst — Records Hiding?

• Which system(s) do you currently treat as the most 'trusted' source, and why might that trust be misplaced?
• List the source systems we plan to include in this pilot (system name, primary owner, and contact) — start with top two.
• Are sample extracts for those systems available now, or will extract work be required? Options: Ready and accessible, Extracts exist but need rework, No extracts yet—access pending, Extracts blocked by security/legal
• What percentage of records in-scope contain the minimum fields we need (name, address, customer ID, contact methods)? Options: <50%, 50–75%, 75–90%, >90%, Unknown
• Describe the three most common data quality problems you see (e.g., misspellings, merged accounts, placeholder IDs) and how long they've existed.
• Are there regulatory cohorts or constrained data subsets we must include or explicitly exclude from pilot extracts? Options: Include specific regulated cohorts, Exclude regulated PII, No constraints, Unsure—need legal input

Who Actually Needs Keys to the Castle?

• Who claims they need full access today but would be satisfied with read-only, sampled, or tokenized views?
• Which data extraction patterns are acceptable for the pilot? Options: Database replicas/DB dumps, SFTP CSV extracts, API connections (REST/GraphQL), ETL/ELT handoff (Informatica, Fivetran), Manual exports
• What level of PII masking or tokenization is required before data can leave the source environment? Options: Full masking of PII, Partial masking (tokenize IDs), No masking—raw data allowed, Depends on dataset
• Who approves credential or connection requests (system owner, IT Sec, data governance)—please list names/roles for each source system.
• Are there network constraints (firewalls, VPN windows, maintenance windows) that limit our extract times? Options: Yes—strict maintenance windows, Yes—some restrictions, No constraints, Unknown—need IT confirmation
• Which communication channel do you prefer for extract requests and status tracking? Options: Shared tracking board (Jira/Confluence), Weekly status email, Daily standups, Ad-hoc via Slack/Teams

What Would Block Even a Perfect Extract?

• What tiny, technical detail keeps you awake at night because it could derail the first extract?
• Have you provided a data dictionary or schema for the systems in scope? Options: Complete and up-to-date, Partial or outdated, No, Work in progress
• Are there known schema mismatches (field types, encodings, multi-value fields) we should plan to reconcile? Options: Many mismatches, A few known mismatches, No significant mismatches, Unknown
• Do any source systems impose API rate limits, export quotas, or performance limits that could throttle our extract? Options: Yes—strict limits, Yes—manageable limits, No limits, Unknown—need confirmation
• Which approach do you prefer for field mapping and transformation during the pilot? Options: You provide mappings, Vendor proposes mappings and we review, Collaborative mapping workshops, Use default mappings and adjust
• If our match-and-merge needs a derived field (e.g., normalized address score), who approves creating that field upstream?

When Will We Fix the Mess You Already Know About?

• Are you willing to accept imperfect source data for faster time-to-value, or do remediation tasks need to be completed first? Options: Accept imperfect and iterate, Fix critical issues before start, Require near-perfect before pilot, Unsure—need discussion
• List the top three data quality issues you want addressed during the pilot and name the owning team for each (owner + rough ETA).
• Which remediation levers are you willing to include in pilot scope? Options: Automated standardization (addresses/names), Rule-based deduplication, Manual review queues, Third-party enrichment (phone/address), None—pilot must avoid remediation
• Do you have data stewards or SMEs available to support manual review and triage, and if so, how many hours/week can they commit? Options: >20 hours/week, 8–20 hours/week, 4–8 hours/week, <4 hours/week, None available
• If we surface systemic upstream issues that require schema or process changes, what is your preferred remediation cadence? Options: Immediate exec notification + stop/hold, Weekly remediation meeting, Triage via ticketing and SLA, Ad-hoc based on severity
• Which data fixes are absolute 'musts' versus 'nice-to-haves' for the pilot to deliver credible results?

How Will We Know the Golden Record Is Trustworthy?

• What would make business users refuse to adopt the golden record even if our algorithms show high accuracy?
• Define the minimum acceptance thresholds you expect for match/merge quality (choose closest). Options: Precision >= 95%, Precision >= 90%, Precision >= 85%, No firm numeric threshold—need business validation
• Which validation methods should we use during the pilot to build trust? Options: Business user manual review, Statistical sampling and reports, Automated scoring with thresholds, Reconciliation vs financial/regulatory reports, Other
• Who will act as the business validators and how will their feedback be captured and prioritized?
• How many manual reviews or what percentage of matched records will make you comfortable signing off (specify absolute or percent)? Options: 100 records, 500 records, 1,000 records, Specify a percentage in the next field, Other
• Which downstream targets must receive synchronized golden records during the pilot and how will we confirm successful ingestion?

Final Approvals, Timelines & What 'Go' Looks Like

• If someone is poised to veto the pilot at the last minute, what evidence or artifact would convince them to change their mind?
• Which artifacts are required for pilot kick-off and sign-off? Options: Signed scope document, Data access approvals, Security assessment/SAQ, Commercial agreement, Governance & escalation plan, Other
• Who holds final go/no-go authority and how do they prefer to sign off (email, ticket, steering committee)? Options: CDAO / Exec Sponsor, VP Data Management, Program Sponsor via email, Steering committee decision, Other
• Which pilot timeline do you prefer for extract → run → validation → signoff? Options: 2–3 weeks (rapid), 4–6 weeks (standard), 8–12 weeks (thorough), Custom timeline—specify below
• Is there contingency budget or backup resources available if the pilot requires additional engineering/time? Options: Allocated contingency budget, No contingency—must stay on plan, Can request additional funding, Unsure
• What are the non-negotiable 'must-haves' we should list in the pilot charter to prevent scope creep and political pushback?

Quick hello: What brought you into this conversation today?

• In one sentence, what is the immediate trigger (e.g., post-merger duplicates, audit/regulatory request, revenue discrepancy)? Options: Post-merger integration, Regulatory request, Revenue/reporting discrepancy, Operational pain (support/ops), Other
• Which entity domain is highest priority right now? Options: Customer/Account, Product/Catalog, Supplier/Vendor, Financial/GL Accounts, Other
• Which source systems contain the overlapping records we should evaluate? Options: CRM (Salesforce), CRM (Dynamics), ERP (SAP/Oracle), Billing/Revenue system, Marketing database, Data warehouse / lake, Custom application, Other
• Approximately how many records exist across the combined systems for this domain? Options: <10k, 10k–100k, 100k–1M, 1M–10M, >10M, Unsure / need to estimate
• Who on your team will most care about the outcome of the pilot (title or role)?
• How soon do you need a demonstrable result that the business can trust? Options: Within days, 1–2 weeks, 3–4 weeks, 1–3 months, Unsure

Are you comfortable running blind?

• How often have duplicate or inconsistent master records led to decisions you later had to reverse or apologize for? Options: Almost always, Often, Sometimes, Rarely, Never
• Tell us about a recent decision or report that felt risky because the underlying master data was untrusted—what happened and who noticed?
• When incorrect master data impacts a decision, which teams feel the pain most acutely? Options: Finance / Revenue Ops, Sales / Account Management, Compliance / Legal, Customer Support, Supply Chain / Ops, Executive leadership, Other
• How does unresolved master data uncertainty usually show up in day-to-day work (manual reconciliations, duplicate invoices, inaccurate targets, etc.)?
• If we fixed the single biggest source of confusion in your master data, what immediate change would you expect to see in the next quarter?
• Which of these feelings best describes how your team approaches master data today? Options: Resigned and workaround-driven, Cautious and experimental, Eager but under-resourced, Optimistic — ready to commit, Skeptical — seen failures before

Where the numbers break your heart

• Has a report, financial statement, regulatory filing, or customer interaction been materially wrong because of master data? What was the outcome? Options: Yes — financial impact, Yes — regulatory/penalty, Yes — customer impact, Yes — internal operational failure, No clear incident, Prefer not to say
• If there was a financial or regulatory impact, can you estimate the scale (order of magnitude) or describe the downstream consequence?
• Which KPIs or reports do you distrust today because of master data issues? Options: Revenue by customer, Customer lifetime value, Risk exposure / credit limits, Product sales mix, Supply chain inventory, Other
• How often do you have to run manual reconciliations or supplemental checks to validate a single report? Options: Every report, Often, Sometimes, Rarely, Never
• When those reconciliations occur, who is accountable for the manual work and how long does it typically take?
• If we could remove that manual reconciliation workload, what would your team do with the freed time? Options: Strategic analysis, Faster closes/reporting, Customer outreach, Process improvements, Other

Who's actually calling the shots?

• Who will sign off on pilot acceptance and who holds veto power for scaling MDM? Options: VP Data / CDO, CIO, Finance Controller, Head of Sales Ops, Legal/Compliance, Business Unit Leader, Other
• Describe any political or organizational tensions that could block a neutral master record hub (system owners, domain stewards, regional leads)?
• Have there been previous MDM or data governance initiatives here that failed or stalled? What was the main reason? Options: Scope creep / too big, Lack of executive sponsorship, Insufficient resources, Resistance from system owners, Technology mismatch, Other
• What does the ideal sponsor look like for this pilot — title, appetite for change, and level of involvement?
• Which stakeholders must be kept informed throughout the pilot (and how often)? Options: Weekly email/status, Bi-weekly review, Monthly steering committee, Ad-hoc updates, Only at milestones
• If someone resists centralizing a golden record, what would convince them to try a phased approach instead of saying no?

If we could snap our fingers and produce one trusted golden record, what would change?

• What acceptance threshold would make your analysts stop manual verification — e.g., a percentage accuracy, types of automated checks, or user confidence signals? Options: >=99% accuracy, >=97% accuracy, >=95% accuracy, Business sign-off required, Automated reconciliation passes, Other
• Which data elements absolutely must be correct in a golden record for you to feel confident (e.g., legal name, tax ID, billing address, consolidated revenue)?
• How should we show provenance and explainability so your team can trust every automated merge? Options: Record-level audit trail, Merge rationale / match score, Side-by-side sample comparisons, Ability to easily revert merges, Automated reconciliation reports, Other
• What business tests should we run against golden records during the pilot to demonstrate trust (e.g., revenue reconciliation, AR cleanup, duplicate reduction)?
• Which user workflows would need to change when golden records are synchronized to downstream systems? Options: Data steward reviews, Automated syncs to apps, Exception handling by business users, New governance approvals, No workflow changes
• Beyond accuracy, what emotional proof point will convince stakeholders—reduced firefighting, fewer audit findings, faster closes, or something else? Options: Reduced firefighting, Fewer audit findings, Faster month-end close, Less finger-pointing, Improved customer experience, Other

What would a safe, non-boil-the-ocean pilot actually look like?

• If we had to prove value in a single phase, which scope would you prefer we lock to start? Options: Single entity type (e.g., Customer), Single system pair (CRM→ERP), Top revenue accounts only, High-risk regulatory segment, Other
• Which sample selection approach do you trust for evaluating match-and-merge quality? Options: Random sample, Top revenue customers, High-risk / flagged records, Representative by region or BU, Other
• What sample size do you consider sufficient for an initial demo that the business can rely on? Options: 5k–10k, 10k–50k, 50k–100k, 100k–500k, 100k specifically (industry benchmark)
• Which matching rules or tolerances are non-negotiable (e.g., exact tax ID match, fuzzy name tolerance level)?
• Who will be the day-to-day point person on your side for pilot execution (data engineer, steward, project manager)? Options: Data engineer, Data steward, Project manager, Business analyst, Other
• What is the minimal success criteria for this pilot that would make you greenlight phased scale? Options: Accuracy threshold met, Business sign-off on reconciliations, Operational syncs validated, No critical issues in audit, Other

Data's dirty little secrets — what are we going to discover?

• What's the worst-case data issue we might uncover in the sample that would surprise your team? Options: Missing critical IDs, Systemically incorrect addresses, Multiple IDs for same customer, Conflicting legal entities, Legacy merges gone wrong, Other
• Which of these data quality problems are present today? Select all that apply. Options: Missing/blank fields, Inconsistent formatting, Stale/outdated records, Duplicate records across systems, Conflicting attribute values, Incorrect IDs (SSN/Tax ID), Other
• Do you have representative sample extracts ready or will extracts require engineering effort and approvals? Options: Extracts ready, Extracts need engineering work, Requires legal/security approval, Unsure
• If extracts require approval, what security or compliance constraints must we observe (data masking, on-prem processing, VPN, anonymization)? Options: Anonymize PII, Masked fields, On-prem deployment only, Encrypted transfer (SFTP/VPN), SOC2 or ISO-only vendors, Other
• Who owns data quality remediation in your org and how much runway do they have to do pre-pilot fixes? Options: Data quality team, System owners, Business units, Shared responsibility, No dedicated owner
• What do you fear most we’ll find in the sample, from a people or politics perspective (blame, budget fights, exposure)?

Red lines, escalations, and success signals — how will we know it's safe to scale?

• What explicit go/no‑go criteria would force you to pause scaling after the pilot? Options: Accuracy below threshold, Critical data lineage gaps, Unresolved compliance issues, Business user rejection, Integration failures, Other
• Who must sign the go/no-go and what level of evidence do they require (demo, reconciliation report, executive summary)? Options: VP Data / CDO, CIO, Finance Controller, Business Unit Leader, Legal/Compliance, Other
• What monitoring or post-pilot dashboards would give you comfort after go-live (duplicate counts over time, sync success rate, exception queue length)? Options: Duplicate rate trend, Sync success/failed counts, Exception backlog size, Audit trail of merges, Business reconciliation KPIs, Other
• If we hit an unexpected data issue during pilot execution, what escalation path should we use and who is the emergency contact? Options: Project manager escalation, Weekly steering committee, Direct exec escalation, Security/legal escalation, Other
• What rollback or remediation controls must be in place before you will accept automated merges into any production system? Options: One-click revert, Staging sync only, Manual steward approval, Versioned golden record with audit, Other
• What post-pilot governance cadence would you prefer for phased rollout decisions? Options: Weekly for first month, then monthly, Bi-weekly through 90 days, Monthly steering committee, Ad-hoc as issues arise

Commitment and next steps — the minimum you need to feel safe to start

• What is the smallest, non-negotiable commitment your organization must make to enable a truthful pilot (people, access, budget)?
• Which of these resources can you commit immediately to the pilot? Options: Data engineer (part-time), Data steward (part-time), Business SME (weekly reviews), Project manager, Security/legal reviewer
• How long will it take to produce the sample extracts once approvals are in place? Options: <3 business days, 3–7 business days, 1–2 weeks, 2–4 weeks, Longer / unsure
• Which timeline feels realistic for a sample run and initial results your business can review? Options: 3–7 days, 1–2 weeks, 3–4 weeks, 1–3 months
• What are the primary blockers we should help you remove before kickoff (select all that apply)? Options: Legal/data access, System owner approval, Resource availability, Budget approval, Data quality cleanup, Other
• What's the single next meeting or decision you want from us to make it easy to move forward?
• Any final concerns or unspoken risks you'd like us to explicitly address before we schedule the pilot?