ISC · Information Systems and Data Management · 9 practice questions
Customer MDM After a Merger: Fix Duplicate IDs Across Systems
Master data management creates a governed customer identity and synchronizes authoritative updates across operational systems. Below: compare proposed responses to duplicate customer records after a merger.
Try one first
Hint
Focus on a solution that (1) creates a single authoritative ID for analytics, (2) preserves existing source IDs and linkages for historical transactions, and (3) avoids intrusive changes to production OLTP systems.
Answer C. An MDM hub centralizes identity resolution and creates golden customer records by using deterministic and probabilistic matching rules; it assigns a global surrogate ID and persistently maps that ID to each source system's native IDs. This preserves referential integrity for historical transactions without requiring intrusive changes to production OLTP systems, and it provides consistent, repeatable identifiers for reporting and analytics.
Why not A: Tempting because it appears to 'solve' the mismatch at the source, but it requires risky, intrusive changes to production systems and extensive updates to historical transactional references, operationally disruptive and unnecessary when an external mapping can preserve linkages.
Why not B: Deterministic composite keys look simple, but names and addresses are error-prone, often non-unique, and subject to formatting/typo differences; this leads to missed duplicates or false matches and cannot reliably preserve referential integrity for historical transactions.
Why not D: On-demand matching minimizes upfront work, but without a persistent mapping it produces inconsistent results across queries, offers no stable identifier for linking historical transactions, and shifts matching costs to query time which can harm performance and governance.
What should you do?
The situation
After an acquisition, Apex retains separate CRM, ERP, and billing applications. Customers have different IDs, inconsistent addresses, and no consistently populated shared email field. Historical invoices reference legacy IDs. CRM is the approved source for contact addresses; billing verifies tax status. A read-only warehouse receives nightly extracts, but operational records still conflict. Management wants an ongoing authoritative customer record without replacing the applications. Which response is best?
Implement an MDM hub with enterprise IDs, stewardship, field-level authority, and operational synchronization.
Best Use probabilistic matching and steward review to resolve identities. Retain legacy-ID mappings, apply survivorship rules using approved field owners, and distribute mastered records.
Deduplicate customer records only in the warehouse during nightly ETL.
Wrong Reports may improve, but CRM, ERP, and billing still use conflicting customer records. A read-only warehouse does not synchronize authoritative operational updates.
Make CRM authoritative for every customer field and export its values.
Wrong CRM owns contact addresses, not verified tax status. Forcing every field into one source ignores approved attribute-level authority.
Clean existing duplicates once, then let each application maintain customers independently.
Wrong The cleanup fixes current records only. Independent creation and updates can recreate duplicates without enterprise identifiers, stewardship, and ongoing governance.
Synchronize records automatically without defining ownership or survivorship rules.
Wrong Synchronization can spread conflicting values. Apex still needs rules and accountable stewards to decide which records match and which field values survive.
Replace all legacy IDs and rewrite historical invoice references.
Wrong This creates unnecessary disruption. Persistent mappings between enterprise and legacy IDs preserve historical links without rewriting every source identifier and transaction reference.
Key points
- Remember: A clean warehouse report does not fix conflicting customer records used in daily operations.
- Map legacy IDs to the enterprise ID so historical transactions remain linkable.
- Use probabilistic matching when no reliable shared identifier exists; send ambiguous matches to stewards.
- Survivorship rules select authoritative field values; one application need not own every field.
How the exam traps you
- Deduplicate only during nightly warehouse loads. Synchronize governed master records to operational systems, not just reporting copies.
- Make CRM authoritative for every customer field. Assign authority by attribute, preserving billing's ownership of verified tax status.
- Clean existing duplicates once and keep local update procedures unchanged. Govern matching, creation, and updates continuously to prevent recurrence.
Question 2
Hint
Focus on which choice best creates one reliable source for the same data element across departments.
Answer C. A centralized vendor master file with unique IDs establishes a single authoritative source for vendor data. That directly addresses duplicate records and inconsistent values across departments. Controlled update procedures also improve data integrity by reducing conflicting changes and unauthorized edits.
Why not A: Keeping history longer may aid audits but does not resolve duplicate current records or conflicting vendor information across departments. It addresses preservation, not present-day data consistency.
Why not B: This is tempting because reconciliation can catch inconsistencies and may seem like a practical fix. However, it is only a detective, after-the-fact process and does not eliminate the underlying problem of multiple competing data sources; the question asks for the most direct improvement, which is a single source of truth.
Why not D: Encryption and access controls protect data confidentiality and limit unauthorized edits, but they do not create a single authoritative dataset; separate secured spreadsheets can still contain conflicting vendor information.
Question 3
Hint
Decide whether the inconsistencies should be corrected before records enter the ERP and who should own ongoing data quality.
Answer D. Pre-migration profiling and cleansing produce a golden master so the ERP receives consistent, high-quality input, reducing downstream reconciliation, operational errors, and reporting problems. Automated matching supplemented by targeted manual review addresses ambiguous matches and incomplete identifiers more reliably than deferring cleanup. Appointing a master-data owner with governance rules prevents recurrence and supports ongoing data quality.
Why not A: Tempting for preserving original records and minimizing upfront effort, but wrong because maintaining duplicates degrades operational reliability and analytics. A cross-reference preserves identifiers but does not produce the single trusted record required for consistent transactions and reporting.
Why not B: Tempting because it appears faster and leverages built-in ERP tools, but wrong because post-migration deduplication is often less effective when required identifiers are missing and it shifts bulk cleanup onto users. It also leaves no assigned owner or governance to prevent the issue from recurring.
Why not C: Tempting as a decisive shortcut, but wrong because one legacy system may not be the most accurate source and overwriting without profiling can propagate errors. Deferring governance until after go-live removes the rules needed to resolve conflicts during migration.
Question 4
Hint
Focus on which option prevents duplicate or conflicting versions of the same data from being maintained in different places.
Answer C. Creating a master data management (MDM) solution with a single authoritative customer record prevents divergent updates by ensuring all systems use the same source for current addresses. This is a preventive, ongoing fix that eliminates the root cause of inconsistent active data across systems, rather than just detecting or recovering from errors after they occur.
Why not A: Backups help with recovery after bad changes but do not stop inconsistent or conflicting updates from occurring across active systems. They address availability and restore, not ongoing data consistency.
Why not B: This may find mismatches, but reconciliation is detective and periodic; it does not prevent departments from continuing to make inconsistent updates during the month. The question asks for an ongoing solution that reduces errors at the source.
Why not D: Archiving affects storage and performance of historical records, but the problem described is inconsistent current address data across systems. Archiving does not create a single authoritative source for active customer addresses.
Question 5
Hint
Classify each dataset by role and lifecycle (master/reference vs transactional vs audit) and choose governance, archival, and retention controls appropriate to each category.
Answer A. Customer master data functions as reference/master data used across processes and therefore requires MDM-level governance (centralized change workflows, validation, deduplication, RBAC and auditability) to maintain data quality. Sales transactions are authoritative financial records that need reliable backups and archival to indexed read-only storage with integrity checks and retention schedules driven by tax and audit requirements. Audit logs support security and forensics and should be preserved in tamper-evident storage with a compliance-driven retention policy rather than treated the same as operational data.
Why not B: Tempting because a uniform approach seems administratively simple and backups are broadly protective, but master, transactional, and audit data have different lifecycles and governance requirements. Purging customer master records after one year risks losing referential integrity and important PII/tax identifiers needed for audits and ongoing business processes.
Why not C: This option appeals to flexibility and cost-cutting but misclassifies master data and removes essential governance, free-form edits and no deduplication will harm data quality across systems. Stopping backups for archived transactions and retaining audit logs for only 90 days can jeopardize recoverability, financial reporting, and forensic/compliance needs.
Why not D: Encryption and RBAC are important controls, but relying on them alone and keeping everything online/editable ignores the need for MDM for reference data, appropriate archival and integrity checks for older transaction data, and tamper-evident retention for audit logs, creating compliance and long-term storage risks.
Question 6
Hint
Focus on the control that improves data quality and consistency across systems, not the ones that mainly protect or recover data.
Answer C. Master data management (MDM) creates a single authoritative source for key entities like customers by standardizing definitions, assigning data owners, and controlling approved changes. Synchronizing approved customer records across systems reduces duplicates and ensures consistent management reports. This directly addresses data quality, consistency, and accountability, the problems described in the scenario.
Why not A: Encryption protects confidentiality of data in storage and transit but does not address data quality, duplicate records, or inconsistent values across applications. Secure systems can still contain conflicting customer information if governance and synchronization are absent.
Why not B: Backups and restore testing protect availability and enable recovery from data loss, but they preserve whatever data state exists at backup time. They do not eliminate duplicate or inconsistent records or create a single authoritative customer record across systems.
Why not D: Quarterly, department-level reconciliation is a manual, delayed, and decentralized workaround that allows multiple versions to persist between reconciliations. Management asked for one reliable version and clear ownership, which requires centralized data governance and MDM rather than periodic local fixes.
Question 7
Hint
Think preventive and enterprise-wide: which single change would create one authoritative customer record across all systems rather than just detecting mismatches or limiting who can change local records?
Answer B. An enterprise MDM program creates a single authoritative customer record and unique enterprise identifier, preventing divergent copies of customer data and enabling automated, synchronized updates across all transactional systems. Because it addresses the root cause, the existence of multiple independent masters, it's a preventive, enterprise-wide fix that reduces duplicate invoices, missed collections, and revenue-recognition errors and lowers ongoing manual reconciliation. Data stewards and governance sustain data quality over time.
Why not A: Monthly reconciliations are primarily detective and labor-intensive: they identify and correct mismatches after errors have occurred. They do not prevent divergence between systems between cycles, so issues can recur and operational costs remain high; they also fail to create a single source of truth.
Why not C: Referential integrity improves local database consistency and helps prevent invalid or orphaned records within a system, which is beneficial. However, it does not harmonize separate customer masters or establish an enterprise-wide identifier, so cross-system duplicates and conflicting records can still persist.
Why not D: Stricter access controls and segregation of duties reduce the risk of unauthorized or improper changes, which is useful, but they do not resolve existing duplicates or inconsistent data maintained independently across systems. Without an enterprise master or synchronization, inconsistencies and related revenue-processing errors remain.
Question 8
Hint
Focus on the control that fixes the source of the inconsistency, not a control that only preserves or later compares bad data.
Answer C. A single authoritative master file addresses the root data management problem of duplicate and inconsistent source data. Assigning data ownership and using edit checks helps ensure accuracy and integrity when records are created or changed, preventing errors rather than merely detecting them later.
Why not A: Longer retention helps with historical research but does not address current data inconsistencies across departments. Retention policies don't create a single reliable current record.
Why not B: Reconciling separate files can detect mismatches, but it preserves multiple competing records and is a detective, after-the-fact workaround. It does not prevent source inconsistencies and is less effective than establishing a single authoritative file with ownership.
Why not D: Backups improve availability and recovery but merely preserve whatever (potentially incorrect) data exists. They do not resolve conflicting master records or improve data governance and quality.
Question 9
Hint
Ask where the inconsistencies originate and when they are detected; prioritize preventive, source-level fixes that establish one authoritative customer record.
Answer C. The root causes are lack of an enterprise identifier, missing lineage/metadata, and quality checks only during ETL. An enterprise MDM program creates a golden record and unique identifier, documents lineage/metadata, and enables source‑level validation and standardization to prevent duplicates and inconsistencies from propagating into reports. Security hardening, retention policies, or warehouse-only reconciliation are compliance or late-stage mitigations, not preventive fixes.
Why not A: Tempting because centralizing reconciliation can quickly produce reconciled reports and may be cheaper short term. But it is a late-stage workaround that leaves upstream systems inconsistent, does not prevent new duplicates, and shifts complexity into the warehouse rather than correcting source data.
Why not B: Tempting because missing taxpayer IDs and PII handling are clear regulatory risks; improving protections and remediation reduces compliance exposure. However, these measures address confidentiality and regulatory remediation, not the lack of a single authoritative customer identifier or cross-system inconsistencies that cause duplicates in reports.
Why not D: Tempting because archival and cleanup can reduce visible duplicates and storage costs. However, retention policies do not resolve inconsistent core fields or missing identifiers in source systems and therefore do not reliably improve the underlying data quality used for reporting.
Common questions
Why isn't a data warehouse enough to fix duplicate customer records?
Apex's read-only warehouse can reconcile reporting copies but cannot correct operational records. MDM governs customer identity and distributes authoritative updates to the applications.
Does a golden customer record require one system to own every field?
No. Apex can use CRM as the authority for contact addresses and billing as the authority for verified tax status. Survivorship rules determine which values populate the golden record.
Should MDM replace legacy customer IDs after a merger?
Not necessarily. Apex can retain legacy IDs and map them to an enterprise ID, preserving historical transaction links without replacing the source applications.
Watch it solved
A full CPA ISC task-based simulation on Data management, worked step by step.
Data Quality Defects & ETL Reconciliation | CPA ISC Simulation (TBS) on YouTube
Practice ISC like the real exam
The free ChatCPA simulator: real exam layout, timed testlets, starting with a question on this topic. No account needed to start.
Open the free simulator →More on Data management
Questions from the ChatCPA bank of 17,658 CPA exam questions, each with a written reason for every wrong answer. ChatCPA is built by Nicholas Miller, CPA (Oregon #14907). How these pages are made. Spot an error? Tell us.