Vendor duplicates and mismatched tax/bank data to MDM; ERP and e?proc get golden records

Overview

A global media company struggled with a vendor master full of duplicates and incomplete records spread across regions and business units. Procurement, Accounts Payable, and legal each maintained variations of supplier data, and updates failed to propagate consistently. Intelligex implemented a Master Data Management (MDM) hub with survivorship rules, validation services, and data stewardship workflows. New requests were screened before creation, golden records synchronized to ERP and e-procurement systems, and tax and banking details remained consistent across platforms. Duplicate creation slowed, remediation became systematic, and audits had a clear lineage of every change.

Client Profile

• Industry: Media and entertainment (broadcast, streaming, and studios)
• Company size (range): Multi-business, multi-region enterprise
• Stage: Mature financial systems with regional autonomy and legacy acquisitions
• Department owner: Procurement, Supply Chain & Logistics
• Other stakeholders: Accounts Payable, Legal and compliance, Tax, Treasury, IT applications, Regional finance, Vendor onboarding

The Challenge

Vendor creation and updates were handled locally, with each region using its own onboarding form and validation steps. Some teams prioritized speed and created suppliers on request; others waited for full documentation. The result was duplicate supplier records with subtle variations in name, address, tax identifiers, or bank details. Invoices matched to the wrong supplier, payments stalled when banking information conflicted, and contract metadata was misaligned with vendor entries.

Replacing the Enterprise Resource Planning (ERP) and e-procurement systems was not an option. The ERP remained the system of record for purchase orders, receipts, and payments; e-procurement held catalogs, contracts, and sourcing history. What the organization lacked was a shared vendor data model, consistent validations, and a governed way to publish a single trusted record to every downstream system. Compliance pressure around tax reporting and sanctions screening made ad hoc data hygiene infeasible.

Why It Was Happening

Decentralized onboarding created fragmented truth. Regional teams collected different fields, accepted different document types, and applied different quality checks. The ERP allowed creation with minimal attributes, and later updates overwrote critical fields without audit-friendly lineage. There was no reliable signal to determine which version of a vendor record should “win” when conflicts arose.

Technical constraints amplified the issue. The integration pattern was batch-based and one-way, so corrections made in one system appeared late or not at all in others. There were limited checks for duplicates at the time of creation, and search results did not surface near matches. Banking and tax details lacked standardized formats, making validation difficult across regions. Without a clear stewardship role and workflow, fixes depended on who noticed the discrepancy first.

The Solution

Intelligex deployed an MDM hub for vendor data that sat above existing systems and became the authoritative source for supplier identities and critical attributes. The hub enforced a canonical vendor model, validated key fields at onboarding, applied survivorship rules to reconcile conflicts, and published golden records back to ERP, e-procurement, and AP automation tools. A data stewardship workflow handled edge cases, with human-in-the-loop review for sensitive changes like tax IDs and banking information.

• Integrations: Bi-directional synchronization with ERP for vendor master and payment runs; connectors to e-procurement and contract repositories; optional links to sanctions screening and AP automation. Address and identity elements aligned to standards such as Global Location Number (GS1 GLN) where applicable.
• Canonical data model: A unified vendor schema covering legal and doing-business-as names, tax and VAT IDs, remit-to details, banking keys, contacts, and risk/compliance flags, with region-specific extensions.
• Matching and survivorship: Probabilistic duplicate detection on names, addresses, tax IDs, and bank keys; configurable survivorship rules that prioritized verified fields, most recent authoritative updates, and trusted sources.
• Validation services: Inline checks for VAT/TIN patterns and country formats; tax ID validation referencing services such as the EU VAT Information Exchange System (VIES); banking validations including IBAN and BIC format checks via SWIFT IBAN guidance.
• Stewardship workflows: Human-in-the-loop review for merges, deactivations, and sensitive attribute changes; dual-approval for banking and tax updates; reason codes and attachments required.
• Onboarding portal: Standardized, permissions-aware intake with dynamic forms by region and supplier type; near-match suggestions to prevent duplicate creation; document capture for tax and compliance.
• Publication and subscriptions: Golden records published to ERP and e-procurement via APIs; event-driven updates so downstream systems stayed aligned; change notifications to AP and category managers.
• Audit and lineage: Full history of attribute changes, approvals, and source provenance; merge lineages retained for de-duplicated records.
• Permissions: Role-based access for vendor onboarding, AP, tax, treasury, and regional stewards; read-only visibility for auditors and category managers.

Implementation

• Discovery: Mapped vendor onboarding processes by region and business unit; inventoried source systems and data attributes; sampled duplicates to understand matching signals; documented validation gaps for tax and banking fields.
• Design: Defined the canonical vendor schema and attribute authority by source; designed matching rules and survivorship logic; established stewardship roles and escalation paths; specified event schemas for create, update, merge, and deactivate.
• Build: Configured the MDM hub (for example, using platforms such as Informatica MDM or SAP Master Data Governance); implemented ERP and e-procurement connectors; built the onboarding portal with near-match search; integrated validation services for VAT and banking formats.
• Testing/QA: Ran deduplication in a sandbox using production extracts; validated merges with regional stewards; simulated updates from multiple sources to confirm survivorship outcomes; enforced human-in-the-loop checks for sensitive changes.
• Rollout: Launched in phases by region and vendor category; kept local ERP creation available for emergencies with post-facto sync; activated soft blocks that warned on potential duplicates before moving to hard gates.
• Training/hand-off: Scenario-based sessions for stewards, AP, and procurement; playbooks for merge and split decisions; quick guides embedded in the onboarding portal; transitioned ongoing operations to a data governance council with IT support on call.

Results

Vendor creation became predictable and governed. Near-match suggestions prevented duplicate entries at the point of onboarding, and stewardship workflows resolved edge cases without lengthy email threads. Downstream systems received a consistent golden record, so purchase orders, contracts, and invoices aligned on supplier identity and remit-to details.

Banking and tax attributes stabilized. Treasury saw fewer payment rejections due to inconsistent account details, and AP processed invoices with fewer holds tied to vendor mismatches. Tax and compliance teams reviewed complete, traceable histories of changes rather than piecing together spreadsheets. Internal audits were quicker to complete because each attribute had clear lineage, approvals, and source provenance.

What Changed for the Team

• Before: Regional teams created vendors with varying forms and checks; After: A single onboarding portal applied standard validations and surfaced near duplicates.
• Before: Duplicate detection happened after payment issues; After: Matching and survivorship prevented and resolved conflicts proactively.
• Before: Banking and tax updates overwrote records without context; After: Sensitive changes required dual approval with full audit trails.
• Before: Corrections in one system took time to appear elsewhere; After: Event-driven publication kept ERP and e-procurement synchronized.
• Before: Audits relied on email threads and file shares; After: A golden record carried attribute lineage and approvals in one place.

Key Takeaways

• Treat the vendor master as an enterprise asset with a canonical model, not a collection of local lists.
• Prevent duplicates at the source with near-match search, standardized intake, and survivorship rules.
• Validate critical attributes—tax IDs and banking—against trusted formats and services, and require human review for changes.
• Publish golden records to ERP and e-procurement via events so downstream systems stay aligned without manual reconciliation.
• Start with high-change regions or categories to prove matching and stewardship, then expand once exception handling stabilizes.

FAQ

What tools did this integrate with?
The MDM hub synchronized with the ERP for vendor master and payments, connected to the e-procurement platform for supplier records and contracts, and shared updates with AP automation and sanctions screening tools. Address and identity elements aligned to standards such as GS1 GLN where applicable, and banking formats followed guidance from SWIFT IBAN.

How did you handle quality control and governance?
We implemented stewardship workflows with human-in-the-loop approval for merges, deactivations, and sensitive changes. Matching thresholds and survivorship rules were versioned and change-controlled. Banking and tax updates required dual approval with documentation. Every action carried a reason code and full lineage, ensuring audit readiness.

How did you roll this out without disruption?
We phased deployment by region and vendor category. The hub ran in observe-and-recommend mode first, flagging likely duplicates while local processes continued. Once matching and publication proved reliable, we enabled gating to prevent duplicates and directed new onboarding through the portal. Emergency creation in ERP remained available with follow-up reconciliation.

How were duplicates detected and resolved?
The hub used probabilistic matching on a combination of signals—legal and DBA names, address elements, tax IDs, and bank keys. Potential duplicates generated steward tasks with side-by-side comparisons and confidence scores. Approved merges preserved lineage and deactivated redundant records, with updates propagated to downstream systems.

How were tax and banking details validated?
Tax identifiers were checked for format and, where supported, validated against services such as the EU’s VIES. Banking details were validated for structure using IBAN and BIC formats per SWIFT guidance. Sensitive changes triggered dual-approval workflows and required supporting documents before publication.