Ad image

Before ERP Migration: Who Owns Product, Supplier, and Customer Data?

Disclosure: This website may contains affiliate links.

Every ERP project eventually reaches a deceptively simple step: moving existing records into the new system. Product catalogs, supplier lists, and customer accounts must all be transferred before launch. Many teams treat this as a technical task, export the spreadsheets, map the columns, and press import.

That approach almost always backfires. The new ERP inherits years of duplicates, inconsistent units, and missing details, and the problems surface at the worst moment: when orders, purchases, and invoices start flowing through the system for real.

Why a Technical Import Is Not Enough

Legacy data reflects years of workarounds. Over time, records accumulate issues that no import script can resolve on its own:

  • The same supplier entered three times with slightly different names
  • Products sold by the case in one system and by the unit in another
  • Customer accounts missing tax details or payment terms
  • Inventory balances that no longer match what sits on warehouse shelves

Successful ERP data migration depends on business decisions as much as technical execution. Someone must decide which supplier record is correct, how a product should be measured, and which fields are truly required. Those decisions belong to the people who use the data, not only to the developers who move it.

Teams specializing in ERP software development increasingly insist on assigning data owners before migration work begins. A view commonly shared by experienced software providers puts it plainly: “We can move any data you give us. What we cannot do is decide what your data should mean. That has to come from the business.”

The Example: An Industrial Components Distributor

Consider a hypothetical distributor of bearings, fasteners, and hydraulic parts. It has thousands of products, hundreds of suppliers, and a broad base of manufacturing customers. Its data lives in an old inventory tool, an accounting package, and a collection of spreadsheets maintained by different departments.

Before moving to a new ERP, the company must decide who is responsible for each type of record.

Assigning Master Data Ownership

Clear master data ownership means every record type has a department accountable for its accuracy:

  • Product data: owned by product management or operations, covering descriptions, categories, units, and specifications
  • Supplier data: owned by procurement, covering contacts, payment terms, lead times, and approved items
  • Customer data: shared between sales, which maintains contacts and pricing agreements, and finance, which manages credit limits and tax details
  • Inventory balances: owned by warehouse operations, verified with finance

Owners do not need to perform every correction themselves, but they must approve rules, resolve disputes, and sign off on readiness.

Coverage on Ars Technica has described how system migrations frequently fail because of data problems rather than software defects. The new platform may work flawlessly, yet still produce wrong results when fed unreliable records.

Common Data Issues to Resolve

Duplicates

Duplicate records create confusion in purchasing, sales, and reporting. Owners should agree on matching criteria, such as tax numbers for suppliers or manufacturer part numbers for products, then decide which version survives and how history is preserved.

Units of Measurement

Inconsistent units cause serious errors. A product bought by the box of one hundred, stored by the box, and sold individually needs clear conversion rules. Define a base unit for each product and document every conversion before migration.

Mandatory Fields

New ERP systems often require information that legacy tools never enforced, such as tax categories, payment terms, or product weights. Identify every required field early, determine who will supply missing values, and set deadlines for completion.

Inventory Balance Verification

Opening balances must match physical reality. Plan a stock count close to launch, reconcile differences with finance, and agree on a cutoff point after which transactions are recorded only in the new system.

Establishing Data Validation Rules

Clear data validation rules turn subjective judgment into consistent checks. Examples include:

  • Every product must have a base unit, category, and active status
  • Every supplier must have a unique tax identifier and payment terms
  • Every customer must have a billing address, credit limit, and tax classification
  • Prices must be positive and linked to a valid currency
  • Inventory quantities cannot be negative

Rules should be written in plain language, approved by data owners, and enforced automatically during test and final loads.

Running Test Loads

Never wait until launch to load data for the first time. Schedule several test loads in a separate environment:

  • First load: reveals mapping errors and missing fields
  • Subsequent loads: confirm that corrections have worked
  • Final rehearsal: mirrors the actual launch process and timing

After each load, owners review samples, run reports, and process realistic transactions such as creating a purchase order or shipping a customer order. Problems found here cost far less than problems found after launch.

Defining Data Acceptance Criteria

Before launch, the company should agree on objective criteria that confirm data is ready:

  • Duplicate records are resolved or documented
  • All mandatory fields are complete
  • Unit conversions are verified for every active product
  • Opening inventory matches physical counts within agreed tolerances
  • Test transactions complete correctly across purchasing, sales, and finance
  • Each data owner formally approves their area

These criteria give leadership confidence that the system will start with reliable information.

How to Choose the Right Vendor

Assess Case Studies and Relevant Experience

Ask for ERP projects where the vendor managed complex data migration. How did they handle duplicates and units? How many test loads did they run? Specific answers signal real experience.

Evaluate Tech Stack Alignment

Ask candidates:

  • Which tools they use for data mapping, cleansing, and validation
  • How migration scripts can be reused for repeated test loads
  • Whether your team could manage data updates after launch

Establish Communication Protocols

Agree on regular reviews with data owners, demos of test load results, and a shared log of data decisions and open issues.

Insist on Cost and Process Transparency

Request estimates that separate data analysis, cleansing support, test loads, and final migration. Clarify how additional cleanup work is priced and confirm access to project boards and repositories.

Plan for Maintenance After Launch

Data quality must be protected once the ERP is live. As coverage in MIT Technology Review has emphasized, organizations gain the most from their systems when data is treated as a continuously managed asset. Before signing, confirm:

  • Who supports data issues after launch, and how quickly
  • How validation rules will be maintained and updated
  • How new data requirements will be scoped and priced
  • What documentation you receive if support moves in house

Clean Data, Confident Launch

An ERP system is only as reliable as the data it receives. By assigning clear ownership, resolving duplicates and unit inconsistencies, enforcing validation rules, running repeated test loads, and agreeing on acceptance criteria, companies can launch with confidence. The effort invested before migration pays off every day the new system runs.

Share This Article
Our Editorial Staff at St. Vincent Times is a team publishing news and other articles to over 300,000 regular monthly readers in over 110 other countries worldwide.
×