Operations & Technology

MDM (Master Data Management)

Master Data Management is the company-wide strategy and system architecture that ensures all critical master data — products, customers, suppliers, locations — is managed in a single, consistent, and authoritative source and synchronized across all connected systems.

Formula
MDM = Product + Customer + Supplier + Location

MDM is not a system you buy. It is a strategy you implement — one that spans multiple systems. A PIM is a tool for product data. MDM is the governance layer that ensures all master data across a company — products, customers, suppliers, locations — is consistent, complete, and usable across systems. When someone says MDM, they typically mean a problem that is bigger than product data alone.

Good sign

MDM makes sense when the same entity — a customer or supplier — exists in multiple systems with inconsistent data, and that inconsistency has operational or strategic consequences. Typical signal: CRM has Customer A at one address, ERP has a different one, Shopify has a third. At transaction volumes with compliance relevance, or during M&A integrations, this is not just a technical problem.

Warning sign

MDM almost always fails due to governance, not technology. Introducing an MDM system without clarifying who is responsible for which data domain — and how conflicts between systems get resolved — produces an expensive database with the same quality problems as before.

For most e-commerce companies in the DACH market below 100 million euros in revenue, MDM as a formal discipline is not an immediate topic. The relevant problem is almost always the product data domain, which a PIM solves. MDM becomes relevant when system landscapes are complex enough that inconsistencies between customer data, supplier data, and product data have measurable operational or compliance consequences.

Industry Benchmark
Growth stage (up to 10M EUR revenue) PIM is enough. MDM is overhead.
Scale-up (10–50M EUR revenue) PIM + ERP integration. MDM optional.
Mid-market (50–200M EUR revenue) Worth evaluating MDM for product domain
Enterprise (>200M EUR, multi-entity, M&A) MDM is an operational necessity
  • Define data domains first: which master data types are business-critical? Product, customer, supplier, material, location — not every domain needs an MDM solution immediately
  • Clarify data ownership: for each domain, a clear data owner must be defined who has the final say when conflicts between systems arise — without this, MDM is not implementable
  • Master record vs. golden record: which system is the 'source of truth' for which domain? ERP for commercial product data, PIM for communicative product data, CRM for customer data — this hierarchy must be explicit
  • Plan integration architecture: MDM requires bidirectional synchronization between systems — ETL processes, API integrations, or a dedicated MDM hub; the architecture determines sustainability
  • Establish data governance: who can change what in which system? Which fields are synchronized via MDM and must not be overwritten in target systems?
  • Confusing MDM with PIM: PIM solves one data domain (product communication). MDM is a company-wide governance architecture covering multiple domains — including PIM as the product domain
  • Introducing MDM too early: companies below 50 million euros in revenue with a manageable system landscape typically do not have an MDM problem. They have a PIM problem or a process problem. The wrong diagnosis leads to the wrong investment
  • Technology before governance: no MDM system resolves unclear data ownership. The system manages data — it does not decide which data is correct. That decision is human and organizational
  • Tackling all domains simultaneously: MDM projects that try to transform all master data domains at once regularly fail due to complexity. The most successful MDM implementations start with one domain and scale from there

MDM vs. PIM: Which system is right for you

Master Data Management (MDM) is the company-wide governance architecture for all critical master data. Master data are data objects that are long-lived, change slowly, and are used simultaneously by multiple systems: products, customers, suppliers, materials, locations. MDM ensures these objects are consistent, complete, and authoritative across all systems — and that when conflicts arise between systems, it is clear which source is correct.

MDM vs. PIM: The distinction that matters in practice

Confusing MDM and PIM is an expensive mistake. Here is the difference that counts operationally:

  • PIM (Product Information Management): Solves a specific problem within the product domain — how products are described, translated, and prepared for each channel. PIM is a solution for one data domain.
  • MDM (Master Data Management): Solves the overarching problem of consistency across all master data, all systems, and all domains. MDM includes PIM as one of its domains — alongside customer, supplier, and location data.
  • Practical rule of thumb: If your problem is 'product data is inconsistent across Shopify, Amazon, and the catalog', the answer is a PIM. If your problem is 'customer data in CRM, ERP, and shop system does not match, causing compliance issues or broken processes', MDM is the topic.

For e-commerce companies below 100 million euros in revenue, MDM is almost never an immediate topic. The relevant problem is the product data domain — which a PIM solves. MDM becomes relevant when multiple data domains (product, customer, supplier) simultaneously generate cross-system consistency problems.

The four MDM architecture models

Anyone serious about MDM needs to choose an architecture model. The choice affects implementation effort, flexibility, and the degree of control over master data:

  1. 1 Registry model: No central data store. An MDM system registers where which master data lives in which systems and makes them findable via unique IDs. Low implementation cost, but no real data consolidation — conflicts are made visible, not resolved.
  2. 2 Consolidation model: Data from multiple source systems is merged and cleansed in a central MDM hub. The hub is read-only and serves analytics and reporting. Source systems are not changed. Good for business intelligence, not for operational processes.
  3. 3 Coexistence model: Master data is maintained in source systems but merged, cleansed, and returned to source systems as a 'golden record' via a central hub. This is the most common model in e-commerce contexts.
  4. 4 Centralized model: All master data is maintained exclusively in the MDM hub and distributed to target systems from there. Highest data consistency, but highest implementation effort and strongest dependency on the MDM system. Typical in enterprise environments with SAP Master Data Governance.

MDM in e-commerce: when it concretely becomes relevant

Three situations where MDM moves from an academic to an operational topic in e-commerce:

  • Mergers & Acquisitions: When two companies with separate ERP systems, shop systems, and customer bases are merged, an MDM problem arises immediately. Which product number applies? Which customer record is correct? Without an MDM strategy, system integration produces data duplication and process breaks.
  • Regulatory compliance: GDPR requires companies to know where customer data is stored and to be able to delete it completely on request. Companies maintaining customer data inconsistently across CRM, ERP, shop system, and email marketing tool have not just a technical problem but a compliance risk.
  • Complex B2B supplier integration: Companies sourcing products from 50 or more suppliers, each with their own part numbers and data formats, need an MDM strategy for the product domain. Technically this is a PIM problem, but governance-wise it is an MDM problem: who decides which supplier part number is internally valid?

MDM software solutions for the DACH market

MDM software solutions differ significantly by target audience and domain focus:

  • SAP Master Data Governance (MDG): Enterprise standard in SAP-dominated environments. High integration density in SAP system landscapes, but very high implementation effort and license costs. Relevant from 500 million euros in revenue or complex SAP landscapes.
  • Informatica Intelligent Data Management Cloud: Leading enterprise MDM platform for multi-domain scenarios. Strong in data profiling, data quality, and governance. Entry from 100,000 euros annual budget, typically significantly more.
  • Syndigo / Salsify / Akeneo (product domain): When the MDM problem is primarily the product domain, specialized PIM systems are often the better choice. Akeneo has MDM-like governance features for the product domain built in.
  • Stibo Systems STEP: Mid-market MDM with a strong focus on the product domain and retail. Present in the DACH market in retail and wholesale. Entry typically from 80,000 euros annual budget.
  • In-house solution / data lake: For most e-commerce companies below 200 million euros in revenue, a pragmatic approach — central ERP as master data hub, PIM for product communication data, and clear integration processes — is economically superior to a formal MDM system.

PIM, MDM, or both — what do you actually need?

Most e-commerce companies start with the wrong system because the diagnosis is missing. In a 30-minute conversation, I will analyze your system landscape and tell you concretely whether PIM, MDM, or neither is the right next investment.

Lieber erstmal schreiben? kontakt@wernerstrauch.de