Master Data Management: a single version of the truth about customers, products and counterparties
When the same customer, product or counterparty is described differently in CRM, ERP and Excel, data is hard to match and reference lists have to be reconciled by hand. Progresia implements MDM in the Microsoft ecosystem: we link records from connected sources, detect duplicates, build master records and set up control over new and changed data.
What a business without master data management looks like
Master data is the core reference data every process relies on: customers, products, suppliers, warehouses, employees. When each system maintains it in its own way, problems appear in sales, finance and logistics at the same time.
One customer, three records. "Agro-Zakhid LLC", "Agrozakhid" and an entry without a company ID in a sales manager's Excel file. Sales, receivables and deal history are scattered across them.
One product has different codes with no reliable mapping. ERP, CRM, the website and suppliers may use different item numbers for the same product. Without agreed links between them, data exchange breaks. MDM builds the mappings between codes, and integrations use them to pass prices, stock levels and other operational data.
Data is hard to compare across departments. Different records for the same customer or product lead to different grouping of figures in reports. Consistent reference data creates a common basis for analysis. For results to be comparable, metric definitions, periods and refresh rules also need to be agreed.
Reference data is maintained by hand. A new counterparty is created separately in every system. Errors in bank details surface only when a payment bounces.
AI receives inconsistent data. Duplicates, incomplete attributes and outdated records can degrade the answers of AI assistants. MDM helps prepare consistent data for Copilot and AI agents. The result also depends on which sources are connected to a given assistant and how access to them is configured.
Every new integration makes things more complex. After a merger, the launch of a new warehouse or an online store, yet another source appears with its own codes.
What is Master Data Management
Master Data Management (MDM) is the set of processes, rules and technologies for consistently maintaining key business entities: customers, products, suppliers and other reference data. MDM helps determine which records from different systems refer to the same entity and which values of its attributes should be considered the master values.
A Golden Record is an agreed master record for a specific customer, product, supplier or other business entity. Its attributes are built according to agreed rules for selecting values from sources and, where needed, taking a data steward's decision into account.
Source systems may keep their own records and local codes. MDM maintains their link to the master record. Depending on the chosen model, the data is consumed through links, APIs, analytical datasets or synchronization of agreed attributes.
MDM is more than software. It requires data owners, quality rules, a change approval process and a person who resolves disputed cases (a data steward). Progresia sets up all of these elements, not just the platform.
| Outcome | Master records for business entities, mappings of source codes and rules that keep data accurate and up to date. |
| Domains | Customers, products, suppliers, locations, employees, reference data |
| Sources | Dynamics 365, Business Central, BAS / 1C, SAP, Excel, SharePoint, website |
| Models | Registry, consolidation, coexistence or centralized. The scope of central storage and the directions of synchronization depend on the chosen model. |
| Platform | Combined architecture: Microsoft Fabric + Power Apps + Dataverse. |
| Start | Data audit and a pilot on one domain |
Which master data we put in order
We start with the domain where inconsistent data has the greatest business impact, for example customers or products. New domains are added on a shared platform. For each one we configure attributes, relationships, quality and matching rules, roles and integrations, reusing the shared mechanisms for loading, auditing and steward work.
Customers and counterparties
Unique identification of a company by its registration code (EDRPOU) and other attributes, consistent legal details, contacts and relationships between organizations. With a separate YouControl integration we add counterparty validation and enrichment; the data scope, access and terms of use of the service are agreed separately.
Products and items
A consistent product catalog with item numbers, characteristics, units of measure and categories. We map ERP, CRM, website and supplier codes and maintain their links to the MDM identifier. Additional product content management and catalog publishing features are scoped separately.
Suppliers
A single supplier master with consistent legal details, contacts and statuses. Validation rules, data enrichment and approval of new suppliers are configured to match the client's business process.
Locations, warehouses and assets
Branches, points of sale, warehouses, machinery and equipment with unified codes and addresses. Essential for distribution, agriculture and Field Service.
Employees and structure
A consistent directory of employees, departments and organizational relationships for exchange between HR, ERP and other systems. Integration with Progresia.HR, Microsoft 365 or Microsoft Entra ID, as well as the related access management processes, is defined as a separate scope of work.
Reference data and financial hierarchies
Chart of accounts, cost and profit centers, currencies, regions, sales channels. The same hierarchies for accounting and Power BI.
How MDM works: from scattered sources to a master record
New and changed records go through an agreed cycle of validation, matching and master data updates. Processing runs on a schedule, on an event or manually, depending on source capabilities and freshness requirements. Quality and error monitoring helps detect new discrepancies in time.
Example: a sales manager creates the customer "Agrozakhid" in Dynamics 365. Once the record reaches MDM, the solution checks its attributes and finds a company with the same registration code in the ERP. Under the agreed rules, the record is linked to the existing master record or passed to a data steward.
Connecting sources
We connect CRM, ERP, SQL databases, Excel, SharePoint and other agreed sources through Data Factory in Microsoft Fabric, Power Platform connectors or APIs. For each source we define record keys, field mapping, the way changes are captured and the load schedule. The load on source systems is taken into account at the design stage.
Profiling and cleansing
We check completeness, formats, code validity and potential duplicates. Names, addresses, phone numbers and units of measure are standardized according to domain rules. Enrichment from external services, including YouControl, is added within an agreed integration scope.
Matching and building the Golden Record
We match records by identifiers, normalized attributes and fuzzy matching rules. Reliable matches are processed automatically; ambiguous ones go to a data steward. Survivorship rules determine the attributes of the master record based on the source, freshness and agreed decisions. AI mechanisms are added where needed after their accuracy and cost have been verified.
Data stewardship and approval
The steward reviews ambiguous matches, compares records and approves master values in Power Apps. The app stores the decision, comment, author and timestamp in Dataverse. In the combined architecture, a Fabric process validates and applies the approved decision to the Golden Record, and the steward receives a completion status or an error. Changes and the lineage of values are recorded in the audit log.
Publishing and synchronization
We publish approved master data for analytics and connected business systems. Local source codes are kept and linked to a single MDM identifier. Attributes are passed to Dynamics 365, ERP and other systems according to agreed directions, rules and schedules. Power BI uses the master data for reporting.
MDM in the Microsoft ecosystem: architecture tailored to your needs
We define the architecture based on data volumes and change frequency, the number of domains, rule complexity, approval, performance and security requirements. We take into account your existing Microsoft environment and the total cost of ownership.
MDM on Dataverse and Power Platform
An option where master records and workflows are maintained in Dataverse. Power Apps provides the steward interface. Normalization, matching and master value selection rules are implemented as part of the MDM solution. Whether existing Dynamics 365 licenses can be used is checked for each specific scenario.
Steward workspace: record comparison, review queue, value approval and comments. In the combined architecture, Dataverse stores tasks, statuses and decisions, while Golden Records remain in Fabric.
Sources we work with most often:
Approaches to managing master data
The result depends on an agreed data model, integrations, quality rules and accountability for changes. We choose between manual maintenance, a specialized platform and an MDM solution on Microsoft, taking into account the required functionality and the cost of ownership.
| Criterion | Excel and manual maintenance | Specialized MDM platform | MDM solution on Microsoft |
|---|---|---|---|
| Master records | Created and maintained manually. | Created by the chosen product once configured. | Created by MDM logic according to agreed rules. |
| Matching and cross-references | Require manual checks. | Depend on the product's capabilities and configuration. | Use identifiers, normalization and configured matching rules. |
| CRM and ERP integrations | Manual import, export and reconciliation. | Out-of-the-box connectors or APIs with exchange configuration. | Data Factory, Power Platform or APIs with exchange configuration. |
| Steward work | Files, emails and manual approvals. | The chosen product's interface and processes. | Power Apps with tasks, statuses and decisions in Dataverse. |
| New domains | Separate files and maintenance rules. | Domain configuration within the product's capabilities. | A shared core plus separate configuration of attributes, rules and roles for each domain. |
| Enrichment and AI | Separate operations or tools. | Product options or additional integrations. | Additional services and integrations as the project requires. |
Start with a master data audit
We will analyze one domain in your systems, such as customers or products. We will show how many duplicates and incomplete records you have, where they come from and where to start with MDM.
MDM for your industry
In every industry, errors in different domains are the most costly. That is where we start.
Distribution
Consistent products, units of measure, supplier codes, points of sale and delivery addresses. Unique identification helps integrations correctly link sales, logistics and inventory data.
Manufacturing
Consistent masters for materials, finished goods, equipment and counterparties. Code mappings support exchange between production, warehouse and commercial systems; specialized production processes remain in their dedicated applications.
Agribusiness
Unique identification of counterparties, fields, machinery, warehouses and other objects for exchange between industry systems. Counterparty validation and extended integrations are added within the agreed project scope.
Car dealerships
Consistent masters for customers, vehicles with VINs and dealership units. Links between records help correctly combine sales and service information from CRM and service systems.
Telecom and media
Unique identification of subscribers, corporate clients, advertisers and agencies. Consistent attributes and cross-references are used for exchange between billing, sales and support.
Holdings and groups of companies
Consistent masters for counterparties, products, departments and financial hierarchies across legal entities. Local codes are linked to shared MDM identifiers for integrations and consolidated reporting.
How Progresia implements MDM
We do not launch a multi-year program right away. First, a pilot on one domain with a measurable result, then scaling to the others.
Master data audit
We inventory the sources, profile the data of one domain and measure duplicates and completeness. You receive a brief with a description and a work plan.
Data model and rules
We define master record attributes, source keys and mappings, matching and survivorship rules. We agree on data owners, steward roles, approval workflows, Golden Record storage and how decisions are applied.
Pilot on the first domain
We connect the agreed sources, configure normalization and matching and verify how master records are built. We launch the steward queue and test the full cycle: source record, mapping, decision, Golden Record update and publishing to consumers. The result is assessed against agreed quality metrics.
New domains and systems
We connect new reference data and sources to the shared MDM core. For each domain we configure the attribute model, relationships, quality rules, matching and access rights. We test integrations and load, reusing the shared processing and steward mechanisms.
Support and data stewardship
We support MDM processes and monitor data quality and exchange errors. Together with data owners we refine the rules and resolve ambiguous cases.
Why clients choose Progresia
- Microsoft-focused integrator: Dynamics 365, Power Platform, Power BI, Azure
- 200+ integrations, including BAS / 1C, Business Central and SAP
- Counterparty validation and enrichment via YouControl within a separate integration scope
- Full cycle: audit, architecture, implementation, training, support
- Architecture sized to your scale, with no unnecessary licenses
- Ukrainian team and documentation in Ukrainian
Ready to get your data in order?
Let's start with a short conversation about your systems and the domain that hurts the most.
Frequently asked questions about Master Data Management
What is master data and how is it different from transactional data?
Master data describes business entities and their properties: customers, products, suppliers, employees or locations. Transactional data records events and operations: orders, shipments, invoices and payments. Master data may change at different rates, but its consistency is essential for correct operations and reporting.
Do we need MDM if we already have CRM and ERP?
MDM makes sense when duplicates, conflicting attributes or significant manual reconciliation costs accumulate in CRM, ERP and other systems. It helps link records of the same entity and manage master values. During the audit we determine whether a dedicated MDM layer is needed or whether improving the existing maintenance rules and integration is enough.
How long does an MDM implementation take?
It depends on the number of sources and the state of the data. We do not sell a multi-year program: first an audit and a pilot on one domain with a measurable result. We give exact timelines and costs after the audit, once the scope and quality of the data are clear.
Do we have to buy a separate MDM platform?
A separate specialized platform is not needed in every project. An MDM solution can be built on Dataverse and Power Platform or on a combined Fabric + Power Apps + Dataverse architecture. The choice depends on volumes, change frequency, the number of domains, rules, process requirements and total cost. We consider a specialized product when its out-of-the-box capabilities better match these requirements. Existing license entitlements are checked separately.
Does MDM work with BAS and 1C?
Yes. BAS and 1C can be connected through supported standard exchanges or other agreed integration mechanisms. MDM links their records to the master data while keeping local codes. Write-back capability, the list of fields and synchronization frequency are defined for the specific configuration of the accounting system.
How is MDM related to Copilot and artificial intelligence?
MDM helps prepare consistent reference data that can be used in Copilot and AI agents. This reduces the risk of answers based on duplicate, incomplete or conflicting records. To achieve this, the assistant must be connected to the cleansed data with correctly configured access. We verify answer quality on agreed scenarios.
We use SQL Server Master Data Services. What should we do next?
Master Data Services has been removed from SQL Server 2025. For existing SQL Server 2022 and earlier installations, we take into account the support lifecycle of the respective version and the roadmap for further development.
Before migration, we inventory MDS models, business rules, history and integrations. The required capabilities are rebuilt in the target MDM solution, we verify the results and agree on a phased cutover of data consumers.
Who is responsible for data quality after go-live?
During implementation we identify data owners and stewards in your team and train them. If you have no internal resources, Progresia can provide stewardship as a service: resolving disputed matches, monitoring the quality dashboard and configuring new rules.
One version of data for sales, finance and AI
Progresia will help you assess the quality of your master data, choose the right architecture in the Microsoft ecosystem and launch a pilot on the domain with the greatest business impact.
We implement technologies known all over the world!
More than 200 integrations with various systems
