Key Steps in the Maintenance Process of Master Data

Maintaining master data within an organization is a complex mix of auditing, cleansing, integration, validation, monitoring and retirement that happens near simultaneously once the organization has a mature master data management process in place. However, you can begin maintenance tasks by following these key steps.

1. Identify and Collect Master Data Sources

Work with all company departments to identify all sources of master data. A thorough search at this stage will prevent data discrepancies created by siloed or otherwise inaccessible data buried within departmental sources.

2. Validate and Standardize Data Inputs

The second step requires the team to analyze data inputs to verify their validity. At this stage, check that data inputs are up-to-date and that they are actually used by anyone in the company. Further, standardize data inputs by deciding what data entities and records teams need in the course of business and what format those records should take. For example, standardization may combine two fields of [firstname] [lastname] into a single field [lastname, firstname]. These decisions should be based on business needs rather than stakeholder personal preference, as they will populate data throughout the company.

3. Enrich and Consolidate Data Records

Once you validate data inputs and standardize formats, begin the process of consolidating and enriching records by combining, deduplicating and cleansing the data and augmenting any missing records with data from third-party verification systems. Using automated systems to perform data matching and de-duplication, and AI-supplemented software to perform more complex functions like fuzzy matching and record augmentation will add speed and accuracy to this process.

4. Integrate Data Across Systems and Applications

Master data maintenance is a two-way street: you pull data from source systems to create golden records, then you populate those records back into source systems and across company applications. Re-populating all company systems with the golden records is important because it:

• Ensures everyone is using the same set of records
• Improves analytic consistency
• Provides a visible template for entering new records

5. Review and Approve Data Changes Through Governance

Once golden records have populated throughout the company software, it’s time to lay down the (data) law by implementing master data governance across the company. Changes will need to be made to data formatting, source connections or technology, but each of these changes should follow a defined (and automated) approval process. When the process is defined, it promotes uniformity for all decisions, keeping the full scope of data use in mind and not making allowances for individual departments or employees that might affect overall data reliability.

By automating master data management approval processes, each decision can take hours or days rather than weeks or months, keeping data moving at a speed that helps the business instead of slowing it down.

6. Synchronize and Update Data in Real Time

Data updates from source software should move just as fast as approvals. Real-time data updates that populate across the company source software quickly ensure that everyone makes decisions according to the latest data. A robust MDM software with automatic sync that works according to the company’s governance rules means the data team can focus on improvements to the overall system, rather than manually syncing and updating data to the latest versions.

7. Audit, Monitor, and Refine Data Quality Regularly

After initial processing, this step becomes the daily work for data professionals. Audit data by performing manual and automated checks on data sets to ensure they fit governance rules and provide the needed insights for business units. Monitor processes to keep APIs connected, source systems exchanging data with the MDM software and check with business units to double-check they follow the latest guidance for data entry. As you perform audits and checks, note any inconsistencies and issues against the KPIs you set for MDM measurement and showing value of the project. Then meet again with the data quality stakeholders to formulate new governance rules and data processes to address those issues and restore the high level of data quality you need.

5 Master Data Maintenance Best Practices

Maintaining scalable master data quality in the long term is a big job. These actionable master data management best practices will help you avoid common pitfalls and prepare the team to tackle the complex and changing landscape of business master data. Use these best practices alongside Profisee’s 6 pillars for MDM success.

1. Prioritize High-Value Data Domains First

It’s tempting to want to fix everything all at once, but start small and scale once your team has figured out what works. Begin with the domains most closely tied to revenue, compliance or customer experience as these will have the biggest impacts on revenue or business operations. Look especially to build master data golden records for domains that do not change often, like vendor master file and supplier information management systems. Once you prove the success of your master data maintenance campaign you can scale to other domains.

2. Balance Centralized Standards With Local Flexibility

It’s tempting to try to force all business units into uniform standards across the board to make things simple. But business is complex, and that complexity multiplies as companies expand to different regions. The data team must learn to balance global rules with regional adaptations needed for clarity, hierarchy or compliance purposes. A bit of local flexibility will lead to wider compliance with governance needs.

3. Build a Data Stewardship Network Across the Business

Data stewardship is everyone’s responsibility. Work with all departments to train employees on new processes and identify data quality maintenance needs as they arise. Finance, marketing, revenue, sales, manufacturing and operations — they all have a stake in quality data. In the retail industry alone, there was an estimated $1.77 trillion difference between expected and actual stock levels, differences that reliable and usable product data could offset. When you go beyond IT and bring other departments into the conversation, you can better keep data aligned with the actual business use cases.

4. Automate Routine Maintenance but Keep Human Oversight

We’re not ready to give everything up to the machines yet. Leverage automated and AI tools for deduplication, enrichment, and monitoring, but assign humans to resolve exceptions and disputes. An exception resolution team of departmental stakeholders and business data users will help the data team align with business needs and build governance that works to improve company operations.

5. Track and Share Data Quality Metrics Transparently

Finally, make the data quality project’s progress visible to the company. Report accuracy, completeness, and timeliness scores so leadership can see ROI and teams stay accountable. It may feel repetitive to the data team because they live in the data every day, but consistent repetition and updates will help other business units remember the importance of compliance and reinforce the value of high-quality and usable data.

Common Challenges of Master Data Maintenance

Master data maintenance projects bring inherent complexity due to the need to build company-wide consensus. In addition to the complexity of the initial master data project, challenges like silos, lack of standards, duplicates and more can affect progress. The table below details the impact of these challenges and how people, process, and technology solutions resolve them.