How do you keep your MDM processes and standards up to date and consistent?

The standards and processes used within an MDM program are typically defined through the data governance policies of an organization. These policies would establish the 'rules' used within an MDM program for things such as quality standards, data definitions, and hierarchy structures. To ensure these standards are current, your governance processes must also incorporate some idea of two things: 1) ongoing audits of your policies to ensure they remain aligned to your business needs over time 2) the ability to add new policies as business needs change. This is why so many people consider governance a change management function - since it needs to continually adapt to changing needs of the business.

Typically, we start with a Technical Health Check for our TIBCO EBX (MDM) Customers. That consists of Software assessment, including versioning. Architectural assessment and Usage assessment to ensure that the best practices are followed and implemented accordingly. Generically speaking, the process/ techniques and products' ability to address the requirement has changed considerably over the past decade and more significantly in the past few years. Hence, revisiting the business objectives and aligning them with the implemented application is critical. Two main outcomes of this phase are: 1. Is your current MDM solution the best in class addressing your requirements? 2. Have you adopted the MDM solution to its fullest?

One thing I’ve found helpful is always communicating the roadmap, value added and efforts carried out by data governance team to all relevant stakeholders. Being as mdm and data governance can easily become a large scope with a lot of moving parts. This ensures mutual understanding between stakeholders, gives a sense of direction and avoids duplicated efforts.

Assess Data Governance: Document data scope, roles, and existing policies. Identify any gaps or outdated practices. Evaluate Data Quality: Check data accuracy, completeness, consistency, and timeliness. Address inaccuracies and missing info. Review Data Integration: Map data flows and synchronization. Fix any integration issues. Analyze Data Usage: Ensure data is accessible and used effectively. Collect user feedback on data quality. Identify Gaps and Risks: Perform gap analysis, assess risks, and create an action plan to address them. Continuous Improvement: Monitor metrics, provide ongoing training, and update processes regularly.

As a seasoned expert in the field, I can't stress enough the importance of defining clear and measurable goals for your MDM strategy. It's like setting up a game plan in sports - you need to know what winning looks like! Aligning MDM objectives with business goals is a slam dunk for driving value. And remember, regular data audits and clear governance roles aren't just good practice, they're your MVPs for keeping data quality at its peak! ?? #DataGoals #MDMWinning

Post assessment Define key performance indicators (KPIs) and metrics to measure the effectiveness of MDM processes and standards. Regularly monitor these metrics to identify areas for improvement and track progress over time. Schedule regular reviews of MDM processes and standards to identify areas for improvement or updates. This could be done quarterly, biannually, or annually, depending on the pace of change within your organization and industry.

Setting of achievable short term and long tern goals at the start of the program is very important to better track the ROI for your MDM implementation. Without them, organizations can get lost and MDM just becomes another system to maintain. Once initial set of goals have been achieved then new areas of improvements can be determined.

Align with Business Objectives: Ensure your MDM strategy supports goals like boosting customer satisfaction, cutting costs, or driving innovation. Set Clear Metrics: Define measurable indicators such as data accuracy, completeness, timeliness, consistency, and compliance. Plan Data Audits and Reviews: Determine the scope and frequency of audits and reviews to ensure ongoing data quality. Define Roles: Clarify the roles and responsibilities of your data governance team to support your MDM goals.

While speaking to customers one of the things I found useful is to draw the holistic approach and emphasis on the fact that MDM is not implemented just by using a great product. That's a great start but Master Data Management implementation is a combination of the right people right entitlement right governance right goals right integration with various sources or targets. The other thing is whenever possible always start with a solid reference data manage before starting to master any other technical or business domain such as Customers, Products, Locations etc… Next is data discovery and analysis. Master the process of identifying the present stage and then enrich. Use governance and business workflows with right set of entitlement.

Part of the implementation of the MDM solution should include the current deployment and backlog for improvements. While most think of the MDM solutions as just technology most of the burden on the business is the implementation of governance, structure and process that often was not present prior. In order for the MDM solution to support keeping the standards up to date there needs to be a deployment of the process for keeping improvements and enhancements possible in the future and deploying a run and manage/operational model to support continuous improvements.

Before diving into setting up your MDM solution, make sure you know what you want to achieve with it. Take a good look at your existing data and figure out what needs fixing. Get everyone on board - from the project owners to the folks who'll actually be using it day-to-day. Finally, Pick a tech solution that fits your needs.

Select an MDM Platform: Choose a solution that fits your data architecture, volume, complexity, and security needs. Design Data Processes: Implement integration and standardization processes like data cleansing, validation, enrichment, matching, merging, and deduplication. Train and Empower Teams: Provide training to data stewards and users to ensure they manage and utilize master data effectively.

Few things to monitor/notice in MDM project are: 1) Changes to Master Data: Usually all the MDM tools provide OOTB features for managing history, source lineage, merge history, etc. 2) Data Access: It is smart to keep track of all the people who are logging in and have access to MDM system. 3) Data Volume: Usually as the time passes, certain implementations can generate large amount of metadata. For example, history tables might accumulate huge volume of data after 5-6 yrs of a program. So a cleanup might be needed to handle it. 4) Data Quality: Try to run periodic data profiles to make sure the data is of good quality since bad data can adversely effect the working/usage of MDM system.

Upon implementation of MDM solution, it needs monitoring and maintenance, especially the supporting elements like Data Quality and Data Stewardship process. A dashboard specifically for Data Quality would help monitor the metrics related to it. Whatever be the MDM architecture, Data Integration is another supporting process that needs monitoring and maintenance. Most of the times, MDM is an initiative from Data Governance program, that data management policies and standards will need attention too.

Track Metrics: Collect and analyze data quality and MDM metrics, including errors, inconsistencies, and anomalies. Resolve Issues: Address data quality issues promptly to ensure accuracy and consistency. Update Policies: Regularly review and revise data governance policies, procedures, and integration rules. Engage Stakeholders: Communicate with data stakeholders and users to incorporate their feedback and address their needs.

AI has been disruptive but has also positively impacted the world of MDM. There are several areas where AI can augment MDM. 1. From data discovery, AI's Data Classification and Taxonomy Management can provide metadata mapping. 2. AI algorithms can automate the process of matching and linking data records across source systems, improving the accuracy and efficiency of MDM. 3. AI techniques such as NLP and ML can be applied to resolve entity relationships within master data, such as customer matching or product hierarchies. 3. AI algorithms can automatically enrich master data by integrating external sources.

To success and maintain the success of MDM , we should manage and establish it as a practice not a tool or a technology adopted. MDM needs to have clear goals, ownership, processes , standards, and authority , it simply requires data governance in place. Without a proper governance , operational systems will not commit or comply with MDM practices and processes, and we may end up having master data duplicate and unconsolidated again and again.

In my experience, it needs to be focused on how efficiently the merging of records are done. It involves to see whether the surviving records are not missing out other non-key but critical values. After identifying the matching records, there should be some kind of merging strategy or merging standards established as part of MDM process. Changes of stewards and custodians should also be maintained to support the stewardship process.