Master Data Management: Implementation and Best Practices

Semarchy and D3Clarity dive into Master Data Management (MDM) implementation and best practices to make your data work for you.

comp

Implementation Barriers

Too many MDM implementations fail because organizations spend months establishing complex industry or vendor-required data models. These result in multi-month and costly implementation cycles that end up being rejected by the end-user (the person it was trying to help all along). A traditional, waterfall approach to MDM frequently results in a solution that models how the business operated last year, not how the business needs to operate next year. Defining solid technical requirements for “how the business needs to operate next year” can be much more difficult than designing for “how the business operated last year.” It’s easy to fall into this trap. Successful teams and strategic data initiatives can alleviate this temptation. 

 

There is a better way: borrow a few agile techniques and add some software development lifecycle best practices and deployment automation.  

 

Best Practice & Approach

After you understand the strategic direction of the business and the high-level success criteria for your MDM solution, find the stakeholders or “business users” that live in the data you are mastering. Pick a specific, high-value business process or use case that is relevant to those stakeholders and begin to map how data flows through each step or decision point.

 

It's important to ask questions like:

  • What data is required?
  • Who owns it?
  • Where does it come from?
  • What happens to it along the way?
  • What happens if it is wrong?
  • What/who can change it?
  • How do you make it right?
  • Where does it go next?

With the goal of prioritizing:

Start by considering the improvements that directly increase the efficiency of these stakeholders or the quality of their output.

 

This should include:

  • Identifying each point of failure
  • Dependencies on people or tribal knowledge
  • The data sets that lack a clear owner and the sources of quality problems

In alignment with the broader strategic objectives:

Recommend a “roadmap” of experiments and improvements that address your findings AND are directly relevant to the specific stakeholders providing input.

 

As a best practice:

  • Break this roadmap into individual tasks that can be completed by a single member of your development team in one day or less (typically ¼, ½, ¾ and 1-day increments)

In collaboration with your stakeholders:

Organize one weeks’ worth of tasks to form a “sprint.” These should be tasks with the highest priority.

  • Set clear expectations with the stakeholders
  • Determine a specific date and time the following week that you will demonstrate the results of your sprint AND deploy your work to a Semarchy test environment so stakeholders can log in and experiment.

Experimentation is critical at this phase:

  • Start your week-long development sprint by being mindful of the demo and deployment
  • Stay focused and find ways to demonstrate value to your stakeholders
  • Finish each sprint with a full deployment to a testing environment

This will get your stakeholders accustomed to experimenting with the functionality and expressing what they like and don’t like on a regular cadence.

 

Impact & Results

It usually takes two or three sprints for the magic to happen. Stakeholders who originally had no available time to participate suddenly find more time. Why? Because they see tangible progress in areas that matter directly for them. At a program level, the real technical requirements start to materialize (frequently stakeholders don’t know their true requirements until they see their own data in the context of a real-world process within Semarchy). As the requirements develop, the tasks grow, but the time each task requires shrinks as patterns emerge and efficiencies are gained. The process matures and repeats project-to-project, use-case-to-use-case, business-process-to-business-process. 

 

However, it’s important to note, this doesn’t work so well if your team has to spend more than a few minutes each week deploying releases to one or more test environments or gets bogged down troubleshooting stakeholder access or configuration errors.  

 

Discover More

The recipe for success relies on the best practice guidelines above, execution of priorities, selecting the right tool for the right requirements, and leveraging the expertise of data management specialists. D3Clarity offers data management specialists, knowledge transfer, and full-service solutions for both on-prem and cloud environments.

 

D3Clarity is your partner in data strategy, data governance, data management, and cloud deployments. Success with D3 provides Clarity.

ASK US ANYTHING