Redesigning master data management framework to save up to $98 million in procurement spend

Research shows that enterprises know robust master data management (MDM) can help them meet key business challenges. Yet there's a disconnect. MDM is one of the least mature procurement functions,¹ even though bad data affects all kinds of operations and decisions. When it's held in disparate systems, it doesn't offer a single version of the truth—and because there's a lot of manual effort involved, it's costly to maintain. Without a structured approach, firms miss opportunities for timely analysis, stronger growth, and accelerated mergers and acquisitions.

This multinational pharma was feeling those ill effects. It turned to us because it wanted to improve spend visibility, eliminate inefficient purchasing practices and have more effective supplier risk assessments. And it knew it needed to address its master data. When we came on board, it had categorized just 27% of its direct and indirect spend and it had low-quality vendor data, putting it in a weak negotiating position when it came to purchasing. In this master data management case study we discuss about helping the company design a master data management business case road map that would lead it to best-in-class MDM processes, governance, and technology.

Many functions played a role in this firm's MDM: IT, procurement, finance, and the data management team were among them. But they were all working in their own silos, and that had to change if the company wanted to make better purchases by getting greater visibility into procurement. The enterprise also wanted to improve vendor risk management and regulatory compliance by introducing third-party oversight, but its vendor data model wasn't in any shape to accommodate that.

Working with the pharma, we capitalized on our experience managing vendor and indirect material master data and taxonomy design. We began by analyzing the existing:

• Data quality
• Data governance
• Target operating model
• Vendor data model
• Indirect material and taxonomy
• Technology
• The vendor process and KPIs

The assessment revealed:

• Incomplete, often inaccurate and frequently duplicated vendor data
• Data-quality metrics that didn't coordinate with business needs
• Missing or incomplete data fields that affected supplier risk management
• Over 100 systems that held vendor data, with no integration between them
• No one had stewardship of MDM
• At 45-55 days, the vendor-onboarding process was too long
• The spend management tool categorized only 10% of indirect materials. Globally, the company made 93% of indirect material purchases through free-text entries rather than catalogs. That represented 65% of the organization's total spend.
• Multiple taxonomies and mismatching commodity codes. No one used commodity codes in 35% of global purchase orders.

We responded to these issues by recommending a comprehensive master data management framework—something that can seem daunting if you think of the process as limitless. So we were careful to align the road map to areas that would generate the greatest business impact.

Developing a powerful vendor data model

To create a robust vendor data model that met the company's needs and that positioned it for third-party oversight, we identified:

• New attributes to improve supplier risk assessment, including site-level details and parent-child hierarchies
• Attributes for better vendor matching, as well as tax codes, tax identification numbers, and banking details
• Redundant or unused attributes for removal
• Principles for determining whether attributes should reside in transactional or centralized master data systems
• A road map of short- and long-term attribute changes.

Reshaping data quality and governance

We scrutinized the supplier data for completeness, accuracy, and uniqueness to benchmark the quality of the firm's vendor master data systems. Then we analyzed related business rules to measure the effectiveness of vendor master matching.

Finally, IT redesigned data governance and stewardship, giving people defined roles and responsibilities. It also proposed key metrics for the data quality dashboard.

Since 93% of indirect material sourcing came through free-form text, the risks of incorrect categorization and inflated inventory were high. As well, the existing system categorized only 10% of indirect materials. So we proposed the following:

• Options for either maintaining materials information in a detailed taxonomy with many commodity codes, or managing it in a master
• The pros and cons of developing a granular taxonomy or maintaining a material master
• Guidance on when a category should reside in a material master or a catalog
• Opportunities for creating categories in catalogs to reduce free-form text purchases
• High-level taxonomy analysis and best practices for taxonomy revision

We also proposed technologies and guiding principles to support the initiatives, and shared how similar companies deal with master data.

Our recommendations identified ways for the pharma to generate benefits of $70–98 million. The vice president of strategy and performance praised our ability to work with teams from many functions to deliver insights and a strong master data management framework. We continue to support the company as the new target operating model and stewardship structure come on line. With effective master data management, the company can now expect world-class procurement.