What is master data management: PwC

Summary

Master data management provides a single source of truth for your company’s data — ensuring uniformity, accuracy and stewardship across the enterprise.
Modern MDM solutions can integrate machine learning and streamline data handling — helping to reduce errors and enhance data security and compliance.
Streamlined collaboration and governance, graph-based search and data exploration, and machine learning for data discovery and stewardship are three key ways to drive better business outcomes via MDM adoption.

For a number of years, companies have invested in technology, tools, processes and people to handle master data management (MDM), which aims to enable the accuracy of master data and represents a single view of truth and is coordinated across the business.

In today’s landscape of mergers and acquisitions and ballooning regulatory complexity, introducing MDM (also called data mastering) is often easier said than done. The rise of cloud-based applications and services along with the high cost of legacy MDM solutions is fueling an evolution in the category. Legacy MDM solutions — even some that are cloud hosted but not cloud native — require an unsustainable level of manual reconciliation and rely on systems that traditionally haven’t been easy to connect or manage.

Moving into a new era of MDM

Cloud, big data and complementary technologies such as artificial intelligence (AI), machine learning (ML) and automation are converging in solutions that make data mastering faster and more affordable, even as data volumes explode. These developments have incentivized enterprise MDM vendors to modernize their stacks for better interplay and an improved user experience (UX). Modern MDM solutions, in particular, offer these benefits and can integrate ML to streamline data handling, help reduce errors and enhance data security and compliance.

Here are three ways MDM is evolving to help your company achieve better business outcomes to see how leading companies and early adopters are putting them to work.

1. Streamlined collaboration and governance

With more business functions such as product and sales playing an active and direct role in managing master data, MDM transitioned from supporting only the back office to the front office too. This has thrown a spotlight on the need for MDM solutions to provide more evolved UX and collaboration capabilities so people who are on the frontlines in sales or product development can explore and manage the data more easily and work together to launch products faster or improve the customer experience. MDM solutions that offer contextual commenting or voting, guided data stewardship, chat integration and conversational user experiences can help make this possible.

How a global consumer products company enhanced overall UX workflow

A global consumer products company implemented a global MDM solution with the latest governance capabilities to enable its data stewards across the globe to seamlessly manage key product master data, with minimal effort.

The master data transformation initiative, led by the CTO, enhanced the overall UX for workflow based routing and notifications to be business friendly, supported by data validation capabilities that referenced common business rules to confirm master data readiness. Once the requests were reviewed and approved, the active data was replicated and syndicated across multiple core ERP transactional systems and PIM (product info management) catalogs.

The result: Significant productivity growth across the global workforce via collaboration on the central data governance platform, while also realizing high data accuracy levels across different business units.

2. Graph-based search and data exploration

A graph-based, cloud-native MDM solution supports visual navigation of relationships between customers, products, contacts and other master and reference data. In particular, companies experiencing high-volume growth that need to scale and reduce latency can benefit from the scalability of a cloud-native MDM platform. Use cases where a 360-degree view of customers (profile, relationships, interactions, preferences) may need to be served up to resource-intensive applications and/or customer data platforms in real time should also consider graph-powered solutions with an API first architecture.

Customer preferences from various sources, such as websites and apps, or insights derived from telemetry from connected devices and the internet of things, can be stitched together with core customer data to create a 360-degree view. Legacy MDM solutions, on the other hand, often can’t scale to meet such demands.

How graph capabilities helped companies take corrective action and understand proper due diligence

PwC helped investigate fraud allegations by a major international steel manufacturer. Graph capabilities were used to stitch together various pieces of data (supply chain, material, QA approvals) and visually represent the relationships between company, clients and external parties.

The result: Enabled the team to quickly identify where the problem clusters resided and take corrective action.

Another example of putting graph capabilities to work is when a company was divesting, PwC leveraged graph based MDM technology to quickly identify which profit centers and territories were driving revenue in the carveout.

The result: Enabled proper due diligence on the deal, with massive benefits to the PE buyer.

3. Machine learning for data discovery and stewardship

Modern MDM solutions integrate machine learning (ML) to help improve employee experience and boost efficiency, especially helpful as the user base shifts to the front office.

There are three primary areas where ML-powered MDM can demonstrate its advantages over legacy systems:

Automatic discovery: Source data analysis and ingestion is one of the most effort-intensive MDM activities. ML can streamline this process in myriad ways, automating data mapping and classification and enabling discovery of data relationships through smart metadata mining and deep learning. ML-powered data protection and classification capabilities are especially helpful in finance and healthcare, which need to comply with a range of data privacy regulations. You could, for instance, automatically classify data on tens of thousands of raw materials or finished products into standard or custom taxonomies — a possible game changer for consumer goods and industrial products companies. This drastic reduction of manual labor can improve efficiency and lower operations expenditures.
Self-healing: ML can help automate the process of identifying input data structure errors and recommend corrective action. Once data has been ingested, built-in anomaly detection and automated recommendations for remediation rules can help further enhance data accuracy.
Intelligent data stewardship: ML can help reduce manual effort by automating match rule identification and providing recommendations. Self-learning algorithms with human-in-the-loop oversight can perform matching, linking and merging to reduce the volume of master data records that need to be handled manually. ML can also automatically orchestrate workflows and offer guided data stewardship to streamline the process of manual review for those records that require it. Most cloud-native MDM solutions can also reduce the effort associated with the traditional ETL approach used in legacy MDM solutions. This is especially critical as data volumes mushroom while pressure to control operational expenses grows.

How a company integrated ML to find migration savings and accurate analytic results

As part of its post-merger integration efforts, PwC helped a SaaS Identity and Access Management company leverage ML-powered entity resolution to provide accurate master data and hierarchies, in support of the data migration and cleansing effort.

The result: All data migration and validation was completed in three days, minimizing business disruption. In addition to significant migration related savings, it enabled more accurate analytic results supported by a single view of the customer.

Getting started with modern MDM

Companies with rudimentary or legacy MDM, or those that want to start with next-gen MDM, can initiate a quick assessment to identify: key business drivers; key gaps in domain coverage and data management capabilities; and a governance operating model to help define the path forward. This assessment can help companies get going with MDM, and especially those with a complex data mastering landscape due to M&As or other events.

The next step is creating an MDM strategy and execution plan. This is especially crucial for organizations that have not yet embraced MDM and are struggling to support operational applications and provide accurate data for external reporting and business planning.

Finally, MDM modernization replaces legacy on-premises (or cloud) systems with cloud-native solutions that can scale as needed to support digital value creation - everything from e-commerce to automated back-office processes to accurate predictive analytics.