Data and Analytics / White Paper
4 MINS READ
Many companies are yet to address the underlying quality of the data they use.
Data makes the digital world we live in go round, empowering consumers while enabling companies to offer the anytime, anywhere services customers have come to expect today. Think of how access to real-time data via the many online retail platforms allows consumers to check product availability, compare prices, or hunt for the best deals on their phone without going anywhere. Or how they make their purchasing decisions be it for grocery items or the latest gadgets based on product ratings on these sites and social opinions. If data is helping consumers make smarter decisions, it’s also enabling companies to get closer to them. Companies are using data to gain a deeper understanding of customers, their preferences and purchase patterns; for micro-segmentation of users; and to find new ways of enhancing customer experience. Airline companies for instance, are using real-time data captured at kiosks in airports to offer customized services like instant upgrades based on customer profiles.
While companies are leveraging data and technologies such as AI to transform their business models and drive growth, many are yet to address the underlying quality of the data itself. As a result, much time and effort are spent on manual corrections, data reconciliation, and improving data accuracy. Inconsistent and inaccurate data impedes business innovation and transformation. It takes a while for transformative ideas to be accepted and adopted across industries. The speed at which these innovations can be adopted to meet emerging market needs depends on the supporting technologies that are, in turn, completely dependent on the underlying quality of data.
MDM serves as the cornerstone of successful business transformation.
As companies increasingly embrace cloud, AI/ML, data virtualization, and advanced analytics to drive business transformation, it becomes even more critical for solutions powered by these new technologies to be integrated to core enterprise master data assets. This is where MDM comes in. It can play a strategic role in driving business outcomes for enterprises.
To understand how, let’s look at a leading engine manufacturer. It decided to adopt an innovative, IoT-based method to analyze large sets of engine-generated sensor data. By analyzing sensor data, the manufacturer wanted to gauge engine performance over time and predict whether the engine required maintenance. However, the company didn’t have a 360-degree view of the engine. It also lacked an empirical context for detailed engine specification, customer contract for which the engine was utilized, and airplanes which could potentially use the engine. The unavailability of context-sensitive ‘master’ information and the inability to map the enormous sets of sensor data to such a master version of truth proved to be the missing link to getting maximum mileage from the project.
In another example, a leading global retailer was launching a new marketing campaign to analyze customer sentiment for specific products. However, the retailer faced two major obstacles. First, it didn’t have specific details on each customer. Second, in cases where customer feedback was gathered, it was unable to link the customer sentiment to the respective product and product category. These were challenges that could have been easily overcome if a master data of customers and products was in place and data from here could have been easily mapped to the new findings. MDM serves as a fundamental component and key driver for successful business transformation. As shown in figure 1, it serves as a trusted source of high-quality data for ERP transformation, IoT-led predictive maintenance, mass personalization, and more.
Figure 1: MDM: A trusted source for enterprise data
Mastering data management
MDM plays an indispensable role in transforming business models, products, and services to reimagine customer experience.
Take, for instance, a leading global retailer with a huge brick-and-mortar presence and rapidly increasing web presence that wants to ‘know’ its customers. Among other things, the retailer wants to gain a holistic understanding of the customer—who the customer is, whether the person prefers online shopping, a particular store, or product; and what the person’s average monthly spend is. It wants to use this information to launch a personalized campaign, offering products to the customer based on past buying behavior, regardless of which channel the person uses.
The data that holds the information the company is looking for is distributed across multiple systems. The retailer can leverage a highly scalable environment similar to a data lake, to tap terabytes of data for actionable insights. As shown in figure 2, it can implement match and merge principles to resolve the challenge of similar customer names, link customers to one or more locations, and analyze purchase history, likes and dislikes, and spending habits. The retailer can also gather customer feedback on social media.
Figure 2: Snapshot of a customer-specific retail campaign where MDM is used to resolve challenges.
The key to mastering enterprise data management lies in understanding where the data resides within and outside the enterprise. Often, the data is embedded within the enterprise’s critical operational systems and continuously passed back and forth as part of operational processes. Companies looking to master data management need to ask themselves some key questions:
A holistic framework
Companies need to develop a framework that focuses on people, processes, and technologies.
Most functions and processes like manufacturing, inventory management, digital marketing, human resource management, and finance have software applications and systems with designated owners and supporting teams. Others have entire departments managing applications and systems, and particularly focused on specific domains.
When it comes to MDM, typically, there are no designated owners and the data managed or that needs to be managed is scattered across applications and systems in the organization. For instance, data around an item that is manufactured and sold originates in several sources and systems in the organization and often continues to be maintained in disparate, siloed systems. It’s no surprise then that the same item is referred to differently across systems and ends up with multiple versions of data. The result: rampant duplication and inaccuracy of data, and high risk of marketing campaigns built on such data not reaching the right customers.
Companies need to approach MDM holistically and develop a framework that addresses these concerns at the organizational level. One that focuses on people, processes, and technology. Here are a few recommended actions for mid to large-sized organizations that have an abundance of data and are looking to drive effective business transformation using the data: