Big Ideas

Master Minded

Business leaders need clean master data to build information-driven businesses.

by Alison Weiss

August 2014

These days, data is a company’s most valuable source of insight for executives looking to expand market share, enhance customer service, and improve decision-making. But modern data management poses huge challenges: a study by IDC reveals that the digital universe is expected to expand to 40 trillion gigabytes by 2020. With such dramatic growth predicted, business leaders need ways to manage and operationalize external and internal data across the enterprise.

Yet many don’t have such strategies in place. According to the “Accenture Technology Vision 2014” report, it’s difficult for IT leaders to eliminate data silos in order to access, share, and analyze existing master data (the critical business information supporting the transactional and analytical operations of the enterprise).


Paul Daugherty, chief technology officer at Accenture, suggests that managers need to consider ways to unlock data silos so that information can flow across the enterprise as a supply chain. “There’s no finish line when it comes to data management. You always have new sources of data coming in,” he says. “The goal is to get faster access to data to gain insights and value out of nonstop digital information.”

But it can take some convincing to get executives onboard with data management initiatives. Support is much more likely when master data management (MDM) can be directly linked to improving decision-making or enhancing customer experience. Even when the business case for data management is clear, it can be far more difficult to determine where to start.

According to Martin Boyd, senior director of product strategy at Oracle, there is no magic starting point, because data itself never stops. Every company begins at a different place, thinking broadly about pain points in current data management practices. Have bad financial decisions been made based on reports with inaccurate data? Is the customer database riddled with duplicate data?

“Master data management is a strategy to make sure you have high-quality, authoritative data,” he says. “There are many ways to build a full master data management capability, including data quality, hubs, and governance, but they all start with a data quality assessment to determine where the biggest and most expensive data problems are so they can be tackled first.”

Championing Data Quality

Indeed, concerns about data quality are what convinced management at Beckman Coulter to improve MDM. The medical device manufacturer had made multiple acquisitions (and was itself recently acquired by Danaher), resulting in approximately 15 locations working on different back-end systems. Several years ago, leaders initiated a multiyear program to migrate all the different operations to a single global instance of Oracle E-Business Suite.

“This generated major benefits in data visibility, streamlined processes, financial reporting, and costs,” says Carl Sandvik, senior manager of information technology at Beckman Coulter. “But every time we acquired a company, we spent a great deal of effort to clean up data from the different organizations before bringing it into the system.”

Once you get the right data quality and the right solution, everybody wonders why you didn’t do it 10 years earlier.”–Linda Martino, Vice President of IT, SCEA PlayStation Business Unit

This was the case when Beckman Coulter made its largest acquisition, in 2009—a division of a Japanese manufacturer with three global systems. Sandvik and his team faced integrating all customer data and standardizing it into Beckman Coulter’s enterprise resource planning (ERP) system—in less than 12 months. “We couldn’t do anything until we had the customer data from external sources cleaned, standardized, and in our global Oracle system to be able to start testing it and transacting on it,” says Sandvik. “We realized that we needed to figure out how to shorten the process.”

Working with Kaygen, an Oracle Platinum Partner in the Oracle PartnerNetwork, the Beckman Coulter IT team addressed acquisition data issues by implementing Oracle Enterprise Data Quality solutions for ERP. The data quality platform reviews legacy customer source data and then standardizes, scrubs it, and removes any duplicate data as it moves into Oracle E-Business Suite. It took just four weeks to technically connect Oracle Enterprise Data Quality with the ERP system.

Sandvik reports that one challenging part of the project was getting the buy-in from partners on the business side of the organization. An understanding of the importance and impact of master data, in a lifecycle approach, was not consistently understood across the organization in all regions.

To start, the IT team needed to collaborate with business partners to define and identify master data, outline impacts and current problems, and subsequently create standards for the system’s master item records. Then, when it came to integrations or system conversions, legacy customer records contained fields that might map directly into Oracle—however, many would not. Additionally, time was required to determine what regulatory data would need to be verified, validated, and updated.

As an example, if English or a local language were selected for a name field globally, the IT team needed to consider the impact on customer names or item classifications when reporting the information in the global reporting system. They wanted to avoid a situation where an executive pulling a global report would see customer names in a local language they couldn’t understand.

“When it came to data integrations, the importance of high-quality data is obvious to most executives,” Sandvik observes. “Our team strives to find opportunities to incorporate MDM advancements into projects that have strong executive sponsorship. This tends to expose and build the MDM business case as a fundamental ‘system’ infrastructure to leverage and provide continuous improvement to business processes.”

The IT team brought in people they targeted as super users to train them on enterprise data quality (EDQ) basics, such as running queries and creating reports in real time. These super users previously helped to define the processes to create the master item records and would own the process and governance going forward. Now, established EDQ rules and processes could be used and reused—improving data quality and reducing the time required to migrate future acquisitions to Oracle E-Business Suite.

Today, Sandvik is investigating ways to further leverage Oracle Enterprise Data Quality and is considering investing in data hub technology with Oracle Customer Hub. He has also cultivated executive champions at the company who are educating other critical decision-makers on how MDM governance will directly lead to better and more-accurate tactical, strategic, and historical reporting—and ultimately better decision-making.

“We started out needing to resolve a tactical data quality problem, and only now are we focused on developing a more complete master data management roadmap,” Sandvik says. “It’s an iterative process. There are always going to be new acquisitions or system implementation with new datasources and changes to the business process that will involve updating and managing data.”

Oracle’s Boyd believes that addressing a concrete data quality issue is an excellent first step toward enterprisewide MDM. “Without that you will have a shaky foundation,” he says. “Beckman Coulter may not be sure yet of the big picture, but they’ve used data quality to get started and are now able to organically and fluidly expand data management governance standards and strategies.”

Cracking the Data Management Code

For a customer-centered organization like Sony Computer Entertainment America (SCEA) PlayStation Business Unit, enhancing customer experience is the main business objective behind an MDM strategy. The company serves more than 95 million PlayStation gaming customers—and the accuracy and completeness of customer records was a huge challenge.

According to Sree Vaidyanathan, senior director of business applications, IT, at SCEA PlayStation Business Unit, customers often create imaginative gaming handles when they sign up for the popular PlayStation online gaming network. These game handles can also appear on product warranty cards. “If a person wants to call herself Mickey Mouse, she has that right,” he says. “But, if a customer called in for customer service, started playing other games, or bought other Sony products, we used to have trouble associating all the activity with the correct game handle.”

40 trillion

Number of gigabytes IDC predicts the digital universe will include by 2020


And although SCEA PlayStation Business Unit gathered a huge quantity of customer data, much of it was siloed in different systems. Vaidyanathan’s team needed to find a foundational and verifiable data element to create a complete customer profile because they couldn’t rely on customer name. This turned out to be the game console serial number, because whatever customers called themselves, they typically played on the same console—making it possible to associate game handles with specific consoles. Vaidyanathan then used the Oracle Customer Hub customer data integration solution to centralize information from all the systems.

Integrating Oracle Customer Hub with the SCEA PlayStation Business Unit’s consumer service and campaign management systems took eight months, with much effort spent cleaning up customer name data. Today, a customer service center representative can access a complete customer record with the correct name or game handle, phone number, and console type. The record even shows what games the customer plays and when.

“Our average call time has gone from 12 minutes to 10 minutes,” says Linda Martino, vice president of IT at SCEA PlayStation Business Unit. “We can also offer specific promotions if we know a customer likes a specific type of game.”

Unstructured Future

In the future, Martino and Vaidyanathan are looking to manage the huge proliferation of external unstructured data available today. If customers use their gaming handles and are doing Snapchats of game screenshots or writing game reviews on Facebook, currently it’s very challenging for SCEA PlayStation to tie those things together.

“We’re working on prototypes to take customer problem tickets or inquiries through Twitter into our help desk process,” Martino says. “Social media offers very rich data if we are able to crack the code.”

Martino is committed to MDM to help her company become an information-driven business. And now, other Sony divisions are recognizing the potential benefits across the entire company. Customers can be tracked not only as consumers of games but also of movies, music, cameras, televisions, and any other products Sony offers.

Martino concedes that MDM’s intangibility can make it more difficult to justify up front. “Once you get the right data quality and the right solution, everybody wonders why you didn’t do it 10 years earlier,” she says.

Action Items
  • Master Data Management Glossary
  • Harvard Business Review Report
  • Photography by Shutterstock