Communities & Groups
Publications & Resources
Career Center

Better data means better decisions

By Mashiku Kuyi Stephen and Brian H. Kleiner

Executive summary
Managers require access to accurate data to help them make important decisions about operational efficiency, competition and regulatory compliance. But having what should be the same data available in various forms complicates matters. Past master data management (MDM) solutions were prohibitively expensive and lacked business support, but the future looks brighter. Six Sigma techniques and new low-cost and free options offer small and midsize businesses the opportunity to improve data accuracy, giving managers the information they need.

Not too long ago, paper was the most common method of communication between businesses and individuals. While paper is not the most effective way to communicate, it enabled individuals and businesses to transmit a wide range of information. Due to the nature of paper, companies who could afford computing power often had to enter the same information manually, possibly into multiple information systems. Businesses spent a considerable amount of money and time re-typing, correcting mistyped information and relaying information to business partners and affiliates. In general, using paper for information sharing is inefficient and prone to errors.

Over the last 10 years, both computing power and networking technology have improved drastically. At the same time, the cost of computers and Internet usage has dropped radically. These advancements in IT and price reductions have allowed people and businesses to take advantage of enormous opportunities. According to Forrester Research, amid a global financial crisis, U.S. online retail growth that topped $155.2 billion in 2009 is expected to reach $248.7 billion by 2014. In its studies, Forrester found that a multitude of factors such as low prices, shopping convenience and a wide range of product selection were the primary growth drivers for online retail businesses. In addition, improvements in the communication industry have enabled businesses to locate cheaper labor on and beyond the United States boundaries, fueling growth in outsourcing and offshoring services.

In the midst of these opportunities, significant business challenges are emerging. The drop in the cost of information has expanded market limits to an unimaginable level, resulting in what is now called global markets or globalization. Global markets are providing consumers with a broad set of choices in products and services. With the help of a cheap computer and the Internet, one can find and compare product prices from different stores in different states, countries and continents.

Opportunities and challenges

Today, customers are looking not just for low prices but also for highquality products and services. With abundant computing power at hand, people can find products and services worldwide, not just at their neighborhood shopping centers. This is a challenge for businesses that must compete worldwide for sales, not just in their city, state or country.

Businesses of any size constantly need to be innovative about finding ways to compete effectively by cutting costs, increasing efficiency, maintaining or surpassing existing quality standards and optimizing operations. Over the years, rapid and uncoordinated innovations in the information technology sector have left businesses and institutions with a pool of heterogeneous computing applications and systems. As a result, it can be expensive and time consuming to manage and use information from multiple vendors and platforms. Large software corporations have realized the need for building interoperable systems. However, development cost and the need for companies to differentiate themselves will continue to hinder these efforts.

Systems integration is about bringing together the components of computing subsystems into one system that functions as a single unit. Integration is meant to drive high performance, reduce system complexity and optimize the whole IT infrastructure. A well-integrated system enables effective information sharing and collaboration between company employees, vendors and business partners. This sharing helps improve the processes of serving customers and delivering products, key issues that help a company achieve its goal of a better bottom line.

Even though systems integration is an integral task in company mergers and acquisition processes, today’s large and sophisticated computing systems compound the complexity of integrating different systems. According to banking merger expert Michael Koetter of the University of Groningen, extremely difficult or complex post-merger IT integration problems often are to blame for banking mergers that do not reward investors.

In the quest for ways to minimize operational costs and increase profits, those who adopted systems integration early implemented electronic data exchange (EDI) systems. EDI systems mostly were installed by large institutions and businesses because of the expensive initial setup and maintenance costs. EDI systems save large businesses and institutions money by replacing information flow that formerly required extensive human interaction and paper documents, meetings and faxes. Such systems also served as document management systems and improved the cost of sorting, distributing, manipulating, organizing and searching information.

While EDI systems have helped, the increasingly global nature of competition forces businesses to seek more operational efficiencies to reduce product and service costs. These efforts have paved ways for exploiting available computing power, hence the explosive growth in systems integration beyond EDI systems. Today, systems integration is a core task of any business. An additional systems integration challenge is the need to find better ways to manage master or reference data.

Master data quality is a business problem

When organizations have multiple copies of information that are different but should be the same, then the business stakeholders — not the information technology personnel — have a big problem. It is a business problem because management uses this information to make decisions that affect the whole organization. For this reason, instead of just relying on IT personnel, stakeholders should pay more attention to the quality of data.

Also called reference data, master data play a key role in the core operations of a business. Master data are shared and used by several applications that make up the whole system. And an organization’s data has value beyond its operational scope. In fact, it is not unusual for a company to acquire another company primarily for access to its customer master data, Roger Wolter and Kirk Haselden wrote for Microsoft.com in 2006.

Master data may include clients, customers, products, employees, inventories, suppliers, stores, assets and contracts. Business operations revolve around master data. The data are shared by multiple users, groups, partners and affiliates across the entire organization. For example, in a financial management firm, master data could include portfolios, securities, analytics and financial research networks. For a bank, such information could include account numbers or a customer list. The data could divide up a company’s normal customers vs. its premier customers.

The ability to access and modify master information from different software applications by different users, groups, business partners and affiliates presents a major problem for a business in terms of data maintenance. After all, management relies on systems to provide high-quality, consistent and reliable information so executives can make critical and sound decisions.

Having unsynchronized copies of the same information can at least cause problems, and at worst it is a recipe for disaster. You might have noticed that your credit card companies sometimes mail you promotions to apply for a credit card that you already possess. This can happen if customer information used by the marketing and servicing departments is out of synch.

Another common example of poor master data management happens when customers move and update their billing addresses. If the billing department does not get the customer’s new address, the customer will not receive the bill. This can lead to unpaid bills, followed by submitting the customer account to a collection agency, loss of the customer and even lawsuits. A system that effectively manages master data can smooth operations, prevent such errors and make sure decision makers have access to the right information.

Master data management to the rescue

One recent development in systems integration is called master data management (MDM) systems. According to the CDI-MDM Institute, the MDM market exploded from $2 billion in 2007 to an estimated $10 billion in 2009.

What entails management of master data? Many definitions for MDM have been coined, but the most appealing one comes from Haselden and Wolter. According to them, MDM is the combination of tools and processes required to create and maintain consistent and accurate lists of master data. In other words, it is the set of tools and processes that allows decoupling master information from individual applications while offering effective, stable, transparent and reliable ways to maintain master information even as it changes at application level or at the master copy.

The need for integrating and actively managing critical information on integrated systems never has been more pivotal. Some experts are proposing application of Six Sigma processes to manage key company data. Joe Danielewicz, an IT data architect at Motorola, argues that even though humans tolerate poor quality data and use context to fill in the gaps, businesses can use the Six Sigma methodology of defining, measuring, analyzing, improving and controlling (DMAIC) to manage their master data and mitigate project risk. Even though Six Sigma standards are too high to achieve in this arena, such steps will reduce issues in management of master data.

Furthermore, the global economic recession has triggered more government regulations in the financial sector, which now is taking proactive steps to mitigate some of the data risks. According to a recent data management survey of 52 senior financial industry executives, 28 percent responded that the No. 1 driving force for strategic investment in reference data management is improving data quality and reducing data errors. No. 2, at 21 percent, was the need to integrate information systems, reported Melanie Rodier for Wall Street and Technology in June 2010. In the same research, 31 percent of the respondents said they anticipated that better data would reduce risk, followed by 18 percent who expected better customer satisfaction and retention.

Even more interesting, businesses are starting to see master data management as a communication problem. They view MDM as another way to get different parts of the business collaborating to optimize operational efficiency for the whole organization. Writing in eWeek in 2007, Chris Preimesberger quoted an Intel data architect and MDM product manager who maintained that installing an MDM project got Intel departments talking to each other about common goals. If different parts of the business cannot speak the same language because of data issues, chances are greater that the entire customer fulfillment process will be disrupted, which could lead to losing customers, problems with regulatory compliance and dire consequences for the bottom line.

Another important key challenge in data management is trust. How do you get business users to trust and rely on the data presented to them, especially if different systems present a differing view of the same information? The ability to track data as it passes through different nodes of transformation to the final destination is called data lineage.

This ability increases data reliability and transparency and minimizes guesswork because users are able to see where the data comes from, what was done to it, who did what to it and when they did it. It allows data consumers to acquire a holistic view of the data lifecycle. Thus, a good MDM system must be competent at addressing data lineage issues. Having a holistic view of data helps businesses answer regulatory compliance audit questions and reliably deal with important business questions and decisions.

MDM options are growing

Until recently, the prohibitive expense of implementing a robust master data management system has been difficult for most small to midsize firms to justify. The cost easily could reach $1 million, which is out of the reach of many non-Fortune 500 companies. But the drive for business efficiency and competitiveness, along with increased requirements for regulatory compliance, has made a more sound case for having an IT department build an MDM system.

Furthermore, a lot of new MDM vendors are entering the market, including Teradata, NCR, Talend, Kalido and ObjectRiver Inc. Even the most recent release of Microsoft’s SQL server product includes a free version of an MDM system called Master Data Services (MDS).

Microsoft’s entry into MDM space is a very significant milestone when you consider that the majority of small and midsize businesses use Microsoft solutions to manage their data and daily operations. It means that small firms that previously could not afford an expensive MDM solution can find an out-of-the-box solution by just installing MS SQL Server R2. And Talend, a California-based MDM software vendor, offers an open source MDM solution that interested parties can download, modify and implement for free.

The benefits of a good master data management system are unlimited. Though cost has been a major bottleneck to many businesses, new vendors entering the market, including Microsoft and open source communities, will reduce the cost of implementing an MDM system considerably. This means that more companies will be able to implement an MDM solution and start reaping the benefits of a system they could not afford or justify just a few short years ago. By implementing an MDS solution, companies will reduce or minimize regulatory compliance issues, improve operational efficiency and add to their bottom line. By taking guesswork out of data through a transparent critical data management process, data consumers will be able to make quick and timely decisions about operational and regulatory issues.

Mashiku Kuyi Stephen is an assistant vice president at the Trust Company of the West (TCW), an asset management firm in Los Angeles. He is an expert in database architecture and SQL server query optimization. He has extensive experience dealing with financial systems integrations. His previous employment includes work as a database developer and software engineer. He has a B.S. in computer science from the University of California, Riverside, and an MBA in finance from California State University, Fullerton.

Brian H. Kleiner is a professor of management at California State University, Fullerton. He received both an MBA and a Ph.D. in management from the University of California, Los Angeles (UCLA).

Print: Share: