Chapter 6 Foundations of Business Intelligence: Databases and Information Management

Panasonic Creates a Single Version of the Truth from Its Data

anasonic is one of the worlds leading electronics manufacturers. It operates under the auspices of parent company Matsushita Electric Industrial Co. Ltd., a conglomeration of over 600 firms that is based in Kadoma, Japan. Collectively, the businesses of Matsushita manufacture 15,000 products for a global market and employ 330,000 people internationally. In Europe alone, Panasonic has 15 sales subsidiaries, 14 manufacturing facilities, five research and development centers, and seven administrative stations. Add in major presences around the world, including Asia and North America, and it is clear that Panasonics operations cover the globe. With so many different sources of data, the company found itself with product and customer data that were often inconsistent, duplicate, or incomplete. Different segments of the company used their own pools of data, which were completely isolated from the data that the rest of the company was using. These conditions combined to be a drag on operational efficiency and drained significant amounts of money from the corporation as a whole. The types of data required to launch a new Panasonic product included photos, product specifications and descriptions, manuals, pricing data, and point-of-sale marketing information. Employees adapted product information to suit the needs of their country or region. It took considerable time and effort to sift through all the data and create a common set of data for launching products globally, which allowed competitors to infiltrate markets that Panasonic did not reach in its first phase of a launch. To solve this problem, Panasonic decided to pursue a single version of the truth. Daily activities required the data to pass though legacy systems, fax machines, e-mail, phone calls, and regular mail. With so many people handling the data in such a variety of formats, inefficiencies and inaccuracies were always a risk. Erasing these problems promised to increase Panasonics speed of bringing products to market. Panasonic was enjoying a number of successes: a market leadership in plasma TVs, a successful transition of company presidents, and a wellreceived marketing identity, Panasonic: Ideas for Life. However, these positives were overshadowed

by the administrative costs incurred by such an immense organization. Thus, when Fumio Otsubo took over as president in June 2006, he took over a company with an operating profit margin of only 5 percent. The board of directors saddled him with the goal of increasing the margin to 10 percent by 2010. In Panasonics industry, consumers expect the price of new technology to decrease over time, which it had for items that were strengths at Panasonic such as plasma TVs and DVD players. Therefore, Otsubo could not expect to increase the companys profit margin by increasing prices. Instead, he had to set his sights on reducing costs and increasing sales. Starting in Europe, Panasonic sought to replace its pull model of data dissemination with a push model. Previously, employees in marketing and sales had to request data from numerous repositories. Under the push model, a centralized data bank sends the information to all employees who need it at the same time, ensuring uniformity. The recipients of the data include retail partners and e-commerce vendors, who receive complete product information at all stages of a product rollout. Panasonic employees receive data on a more targeted basis. The benefits to Panasonic Europe are more consistent product rollouts and product information. The latter ensures that customers do not become confused while researching their purchases, which could motivate them to abandon Panasonic for a competitor. The technical force behind Panasonic Europes data management overhaul was master-data-management (MDM) software from IBMs WebSphere line. The software enabled Panasonic Europe to consolidate data, as well as systematize the business processes related to the data. Overall, the company gained better control over its internal data. Generally speaking, MDM software aims to merge disparate records into one authenticated master file. Many companies have adopted MDM to fix discrepancies among the databases used by their various departments (e.g., the accounting department having record of fewer customers than the number of customer IDs in the CRM database). MDM is particularly useful for companies that have data integration issues as a result of mergers or acquisitions. Small-

Part Two Information Technology Infrastructure

and mid-size firms generally do not have the kinds of challenges that would require an MDM solution. Implementing MDM is a multi-step process that includes business process analysis, data assessment, data cleansing, data consolidation and reconciliation, data migration, and development of a master data service layer. These steps produce a system of records that stores the master file for all of the companys data. It is critical for the organization to institute strict policies against computing activities that could compromise the authenticity of the data. Once the MDM is in place, employees and applications access a consolidated view of the companys data. The implementation should enforce standards for the formatting and storage of data, such as the number of fields in an address record or the number of digits in a ZIP code. The service layer of the MDM preserves the view of the master data for applications and synchronizes updates to the master file. In the case of Panasonic, the deployment of the IBM MDM software paid quick dividends. Within a year and a half, Panasonic Europe was getting products to market faster and spending 50 percent less time creating and maintaining product information. Time-to-market for a product was reduced from five to six months to one to two months. According to internal calculations, Panasonic Europe improved its efficiency by a factor of 5 and anticipates saving a million euros a year while increasing sales by 3.5 percent. However, analyst Paul Jackson of Forrester Inc., cautioned against high expectations of boosted sales based on data management improvements. He pointed to pricing, innovation, and strategic partnerships as better strategies for long-term market share increases. When Panasonic North America had to reconcile its data, it did not have to confront the challenge of multiple countries with multiple languages and currencies complicating product launches, as had its European counterpart. However, the challenges of reorganizing workflow and consolidating product information were just as daunting. Panasonic faced this issue when it needed to provide a consolidated view of product information for retail giant Wal-Mart. Panasonic started by identifying the information that Wal-Mart would need, which was data that adhered closely to industry standards. Then, the electronics maker searched its legacy systems for the sources of the required data. Finally, Panasonic worked with IBM to create an interface apparatus to collect the required

data for a repository. Some information, such as that produced by newer business processes, was not available in the legacy systems. Panasonic had to add new interfaces in order to include this information, and then build an application-integration layer to send the whole package to Wal-Mart. Each of the companys multiple facilities made its own contributions to new products. More importantly, the facilities had their own cultures and information infrastructures. They also valued their autonomy and the flexibility it furnished. Different Panasonic entities might be unwilling to give up control over information due to the perceived loss of power. The company required clear master data management rules to prevent too many hands from manipulating the data so that the master file would remain pristine. Panasonic North America Information Technology vice president Bob Schwartz hoped that the fierce competition threatening the standing of his company would help convince the traditionalists to support data-sharing. However, he expected that convincing the enterprise of this would be an uphill battle. Besides all the units of Panasonic North America, there were manufacturing partners to bring aboard. Without them, the system could not fulfill its complete potential. This had been a serious challenge for Panasonic Europe, where most partners were based in Asia and were content with their manual processes for managing product data. Paul Bolton, senior manager for e-commerce and customer relationship management solutions, deployed the product information database at Panasonic first. Once it proved effective, he then presented its capabilities to the other manufacturers and won them over. Schwartz therefore had a strategy and a roadmap to clear that hurdle. What remained was perhaps the biggest hurdle: convincing the corporate office in Japan that their data management strategy deserved global adoption. Only then would the application of MDM principles achieve its full benefit. In the meantime, Schwartz reached out to Panasonics vendors in the U.S. and gained additional profits from the companys improved data. Panasonic was able to use the data to reduce the amount of time that vendors such as Best Buy and Circuit City kept high-cost inventory, such as large-model TVs, in stock from 35 to 7 days, thereby increasing their profit margins.
Sources: Duff McDonald, Case Study: Panasonic Searches the Master Data for a Single Version of Truth, CIO Insight, May 22,

Chapter 6 Foundations of Business Intelligence: Databases and Information Management

CASE STUDY QUESTIONS

ability to execute its strategy? What management, organization, and technology factors were responsible for those problems? 3. How did master data management address these problems? How effective was this solution? 4. What challenges did Panasonic face in implementing this solution?