Mating the elephants --
ERP implementation in Japan
by Matthew Mai

Mating elephants is one of a zookeeper's toughest jobs. It involves a lot of bellowing and trumpeting, occurs at a high level, and takes two years to produce results. This, oddly enough, is a pretty fair description of the typical ERP implementation project in Japan, except for the part about happening at a high level -- which perhaps is a cultural clue as to why Japanese corporations are facing extra costs, increased time, and reduced functionality from their ERP efforts.

Japan's market for enterprise resource planning (ERP) software packages, like those provided by SAP, Oracle, and JDE, has grown significantly over the past several years. Many large companies in Japan are now in the process of implementing or have implemented these packages, and in addition, many foreign companies are implementing global or regional templates in Japan. Many of these implementations are limited in scope, take more than a year to put into service, and are extremely costly. In addition, the potential savings of implementing a package-based integrated system are not being fully realized, in contrast to experience in the United States and Europe. Why the extra costs in Japan?

Often, the cost increase is blamed on allegations that the software fails to meet Japanese business requirements, and, to some degree, this may be true. However, based upon a closer examination of successful implementations in Japan, it appears that the approach to implementations here is the biggest culprit, largely due to differences in culture between Japan and the United States and Europe. These differences are seen in the decision making process, in differing views toward change and reengineering, a dissimilar focus on customer service -- including the concept of users as "customers" of the ERP system, and in quality versus time management and functional approach issues.

How the elephants do business in Japan

One of the key success factors in implementing an ERP package is to use the standard functionality provided by the vendor. This typically requires a company to change the way it does business, and this approach often conflicts with the approach taken by many Japanese ERP implementations. As a result, there are typically many significant modifications, requests to add functionality, and more reports must be developed for a Japanese ERP implementation. This increases the cost and duration of the implementation, and in addition, the ongoing support costs are much higher and later upgrades become more difficult and time consuming. In other words, there's quite a thud when the baby elephant hits the ground.

There are several cultural factors that contribute to the large development effort associated with these implementations. The primary factor is the Japanese view toward reengineering and change. For many companies in the United States and Europe, management's focus is on significantly changing their business to become more globally competitive. ERP package implementations are often used as the tool to effect such a change. This focus is atypical in Japan. To some degree, there is a fear of change. Likewise, these projects are often viewed as IT projects, and lack top-level management commitment or understanding from the very beginning. Therefore, the driver for change is missing. As a result, the ERP implementations in Japan often mirror the current state of the business.

Another factor that leads to increased development costs is the Japanese focus on customer service. For many project teams, a requirement issued from any level of management must be delivered. Frequently, user needs will not be questioned. The project teams take great pride in developing elaborate solutions to meet these requirements without ever taking the time to understand the true cost, business benefits, or true requirement of each request. As a result, the solution is often complex, involves intricate configuration, or worst yet, requires modification to the standard ERP system. Management is usually unaware of the amount of modifications, added functionality, or reports created, and, more importantly, unaware of the future ramifications of such an implementation approach.

Zoo culture is key
Another cultural factor that results in longer and more costly ERP implementations is the decision-making process. On a recent project, a system integration firm planned to implement most modules of an ERP package in less than a year. The ERP package vendor told the firm that it takes a typical project in Japan twelve months just to get the project started. This is clearly the result of the need for consensus on almost all decisions. Given the magnitude of most ERP projects, the project team must be empowered to make decisions. Likewise, a strong and active steering committee is required to make decisions that the project team cannot make. Top management and the project team must understand the impact of delays caused by its inability to make decisions. One technique employed in the United States is to describe the delaying of critical decisions in terms of dollars lost.

Real elephants don't mind the wait
In addition to the decision making process, a lengthy and costly ERP implementation is often the result of the Japanese view of quality versus time management. On any project, there is, to some degree, a trade off between quality and time. In the United States, the concept of timeboxing is often used on ERP implementations. The timeboxing concept was originally developed as a control technique for projects using prototyping (see sidebar). Because of the creative nature of prototyping sessions, project "scope creep" pressures can be great and the project can easily get out of control. Timeboxing focuses the prototyping effort on a set of objectives and restricts the time in which the development team has to meet those objectives. The exact functionality of the deliverable is not as important as the fact that the deliverable must meet the original objectives. In a timebox, the deadline is fixed and the functionality is flexible. If the endpoint is flexible, the discovery process goes unchecked and the control mechanism is lost. This concept is often hard to implement in Japan where the typical project team would prefer to extend the discovery process to get the "ideal" solution regardless of time deadlines.

One elephant asked: "What's the aim?"
Finally, a functional view of ERP implementations contributes to long and costly implementations in Japan which manifests itself in two ways. First, the scope of many ERP packages is very narrow and functionally oriented. As a result, the ERP package is interfaced to many legacy systems to support the business. This results in increased custom development and testing time. As with all interfaces, there are significant support costs to maintain them, and to maintain the data consistency. With these implementations, it becomes very difficult to add functionality to the ERP system since an overall enterprise view was never taken. Each additional function takes that much longer to implement. Second, even companies that implement most ERP modules face difficulties related to functional activities. These companies often modify or add-on functions to optimize individual functions and do not look at the overall cost-versus-benefits impact of such modifications to the enterprise as a whole.

Showing the elephants how to do it
In summary, there are many cultural issues that make implementing an ERP package in Japan expensive, time consuming, and challenging. If these factors are understood up front, these issues can be dealt with and hopefully minimized. However, this will require a strong project manager and project team who have the ability to know when to push hard and when to compromise. It will also require a strong and active steering committee. The steering committee must include key business executives, as opposed to IT executives only, who can champion change and truly understand the value of using the standard ERP system even if it means changing their own business processes. This will set the tone for the entire project. With a little luck, that baby elephant will grow up to be a slimmed-down beast requiring few peanuts and even less cleaning by the zoo keeper.

Matthew Mai is a consultant at Ernst & Young, Tokyo.
Contact him at: matthew.s.mai@jp.eyi.com

Back to the Table of Contents

Comments or suggestions?
Contact cjmaster@cjmag.co.jp