There are few information systems whose design and implementation challenge a company like ERP. Done right, a new ERP implementation can dramatically improve business processes. But when an implementation fails -- or takes a prolonged and arduous course -- tens of millions of dollars could be misspent.
I know of a company that spent $85 million and more than three years on a suite of applications that ranged from finance to human resources to supply chain management -- and the system still didn't function well. I also know of an even larger company that invested more than $200 million in an ERP system and later abandoned the project.
In spite of the risks and challenges, companies are increasingly redesigning the infrastructure systems that run their business. In most cases, they have no choice. To stay competitive, those companies must leverage IT in new ways.
Here are some ideas on how to get real value from an
ERP must be driven by a business case.
In other words, the work must be directed toward improving specific business metrics: improved cash flow, faster hiring, reduced costs and accelerated shipments.
Both line and IT managers must identify the processes that need improvement. Focusing on business process improvement is the only way to know whether you got your money's worth.
Active executive direction.
Executive direction in this case means more than just oversight or support. People don't adapt to doing work in new ways without some inspiration from the top. ERP systems are huge, influencing thousands of people, processes, practices and policies. There are thousands of decisions to be made -- both large and small. Someone at the top of the organization has to make the calls.
I have seen ERP projects get delayed for months as shop foremen argue about the form of a new report. If there isn't active executive management of an ERP implementation, the IT manager should stop the project. It's doomed to fail.
Always favor the standard.
Be prepared for people to argue that they need to customize a report -- or tell you that the way your company pays invoices is not accommodated by the new software. However, you should always challenge those arguments aggressively. Most companies can live with standard business processes. When you overcustomize an ERP installation, you not only add time and costs, but you also make installing the next release of the ERP software increasingly difficult and costly.
Remember that there are very few business processes that truly differentiate your company or require customized software.
Make ERP-related decisions quickly.
How about a rule that all decisions must be made within three days? This may seem arbitrary, but setting a fast and deliberate implementation pace is critical.
Experience has shown that when an ERP project stretches more than 10 months, it's at risk -- mostly because key team members move on.
Put the very best people on the implementation team.
The natural tendency is to staff projects with people who are the most available. But these people may not have the skills or process knowledge required to get the job done. This is complex stuff. An ERP project will require the best talent that you have. And if you assign only part-time talent, then work gets prolonged and the project is at risk.
ERP work is not just a project. It is the beginning of a continuous improvement process. There are at least five years of benefits to get out of any ERP implementation.
Although you should set an objective to have the technology platform up and running within 10 months, with demonstrable business results, keep a team of people focused on continuously using the technology to improve business performance -- lowering costs, increasing profits, improving customer satisfaction. There are at least five years of benefits to get out of any ERP implementation.
James Champy is chairman of Perot Systems Corp.'s consulting practice and head of strategy for the company. He is also the author of the best-selling books Reengineering the Corporation, Reengineering Management, The Arc of Ambition and X-Engineering the Corporation.