Enterprise Resource Planning

Factors Affecting Success and Failure  

Patricia Barton

November 25, 2001

Table of Contents

Introduction

Factors Contributing to Failure

ERP Success Stories

Conclusion

References

Click here for Printer Friendly Version

Introduction

What is Enterprise Resource Planning (ERP)?  “Enterprise Resource Planning” is a term originally coined in 1990 by The Gartner Group to describe the next generation of MRP II software.  The purpose was to integrate all facets of the business enterprise under one suite of software applications.  The definition of ERP would be broadened to include almost any type of large integrated software package.[1]  Webopedia provides a generalized definition of ERP as “a business management system that integrates all facets of the business, including planning, manufacturing, sales, and marketing.”[2]  Some of the more well-known ERP software developers include SAP, Oracle, and PeopleSoft.

This paper will look at both successful and unsuccessful ERP implementations and what contributed to their success or failure.  Many lessons have been learned by failed ERP projects, as evidenced by the volume of information available.  Many of the failures occurred in 1999, in an attempt to manage Y2K issues, which may suggest that the companies had pressing needs which forced the implementation.  Apparently, late adopters are benefiting from the mistakes of their predecessors since the most current research describes successful implementations.

What constitutes an ERP implementation failure?  There are degrees of failure with ERP projects.  The most obvious failure is never actually implementing the ERP system.  But a project can be considered failed if the new system is not fully utilized.[3]  According to a survey conducted by Forrester Research in April 2001, only six percent of 500 companies surveyed considered their ERP systems effective, while 79 percent said they were not effective or only somewhat effective.[4]  Five of the top ten Corporate Information Technology Failures cited by Computerworld involve ERP projects.  See chart. (requires Adobe Acrobat reader) 

 Factors Contributing to Failure

Apparently, no single point of failure can be attributed to unsuccessful ERP implementations.  Some of the causes cited for failed ERP projects include:

Inherent complexity of ERP implementation

The Problem

ERP Systems are complex, and implementing one can be a difficult, time-consuming, and expensive project for a company.  The technology is tightly integrated and requires a commitment from all divisions and often a change in the way a company does business to make it work.[5]  It can take years to complete and cost as much as $500 million for a large company.  Moreover, there is no guarantee of the outcome.[6]

A notorious example of a failed ERP implementation is the Hershey Foods’ SAPAG’s R/3 implementation.  The company spent $112 million and 30 months on their ERP project.  When they went live in July 1999, the company experienced problems pushing orders through the system, resulting in shipping delays and deliveries of incomplete orders.[7]

Many reasons have been cited for the Hershey ERP failure.  One, the project was originally scheduled to take four years, but the company forced the implementation to go live in just 30 months.  Two, the company simultaneously implemented a customer-relations package and a logistics package, substantially increasing the overall complexity and employee learning curve.  Three, the company went live at their busiest time of the year, just before Halloween, and the resultant delays caused third quarter profits to fall by $151 million compared to the previous year.[8]

One step often taken to reduce the complexity and time involved in an ERP project is using process templates.  Managers are not willing to spend the time, effort, and money to work through the complexities inherent in configuring an ERP system to company- specific processes.  They use process templates to short cut the process and accelerate implementation.  Although the templates are simpler and speed up the process, they promote generalization, which can limit performance gains and competitive advantage.[9]

The Solution

If possible, plan the ERP project go-live date during the company’s slow periods.  Roll out the modules in stages and don’t attempt to implement other applications simultaneously.  Don’t speed up the timeline – critical testing and training of users will likely pay the price.

Some degree of conformance to process logic will likely be required.  The company must ensure that process templates utilized in the implementation reflect best business processes for their organization.[10]  In some modules, Human Resources, for example, the processes tend to be relatively standardized so utilizing templates in this module may make more sense than utilizing them in processes that are highly unique to the company.

Outside Consultant Issues

The Problem

In W.L. Gore’s lawsuit against PeopleSoft and Deloitte & Touche, the company alleges PeopleSoft sent in unqualified consultants to do the job, forcing the company to rely on the customer service hotline to set up the program after major problems occurred when the system went live.[11]  Deloitte & Touche paid referral fees to PeopleSoft, which prompted PeopleSoft to recommend them as consultants even though they didn’t have the expertise required to implement the software.  Gore was then required to hire another group of consultants to fix the damage, which cost hundreds of thousands of dollars more.

An Atlanta-based forestry products manufacturer used four consulting firms in various phases of its SAP implementation project.  According to the CIO, the consultants “bickered among themselves over how to approach the project”.  Rather than partnering with the manufacturer, they were fighting for control and overstaffing the project, which the company eventually shelved. [12]

The Solution

It’s important to determine prior to project initiation the expertise currently available within the company. This will help define the role of the consultant.  The project lead should interview the staff proposed for the project, ensure that the contract stipulates that those same people will work on the project, and in some cases remuneration may be based upon the successful completion of the project.  It’s also important to integrate consultants with corporate staff to ensure a smooth knowledge transfer at the conclusion of the project.[13] 

Inadequate Training

The Problem

Increasing reports point to poor training as a major cause behind failed ERP projects.  Not just education of the technical staff, but of the user community who are supposed to actually work with the system.  ERP changes the way companies do business but, instead of training everyone in the company on how to do business differently, they are trained on new computer software.[14]

Companies must find the right person or organization to conduct the training.  At Purina Mills, the company contracted with a third party consultant to conduct the training, but users complained almost immediately that they needed a translator to help them understand the training.  The company needed someone to train the users that understood the current process and could relate it to the new one.  They fired the third party trainers and developed the training course internally. [15]

The Solution

One CIO differentiates between education and training as education being the why, who, and where and training as the how.[16]  Rather than basic, software-specific training, a more broad-based understanding of the flow of information through the system is needed.

Mead Corp has undertaken an ambitious education program.  Prior to project go-live, every employee will attend a course explaining why the company is becoming more standardized and what an integrated supply chain looks like.  Users affected by process changes will also be further educated as to how their process fits in to the overall supply chain.  Just prior to “go live”, task training will occur so that it is fresh.  The CIO suggests that “it requires leadership after implementation and a lot of reinforcement, retraining, and reeducation.”[17]

Process Risk and Process Barriers

The Problem

“Process Risk” is the risk that a business will suffer significant financial losses or harm to its reputation as a result of significant changes in the way the company does things. 

There are several types of process risk:

• Performance dips – efficiency drops as employees learn new jobs and technology

• Project fights – when problems occur, top management drops the project

• Process fumbles – the new implementation is more than the company can handle, timelines slip and performance problems crop up

• Process failures – after go-live, the new process simply doesn’t work[18]

 According to Barry Calogero,  “...the real culprit is the process.”  Three process barriers cited as contributing to ERP failure are:

1. Focusing on technology – software alone will not solve business problems
2. Ignoring requirements definition – processes are adopted to fit the software or legacy processes are forced into software that’s not designed to handle them
3. Skipping the implementation plan phase – jumping from requirements definition to development phase[19]

The Solution

No amount of training can take the place of hands-on, real-time interaction with a new system.  Performance dips are likely.  The task is to minimize the length and depth of the dip through adequate training of the user community.  Top management must remain fully committed to the project.  Process fumbles and failures can be avoided by putting a good project team together, headed by an experienced project lead, as well as a detailed, well thought out implementation plan.

Also, develop good performance metrics.  Managing risk involves changing the way people are measured, managed, and rewarded.  Good metrics help managers assess performance more accurately in order to identify problems before they get out of hand.[20]

Corporate Culture

The Problem

Corporate culture refers both to the leadership sponsors’ involvement and accessibility in the project and the recognition of the role the employees play in a successful implementation.  Many ERP projects fail because the employees do not realize the needs and benefits of the project and will be resistant to change.  Additionally, the users do not take ownership of the project.

The forestry products manufacturer mentioned above didn’t consider the effect of an ERP implementation project on its divisional VPs.  In an attempt to centralize operations and integrate back-office systems, the company elected to implement SAP across all divisions.  The Vice Presidents of the 12 major divisions were unaware of the company’s long-term strategies and how those strategies would cause them to lose some of their autonomy.  When they finally realized this, 11 months into the project, they balked and the project was scrapped.[21]

Bruce Webster, director at Pricewaterhouse Coopers suggests that “IT culture is such that problems…are hidden.  Programmers write around buggy code rather than tear it apart.  Managers revise project specifications to reflect what they did instead of what they should have done.  Senior IT leaders neglect to tell their bosses the bad news.”[22]

The Solution

Robert Switz, CFO of ADC, suggests that CEOs and CFOs have to visibly support and monitor the progress of the project.[23]  Active business leadership in all project phases decreases the likelihood of project failure. 

According to Allen Web, author of “ERP Project Management Basics”, every project needs a sponsor, someone who can make things happen.  The sponsor must come from the upper reaches of the organization for the following reasons:

1. the sponsor must have the clout to obtain needed resources without scaling many approval layers
2. the sponsor must show that upper management has a stake in the project just as much as those operating the day-to-day chores, and
3. the sponsor must be able to bring warring parties to the table and force necessary compromises in a timely fashion.[24]

Alpha Industries experienced a successful ERP implementation that has resulted in decreased inventory levels, increased customer service, accurate information on finished inventory and work in process.  All of which has give management the ability to make better and timelier decisions.  One factor that contributed to the success of the project was involving the users from the outset.  They took the time to prepare the staff for change, allowing them to buy in to the change process.  They involved everyone and gave the users ownership of the project.[25]

Mead Corp’s initial phase of its ERP implementation has begun successfully.  Eventually, ERP will be phased into all eight of its divisions.  The project already had strong executive management support.  They also recognized the importance of employee participation in their ERP project from the outset.[26]  They understood that “people make these projects work”.  They gave the project a name, educated the users as to how they fit in to the overall process, and communicated regularly through brochures, newsletters, and training sessions. 

Unrealistic Expectations

The Problem

At an unidentified company, the user community wanted a Business Intelligence application installed in conjunction with the implementation of an ERP system in order to facilitate information from the ERP system.  By the time an outside consulting firm was selected to implement the project, months had passed.  However, the project go-live date had already been communicated to the users and the company was unwilling to change it.  The consulting firm, recognizing that user expectations would be impossible to meet, declined the project.  The firm that was hired failed to meet user expectations and the project was shelved.[27]

Sobesys, a leading grocery chain in Canada, also experienced problems with a SAP R/3 installation.  Dave Boulanger, a senior director with AMR Research, suggested that some Sobesys “executives have come from other organizations that may have installed competing systems, and as such, they may have different expectations.”[28]

The Solution

To avoid the problem of mismanaging user expectations, the project manager must be able to:

• Explain what can and can not be achieved
• Understand the user expectations
• Prioritize the user expectations and communicate the priorities to the user community
• Formally document the user expectations
• Continuously validate project deliverables to expectations[29]

Over-customization of Software

The Problem

A technological mistake often made in SAP implementations, says Graham McFarlane, director of Western Management Consultants, is that organizations modify the software more than they should, rather than modifying their business processes.[30]  Kevin McKay, SAP America’s CEO concurs.  He suggests customization almost always means trouble.  It hurts performance and can cause upgrade and testing headaches.

The Solution

The Military Sealift Command (MSC) successfully implemented Oracle’s ERP solution.  One rule they cite for achieving ERP success is that agencies must find an ERP package that mirrors their business practices as closely as possible, then resolve to implement the package without significant modifications.  MSC managers made a key decision to minimize the risk of ERP implementation by taking a “vanilla” approach: They committed to installing the software as it was packaged, without any modifications.  The MSC put together 1,000 requirements the optimal software would fulfill.  There were only 11 areas where their processes didn’t match the software and the commander made the decision to modify MSC processes to fit the Oracle software.[31]

Using IT to solve the problem

The Problem

Most companies are discovering that a quantified business need is a prerequisite for a high level of satisfaction with enterprise resource planning initiatives.  Many ERP initiatives are still system driven and these are more likely to fail than those that are business led.[32]  Often, management applies the technology as the solution to correct fundamental flaws in underlying business processes.  Companies implementing an ERP package often view the new technology as a new core competency, but it should only be viewed as a means to achieve the competency through better business processes.[33]

FoxMeyer Drug declared bankruptcy and cited a failed ERP implementation as the cause.  They jumped on the ERP bandwagon early, when the software was designed specifically for manufacturing, not distribution, companies.  The software was unable to handle processing demands.[34] 

The Solution

New business processes must be established, thought through, and implemented before software tools are selected, purchased, and rolled out.  Business evolution to ERP is about more than software tools.  An organization’s success will depend on redesigning the process and customizing the technology to fit that process – rather than the other way around.[35]

Before committing to a specific ERP software package, companies need to take the time to evaluate their ERP needs.  They need to define in advance:

• How they want to run their business
• What problems need to be resolved
• What are the priorities
• The current process – what works and what doesn’t
• An implementation plan – including timelines and deliverables
• What software will best resolve their problems, meet their goals and priorities[36]

Timeline flexibility

The Problem

Although many industry experts insist that ERP implementation timelines are closely followed, the company must ensure that the system is fully tested and ready for implementation or deal with negative consequences.

Whirlpool suffered delays in shipping product after it went live with an SAP system implementation.  Even though red flags had been raised late in the testing phase, Whirlpool elected to stick to their schedule.   The decision resulted in a crippled shipping system that left appliances sitting in warehouses and stores with six- to eight-week delays for shipping orders.[37]

Meritor rolled out an ERP system called Passport 21 in phases starting with a manufacturing plant in Wales.  Although they believed their training was adequate, it took 30 days and additional staff to ensure that orders were entered and the manufacturing schedule was not further delayed.  In order to ensure the problem never happens again, Monte Nuckols, the VP of IS, added two weeks to the timeline for all ERP deployments to allow for what he calls “a day in the life” testing.  Although this plan takes extra time, therefore extra expense, he believes it’s necessary to ensure a successful rollout.

The Solution

It is important to stick to schedules in ERP implementations because the process is lengthy and complicated and delays can increase costs substantially.  However, management must review closely the need for extending the timeline to ensure success of the project.  It is counterproductive to insist on a project deadline if meeting that deadline results in a failed implementation.

Infrastructure

The Problem

When Bio-Rad Laboratories implemented an ERP system, the WAN was brought to a crawl by conflicts between ERP and email traffic.  It was determined that email alone was using up the entire bandwidth between locations.  As a result, mission-critical data in the ERP system’s distribution and financial modules would languish at the various sites.  This stalled order taking and the shipment of product.[38] 

The mainframe at Cleveland State University was unable to handle the PeopleSoft application they installed and it had to be replaced by a Unix system.  This was one of the issues that caused the university continued problems more than a year after the initial rollout.

The Solution

Bio-Rad Laboratories switched to a thin client technology to reduce WAN traffic and installed a Packetshaper, a network management device that manages bandwidth. 

Infrastructure issues must be considered in the rollout of any large system.  Server technology, bandwidth measurements, and database requirements must all be researched prior to system installation.

Other Factors Attributed to ERP failures

There are many other factors cited in the research that contribute to ERP implementation failure including: 

Scope creep – Scope creep was another factor cited in Whirlpool’s failed ERP project.[39]

Inexperienced project managers – one mismanaged ERP implementation left a southeastern electronics manufacturer unable to accept deliveries and nearly closed a plant.[40]

Brain Drain – A semiconductor manufacturer in Silicon Valley lost 70-80% of its project core-team within three months after “go live”.  The main factor that influences success or failure after the consulting team’s exodus is the intellectual capital of the project core-team.  Core-team retention must be planned and executed.[41] 

ERP Success Stories

Web-enabled interface – Hunter Douglas was interested in automating its order processing system to give the assembly plant the ability to enter orders directly, as well as faster access to order status information and estimated delivery details.  They implemented a data integration tool with a web front end to extract business data from SAP.  The result is that customers can check order status and confirm delivery times in minutes, saving the company time and money.[42]

Communication – Georgia’s Department of Administrative Services (DOAS) implemented an ERP system on time and within budget.  There were considerable benefits.  Queries that used to take a month are completed instantly.  Annual contract reviews that could’ve taken weeks are now done in hours.  And it reduced the audit preparation time by at least 50%.  The DOAS attributes their success to constant communication via a Web page, email, instant messaging, as well as face-to-face meetings and extensive planning.[43]

Change the Process – Bradley Corporation implemented an ERP system and has realized significant benefits including lower inventory levels and warehouse space requirements, increased sales without added staff, decreased lead times and increased on-time deliveries.  Bradley attributes their success to reengineering the business processes rather than automating poor processes.  This enabled them to select a software package that reflected their new business processes.[44]

Phased Rollout – McKesson HBOC attributes their successful ERP project to a cautious rollout and strict adherence to plan.  The different pieces of SAP’s backoffice applications are being installed one by one.  Managers must sign off on each phase and any changes must be approved by an executive steering committee.[45]

Conclusion

There are as many reasons for successful ERP implementations as there are for failed projects.  However, success seems to often be measured by whether or not the project came in on time and under budget.  Whereas, fully utilizing the system to achieve improved business practices appears to be ignored.  Performance measures must be developed and standardized to give organizations a clearer picture of the benefits derived from Enterprise Resource Planning implementation.  Much has been written about and learned from some well-publicized successes and failures in ERP implementations.  Some of it has even been directly contradictory.  However, most agree on some basic rules:

• Establish the business processes prior to selecting the software.

• Staff the project team with members of the user community in addition to IT staff.

• Develop an implementation plan and stick to it.

• Train the users thoroughly on the process changes and flow of information in addition to the actual software.

• The project doesn’t end with “go-live”, but must be continually monitored.

[45] Craig Stedman, “Flash! ERP works if you're careful”, Computerworld, 12/13/99, v33 i50 p1,14