The start of a new school year is bound to be stressful. However, in September 2004, students at the University of Massachusetts-Amherst (UMass) were more stressed than usual.
That year, a buggy system left around 24,000 students unable to register for classes or collect their financial aid checks. The culprit? A crash that occurred in the administration’s student information system (SIS) earlier that month.
What led to the University of Massachusetts ERP failure, and how can you avoid similar mistakes? Today, we’re taking a closer look at the events leading up to the breakdown.
The University of Massachusetts ERP Failure: What Happened?
This failure started as an innocent attempt at a system upgrade. Historically, UMass automated much of its back-end business management activities through a platform known as Spire. The facility used Spire to handle core administrative activities, including:
- Student registration
- Student aid
- Class schedules
In 2004, executives decided to implement PeopleSoft Web Portal onto the Spire platform. The effort officially kicked off in July.
While project team members identified some minor errors during the testing phase, they believed they resolved them. The SIS system went live the first week of September, just in time for classes.
The Crash Heard Around Campus
For a few weeks, the system was slow but operated satisfactorily. Then, on September 7, it crashed completely. Though the cause is still unknown, many attributed the event to a heavy workload that strained the software.
Though it eventually came live again, the service continued to operate slowly and intermittently. From that point forward, the damage seemed to be done.
Soon, confusion spread across campus, with frustration quick on its heels. In all, the outage affected nearly 18,000 undergraduate students and 6,000 graduate students.
By the time the first day of school arrived, the system was still barely hanging on. Without class locations, students were left stranded, waiting in droves by the registrar’s office or visiting the registrar’s website to find answers.
Without an automated way to contact students, school officials were forced to use e-mail blasts and post updates on the university website.
Eventually, the UMass IT staff worked with PeopleSoft officials and a third-party consultant to troubleshoot the issue. Part of the resolution came from a makeshift workaround that required students and faculty members to log into the portal at staggered times to prevent a second overload.
PeopleSoft officials denied that the issues were software-related, suggesting that implementation issues were the more likely culprit.
5 Lessons Learned from this SIS Failure
1. Plan for a Lengthy Timeline
When you plan an implementation, it’s understandable to expect your system to be up and running by a certain date. However, it’s important to keep your milestones flexible. It’s better to go live later with a fully functional platform rather than push forward with a buggy one.
Instead of relying on the timelines proposed by your ERP vendor or systems integrator, take the time to benchmark your project against similar efforts in your industry. How long did those organizations take to implement their system? Use this data to set realistic expectations and plan accordingly.
At the same time, don’t underestimate how much time you’ll need to properly maintain your ERP system down the road. Make sure there is a solid maintenance plan in place, so you minimize any downtime.
2. Lead with a Solid Project Plan
Your ERP project plan should include key details such as project scope and roles and responsibilities. When all details are ironed out, you can be sure the plan realistically represents the effort in terms of budget and timeline.
Without a project plan, the project can feel like blindly sailing in the middle of a storm. You might have a general idea of the direction you need to go, but you don’t have the resources in place to get there.
A more thorough project plan could have saved UMass from such a long system outage. In addition to providing more specific verbiage around testing, the plan should have established a clearer chain of command. This may have prevented the registrar’s office from being flooded with questions from concerned students and faculty members.
3. Plan for More Testing than You Expect
The writing was on the wall with the UMass failure. There were bugs in the system from the beginning, while the project team mistakenly believed they were all resolved before go-live.
This is a case where more rounds of ERP system testing could have made all the difference.
For example, conference room pilots could have allowed end-users to try out the platform while it was still in development. These pilots often reveal issues within the system that aren’t always apparent when it’s tested in a non-operational environment.
Had the IT staff at UMass taken this approach, they may have been able to prevent the September 7 crash that led to the weeks-long domino effect.
4. Ensure Software Scalability
One of the factors that contributed to the UMass failure was a system overload. Put simply, there were too many people trying to access the PeopleSoft platform at once. Even when the main bugs were fixed, student access was still staggered to prevent the same issue from occurring again.
The lesson here is that software scalability cannot be overlooked. Instead of selecting ERP software that just barely meets your current needs, look ahead and consider how you can plan for an increased number of users or transactions. Then, choose software with built-in scalability that can support this expansion.
Keeping a long-term perspective can save you time and money in customizations and upgrades down the road. This also maximizes your return on your investment because you can use the system for years to come.
5. Prepare for System Problem Solving
Rarely will any ERP implementation go off without a hitch. If you expect everything to work perfectly right from the beginning, you may fail to put the proper protections in place.
We recommend developing a contingency plan to ensure your team has the capacity and bandwidth to resolve issues as soon as they occur. This will require coordination between internal resources, software vendors, and third-party consultants.
If the University’s teams had been prepared for problem solving, they could have addressed problems as soon as they occurred.
Your Ticket to Long-Term ERP Success
While the system eventually came back online, it was a major step backward for the campus. The University of Massachusetts ERP failure affected tens of thousands of students, sending them scrambling for answers during a season that was already nerve-wracking.
If you’re planning an SIS or ERP implementation, consider how this story underscores the importance of scalability, project planning, and system testing.