Pennsylvania’s ERP Selection Process

ERP Vendor Selection in general
ERP Vendor Selection in general

The Commonwealth of Pennsylvania had made a bold decision. Instead of replacing their aging legacy systems in a piecemeal fashion, they would implement an ERP system that would replace these systems and also integrate the 53 different governmental agencies in a single system. Like most state governments in the U.S.A, procuring things as simple as notepads and paper clips for the Commonwealth was fraught with regulations. On top of this it was estimated that this project would cost $250M and would take up to 5 years to complete. All of this too in a state where the Hershey ERP project was implemented just down the road from the state capital in Harrisburg. Generally, there was considerable distrust and suspicion surrounding this effort. The fact that no public sector agency had ever implemented an ERP system on such a scale was also a point of concern.

Vendor selection in steps
Vendor selection in steps

If the Commonwealth of Pennsylvania were a business, it would rank right up there on the list of Fortune 500 companies. In fact, with annual tax revenues or tax receipts of more than $20 billion, it would rank in the top 100. But few – if any – businesses are as multifaceted or complex as the Commonwealth, which is responsible for:

  • Educating 1.8 million children
  • Maintaining 44,000 miles of highway
  • Protecting public safety in 67 counties and 2,600 municipalities
  • Operating parks, public hospitals and correctional facilities
  • Attracting new businesses and encouraging economic growth

To fulfill these and other missions, the Commonwealth must manage its $20.8 billion budget; purchase $12 billion in goods and services; manage 84,000 employees; disburse $3 billion in payroll; and maintain records for more than 150,000 vendors.

ERP Vendor Selection for the ImaginePA Project

Because of the size and complexity of the proposed project it was decided by the Selection Committee that the software vendor and the systems integrator would be selected separately. And this later worked to the Commonwealth’s advantage. To assist in these two procurements it was also decided that Gartner Group should be brought in. After an industry study, Gartner Group helped to identify a short list of ERP vendors that included AMS, PeopleSoft, Oracle and SAP. In the meantime, the various subject matter experts in the Commonwealth had been meeting for some time and had already developed a list of about 700 functional requirements that the new software would have to support.
These four vendors were then provided with copies of the 700 functional requirements that had been identified. After studying these functional requirements the vendors had a conference call with the Selection Committee to clarify things such as “What does the Commonwealth mean by a voucher transmittal?” and any other requirements that might not have been clear. Then each of the four companies made their decision as to whether they would participate in the demo and of the four, AMS and PeopleSoft decided not to compete.
Two days were given for the remaining vendors to demonstrate the functionality of their software. This was to be done in front of about 150 people and the selection committee. The Selection Committee itself was composed of six different teams of two people each. Each team included two people from each functional area and two from the Commonwealth’s IT area. Other people who were invited to the demonstrations were division chiefs, bureau directors, and people at the policy level. Subject matter experts who were mainly staff or from back office operations were also identified and invited to what was becoming known internally as the ERP “bakeoff”.
To aid in evaluating the vendor “bakeoff” a scoring system was devised where the software was rated on a scale from zero to four on the desired functional requirement. So if one of the requirements wasn’t demonstrated a zero was given. If they demonstrated it but it didn’t meet the needs of the Commonwealth it was given a “one”, and so on up to a “four” for a really impressive demonstration. So a 2, 3, or 4 score on an end-to-end process meant that the software was acceptable. The goal going in was that 80 percent of the functional requirements for the nine different end-to-end processes that were identified should be met by the vendor right out of the box. After compiling the scores and converting them to a percentage, it was determined that SAP met 78 percent of the functional requirements and they were the bakeoff winner.

ERP System Integrator Selection

Having gone through the process of selecting the ERP software vendor, they applied a similar process to selecting a system integrator (SI) and their partners. With the assistance of Gartner potential integrators were asked to submit a proposal based upon the initial functional requirements. Four proposals were received from Andersen Consulting, Price Waterhouse Coopers (PWC), Deloitte Consulting, and KPMG Consulting. These were then scored by the Selection Committee and it was deemed that all four could go forward.
At this point it was decided that a Business Information Day would be held with all four vendors present. System integrators could take this opportunity to ask each of the process teams further questions based on the RFP. After this another two days were spent with the SI doing scoping exercises with the project teams using information from the RFP and the Business Information Day. They were then instructed to return with an updated proposal and improved document with more detailed scoping rules. One more day of meetings with each team was held to do a final validation on what the SIs had determined for the scope of the project. At this point, integrators were asked to complete their best and final offer document and cost proposal.
Before the Selection Committee started looking at the set of final proposals, they decided to conduct another “bakeoff” for the system integrators. However, it was determined that the basis for this bakeoff would be to include all the functional requirements from the first bakeoff that either weren’t demoed or that were rated as a one for being unsatisfactory along with other selected critical functional requirements. So scripts were written for these functions and the integrators were asked to demonstrate how these requirements could be met with SAP. They were also given scripts for some features that clearly would be very popular and asked to demonstrate how they would build “Z” tables or user exits to make SAP R/3 work better. Scoring was done in a similar way and when this was compiled it showed that a clear consensus was for KPMG Consulting (which later became BearingPoint).

Lessons for the Public Sector

The Commonwealth of Pennsylvania had never been through anything like the process of selecting vendors for their large ERP project before. The only things close to it was when they chose to standardize desktop computers for the whole state and the outsourcing of the major agency legacy mainframe systems. By taking a very careful approach to the vendor selection process they learned a lot and accomplished several important, but little discussed objectives:

  • Creating a sense of positive momentum for the project. This was important since there were many concerns about whether this project could be done on this scale in the public sector.
  • The selection process helped to build a consensus of support for ERP in the state government. Seeing the functionality demonstrated so thoroughly went a long way to showing that it could be done in government.
  • Having the vendors carefully refine the project scope with the process teams essentially created a solid working foundation for the rest of the project.
  • By identifying key domain experts for scoring the “bakeoffs” they also identified people who would be project leaders in their respective departments for the actual project.
  • What kind of setup did the software vendors bring? Did they create a special LAN to demo their system or just dial into an existing server? How sophisticated was their setup? Did they bring a Porta Potty or a tiny machine?
  • The system integrators were required to submit the resumes and length of time they would stay on project of their team members. This was used to help map out the staffing requirements for the Commonwealth ERP teams. The goal was that the Commonwealth wanted a 2:1 size advantage for their parallel project teams.
  • ERP knowledge transfer is crucial. The Commonwealth tried to judge this by comparing how well the integrators communicated what they were demonstrating to the audience. How clear were their responses to questions? Did they bring good closure to validate that ImaginePA people understood what was they were told?
  • How well did the integrators work with their team of business partners in the demonstration? When the system goes down in a demo, as it inevitably does at some point, do they point fingers or try to work together to get it back up? Can you identify the different business partners on the team or do they work together like seamless teams?

Clearly, because there is so little experience with ERP in the public sector, much more care needs to be taken in selecting the best vendors. The innovative approach that the Commonwealth of Pennsylvania took in having vendors present in a “bakeoff” contest helped to better define the state’s processes, generated a core group of supporters among key departmental decision-makers, and created an air of excitement among the state employees. This was to prove invaluable as the state went live with the different functions throughout the state in the following years.

Vikram Pant
Vikram Pant
Ph.D. William P. Wagner
Ph.D. William P. Wagner