CCSD’s systemic problem and its expensive consequences

Part 3: Internal CCSD memo admitted ERP failure 'directly related to lack of proper management'

Steven Miller

“The 17-year saga of CCSD’s highly expensive ERP adventure — where the estimated cost … ballooned from $15 million in 2001 to well over 10 times that amount today,” is how Part One of this series began.

Such runaway spending shows that Clark County School District — fifth-largest district in the nation — is no exception to the waste, incompetence and abuse that, as Part Two showed, plagues America’s other large districts.

However, for readers and policymakers to fully grasp what went wrong at CCSD requires first understanding what the district was attempting.

Which brings up the question: What in the world is an “ERP”?

ERP stands for Enterprise Resource Planning, the corporate term of art for mainframe-based software systems that integrate the internal business-data operations of major enterprises, such as large corporations and government agencies.

When data from core divisions — usually purchasing, accounting, payroll and human resources — can be digitally organized and combined, enterprise leaders can get the critical reports they need much more quickly and accurately than through older, unintegrated, legacy systems.

A primary source of the new organizational efficiency that ERP systems can bring — and a major reason why firms install the systems — are the “best business practices” that are integrated into the software. Thus, a key part of an organization’s successful transition to its new ERP system is an initial process of re-engineering, or upgrading, the organization’s internal business practices to reflect the state-of-the-art procedures built into the software.

Not only do these “best business practices” facilitate quicker and more accurate decision-making. They also provide organizations with significant protections against internal fraud, theft and waste.

Of course, the “re-engineering” of an organization’s work processes also, quite frequently, elicits in-house resistance. And private- and public-sector organizations differ as to their ability to meet that resistance.

Usually, private-sector organizations go to ERP systems, say industry specialists, in order to better meet customer demand, cut unnecessary costs and meet the competitive challenge from rival firms.

For CCSD, however — a government monopoly facing little serious competition — those sorts of motivation were not primary, district school board records show. Rather, its interest in enterprise-resource planning largely reflected the repeated prodding and warnings by the outside auditors that state law requires the district to employ.

When an organization seriously confronts the real challenges involved in ERP evaluation and implementation, says the Colorado specialty consulting firm Panorama, “ERP software can revitalize an organization by streamlining and synchronizing its separate departments into one unified software system.”

However, also says the company, “When these tasks are done without focus and solid planning, the ERP software project can have severe repercussions,” negatively affecting the organization for years afterward.

The public record shows such has clearly been the case with the Clark County School District.

The 2007 district memo

A closely held internal district memo from 2007 — never generally distributed but recently provided Nevada Journal — summed up the situation at the end of the first six years:

[T]he project has not been well managed. The delays in the project and the cost overruns can be directly related to the lack of proper management.

According to the memo, the major causes of the delays and the resulting additional costs to the project were, as of 2007:

  • CCSD administrators failed to grasp the large inherent risks and challenges of ERP projects. District management declined to assign staff to the project full-time, not appreciating the need of CCSD staff assigned to the complex and demanding project to be able to focus entirely upon it. Staff members were required to work on the project only part-time, while also continuing to do their previous, pre-project CCSD jobs. The lack of staff time spent on the project contributed significantly to its failure to meet its original, and then even its revised “go-live” dates. It also meant much work that could have been done by staff was routed to the highly expensive SAP consultants.July 1, 2006 had been the original go-live date for the Finance and Purchasing modules, before being pushed forward to first October 16, 2006, then November 20, 2006, and then to “a date to be determined.” The Human Resources and Payroll modules were initially supposed to go live in December 2006, but after that date was missed, implementation of both modules was abandoned, being given no new go-live dates. Those two modules, constituting Phase Two of the project, remain uncompleted to this day.
  • No plan. Neither CCSD nor SAP ERP project managers ever developed a detailed project plan, to which they would then manage. “Failure to have a plan,” said the memo, “resulted in the project managers never knowing whether they were ahead or behind schedule or whether they had sufficient resources assigned to the project. There was also a lack of decision-making by the CCSD project manager, and even when decisions were made, they were often changed when the next person walked through the door and didn’t like the decision the project manager had made.”
  • Lack of urgency. Through 2006, project directors — initially CFO Walt Rulffes and then the late Phil Brody, assistant superintendent and chief technology officer — failed to see the importance of meeting deadlines, Brody stating several times that “there are deadlines and then there are deadlines.” This lack of urgency among top management led to a lack of urgency throughout the whole project team.
  • Indifference to costs. Neither the project manager nor project sponsor tracked costs against a budget. “There seem[s] to be a feeling that money was unlimited [and] no concern for what additional costs were being incurred.”
  • The mainframe hardware necessary to house the ERP software was put in place only after significant delay. CCSD had purchased and received the hardware in the fall of 2005. To meet the agreed-upon July 2006 go-live date, the hardware needed be in place and operational by November 2005 at the latest. Although it appears that both Rulffes and Brody knew in June 2005 that a significant upgrade of power at the installation site would be required to run the mainframe, that condition was not met until April 2006 — resulting in at least a five-month delay in configuring the software for implementation.
  • CCSD’s project leaders were content to move slowly. According to the internal memo, when the problem with the hardware setup had been brought to the attention of then-project-manager Greg Halopoff and project sponsor Brody, the response given had been “there are deadlines and then there are deadlines,” and that “They are working on the issue and not to worry about any deadlines.”
  • CCSD management allowed a single department head to repeatedly revise the project. Against the advice of software vendor SAP, top CCSD management elected to go with a “big bang” implementation for the Materials Management (i.e., Purchasing) module, in the first part of Phase One, even though CCSD and SAP had previously agreed that phase would only implement the core Finance and Purchasing modules. However, CCSD’s in-house purchasing department then demanded, and got approval for, an immediate, simultaneous, implementation of SAP’s Supply Chain Management (SCM) module. The justification given for the “big bang” implementation for CCSD’s purchasing department was that something “fuzzy” was needed for district Principals. And the “fuzzy” thing was the Shopping Cart application included in the SCM module.During implementation of the SCM module, constant changes were demanded and approved to customize the configuration. Because of those changes, the module was not available for integration test cycles #1 and #2. Even during integration testing cycle #3, changes were still being made to the SCM module, and as late as January 2007, noted the memo, changes were still being made — causing additional delays to the project and added costs.
  • Project in chaos. CCSD and SAP project managers, failing to control changes, were not following the change-order procedures that previously had been outlined in the district’s contractual agreements with SAP. “The project manager had lost all control of the project,” said the memo, “was not following a project timeline, was not managing the budget, and was unable to give status reports and estimated dates for a delivery of the product.”

In mid-January 2007, Halopoff was taken off the project and replaced as manager by CCSD Assistant Superintendent for Business Keith Bradford — who would himself leave CCSD at the end of 2008.

History of the cost estimates

According to the memo, by the end of 2006 CCSD had spent or encumbered $25 million, while another $2 million was going out each month to SAP consultants whose hourly pay ranged from $202 to $317, plus expenses.

In 2001, the initial estimated cost of the project had been $15 million. A later CCSD estimate given eSchoolNews.com — in 2003 or early 2004 — had been $25 million. A CCSD spokesman told the web publication that the district would be avoiding the high costs and delays that other organizations had run into in their ERP implementations, because CCSD would have the benefit of “learning from others’ failures.”

However, by December 2004, when the district actually signed its first agreement with SAP, total projected costs were now put at $33 million, according to an April 12, 2007, Las Vegas Review-Journal report.

That story was based a status report given by Bradford, the day before, to a CCSD board of trustees work session. That report estimated that project costs by the end of that year would total $29.3 million, and by the end of 2008, would total $35.7 million. Total costs of bringing the entire project to completion, Bradford estimated, would be approximately $46.25 million.

Eleven months later in March 2008, however, CCSD officials told the Las Vegas Sun that the project, after spending $35 million, was out of money and was being placed in “hibernation” with Phase Two — Payroll and Human Relations — still not operational.

Seven years after Walt Rulffes had told district trustees that CCSD needed to begin investigating system specifications for an ERP system, the Great Recession had arrived in Nevada. And then-gov. Jim Gibbons’ administration was pressing CCSD to cut its spending by five percent.

“It would have been nice to see the project through this year,” district CFO Jeff Weiler told the Las Vegas Sun. “But, given that the district is struggling to comply with Gov. Jim Gibbons’ call for it to cut $66 million from its budget over the biennium, postponing the computer system “is the prudent thing to do.”

CCSD’s final amended budget for 2008 was $2.07 billion.

“So far only two departments, finance and purchasing, have switched to the new system,” wrote Sun reporter Emily Richardson. “For the time being, the payroll, human resources and teacher recruiting departments will continue to use older computer systems.”

That “time being” has now lasted for a decade.

Although Weiler and other CCSD officials continued to use Bradford’s 2007 estimate of final costs when they were speaking to the Sun in 2008, better informed estimates today place the eventual final total much higher — at well over $100 million.

For one thing, multiyear records of CCSD spending warrants reviewed by Nevada Journal show that after the district had announced the exhaustion of the initial $35 million, another $32.6 million at least has gone into the project. That would bring the total amount spent by the district on merely the first two modules of four, or Phase One of a two-phase project, to at least $67.6 million.

Moreover, even that $67.6 million figure may significantly underestimate the district’s ultimate costs, for several reasons:

  • Phase II, which was to implement ERP capability for CCSD’s Human Relations and Payroll departments, was estimated in discussions with district trustees to constitute as much as 60 percent of the entire ERP project. Thus, if 40 percent of project costs translated into $67.6 million, 60 percent, for Phase II, could be as high as $101.4 million. And total costs for just getting the entire system off the ground could then arguably be as high as $169 million.
  • Although CCSD committed, in its contract with SAP, to “use the current generally available ‘GA’ version of the SAP software,” and to “adapt its business processes as necessary so that the system will not require modification of the base SAP software code,” and agreed that “CCSD is willing to implement the necessary business process changes in order to follow the SAP best business practices approach subject to applicable state or local laws and regulations,” the district nevertheless sought extensive — and expensive — customization of the software.Elaborate customization of ERP software, industry experts agree, is nearly always a serious mistake. One problem is that whenever new updates or upgrades to the basic ERP software are issued, the organization with a highly customized system now has to make sure all those customizations will work with the new upgrades. Thus every time updates are issued, the organization can face significant new costs from highly paid programming consultants, as well as in-house delays in core departments, while the consultants tinker with the custom code.
  • Another cost issue for the district is that at the very start it had selected the industry’s most expensive ERP software provider, SAP.Ken A. Forrest, a highly experienced industry consultant who has worked on ERP projects for multiple school systems, including not only CCSD, but also SAP projects at Los Angeles Unified and Detroit, argues that SAP is actually too expensive for nearly all K-12 districts.“K-12 just isn’t funded at the appropriate level to afford that Cadillac of a program,” he told Nevada Journal. “I mean, SAP will do anything you want it to do, and I mean that in sincerity. But you have to have the money to make it do it.“This system is expensive, it is the world’s largest ERP system, and it does an outstanding job if you have the resources to fund it and staff it. I don’t believe that K-12 districts have those resources at the appropriate level…“Obviously, General Motors, Walt Disney Corporation, Warner Bros. — they’re all doing great with it… If you talk to the people who work with it, it does a lot of really cool stuff. But you have to have the technical expertise and the money to get it there…

    “I’ve seen ERP systems for a while. So could Clark County [School District] have made it work? Yes, they probably could have. Did they have the necessary resources and personnel in place to do it? I didn’t think so.”

Coming in Part Four: CCSD’s ERP failure is actually par for the course for government entities

Links to the entire eight-part investigative series

———-

Steven Miller is managing editor of Nevada Journal and senior vice president at the Nevada Policy Research Institute

 

Steven Miller

Senior Vice President, Nevada Journal Managing Editor

Steven Miller is Nevada Journal Managing Editor, Emeritus, and has been with the Institute since 1997.

Steven graduated cum laude with a B.A. in Philosophy from Claremont Men’s College (now Claremont McKenna). Before joining NPRI, Steven worked as a news reporter in California and Nevada, and a political cartoonist in Nevada, Hawaii and North Carolina. For 10 years he ran a successful commercial illustration studio in New York City, then for five years worked at First Boston Credit Suisse in New York as a technical analyst. After returning to Nevada in 1991, Steven worked as an investigative reporter before joining NPRI.

News You Can Use from Nevada Journal

Editors: Reprint permission, in whole or in part, is granted under a Creative Commons Attribution-ShareAlike 3.0 license.

Nevada Journal, a member of the Nevada Press Association (NPA) and Investigative Reporters & Editors (IRE), is an independent nonprofit reporting effort that adheres to the SPJ standards of professional journalism and specializes in in-depth and investigative journalism.

For the last 20 years, Nevada Journal has been published by the Nevada Policy Research Institute, a non-partisan public-policy think tank.