Doug Hadden, VP Products
I recently encountered the notion of “tried and true” related to the use of ERP software in government. This notion contrasts with reality. That’s why I’ve been making the case that Government Resource Planning (GRP) software designed for government is less risky and has a lower Total Cost of Ownership (TCO) than software originally designed for the private sector, in government. More evidence comes from an analysis of the US Government Accountability Office (GAO) report on ERP implementations in the US Department of Defense from March 30th. Chris Kanaracus of IDG News reported on April 2nd:
“Public-sector ERP (enterprise resource planning) software projects historically have experienced some of the industry’s most dramatic cost overruns and delays, a fact that a new report by the U.S. Government Accountability Office brings into sharp relief.”
Low Risk
The GAO report makes for some interesting analysis. Further investigation reveals what ERP software was used and what systems integrator won the contract. The ERP software comes from the two leading vendors. The HR software comes from one of those leading vendors and is the leading HR package. All of the integrators are well-known with established methodologies. In other words, “tried and true”. Low risk. The results show anything but.
I’ve had to make some assumptions because there is missing information on some of the 11 projects. Some of them do not have new “actual” estimated finish date or budget. And, when something is meant to finish during a fiscal year – when is that? I assumed fiscal year as mid fiscal year. Cost overruns of 10% if not documented and time overruns of 18 months from the estimated date or 24 months should the estimated date have already passed.
Nevertheless, the results show the following:
- Average time overrun was 43% and budget overrun was 95% which means averages more than 2.5 years and $607 Million.
- Overall figures are somewhat skewed by a specific project
- The HR implementations seem to be meeting objectives better, but it should be noted that it is essentially the same implementation adapted for three services
Military Complexity?
Military implementations are more complex than many in the private sector. And, non-military in the public sector. So, what does this tell us about risk and ERP failure in the public sector?
Yet, the primary objective for these projects was to replace legacy software with Commercial-off-the-Shelf (COTS) and adapting processes to suit the “best practices” of ERP. This isn’t about inventing stealth aircraft. There isn’t a high degree of uncertainty or innovation here.
And, the timelines of these projects were quite long to begin with. As were the budgets that ranged from $209 Million to $3.9 Billion!
Could it be worse than reported?
Michael Krigsman, the CEO of Asuret is quoted in the IDG article:
‘Krigsman cautioned that the report’s figures may not tell the whole story, noting that in a number of cases, the reported expenditures on a project don’t seem to match up in a logical way with the timelines. “Clearly, detailed audits should be performed to untangle the details and get at the truth,” he said.’
Why failure?
It is frequent that the victim is blamed for ERP failures. I don’t think that the Department of Defense can be blamed for failures in 10 of the 11 cases. My research into the projects and the GAO report suggests that the ERP software is a major contributor to these problems:
- Flexibility: Adapting private sector processes to the public sector that has required more code customization than anticipated
- Integration: Integrating with existing software which is an ironic problem given that ERP vendors spout so much about integration
- Complexity: Dealing with complex ERP software for configuration and customization.
- Legacy: Dealing with proprietary development languages and infrastructure, much of it based on legacy client/server technology.
This puts even more doubt into the business case of ERP in government. ERP failure in government has become an open secret. It’s been tried and truly – rarely succeeds.
I made an observation about risk at presentation last year about if ERP risk management. It is almost a certainty that one of the following will be true of an ERP government implementation:
- It will not achieve the planned benefits or the scope will be reduced
- It will take longer than the plan to implement
- It will cost more to implement than planned
- It will cost more to maintain than planned