The UK National Offender Management Information System project (called C-NOMIS) failed amid scathing attacks, accusations of mismanagement, and vast budget overruns. The project offers an excellent case study relating failure directly to inadequate governance and oversight.
The project was supposed to create a single database allowing UK prison authorities to track and manage offenders while they are in custody and following their release. After a three-year delay and doubling of costs, authorities abandoned the critical, single database concept.
A National Audit Office (NAO) analysis of this project concludes:
Overall the C-NOMIS project was handled badly and the value for money achieved by the project was poor. Many of the causes of the delays and cost overruns could have been avoided with better management of well known issues. NOMS’ failure at the start to appreciate the product customization and business change required, its inadequate oversight of the project and weak relationships with suppliers led to a doubling in program costs, a three year delay in program roll-out and reductions in scope and benefits. In particular the core aim of a shared database to provide a single offender record accessible by all service providers will not be met....
The NAO press release describes the cost overruns on this billion-dollar project. [Note: these figures are in Pounds Sterling; at current exchange rates, the dollar cost is about 50% higher]:
The project to provide an IT system to support a new way of working with offenders was to be introduced by January 2008, and had an approved lifetime cost of £234 million to 2020. By July 2007, £155 million had been spent on the project, it was two years behind schedule, and estimated lifetime project costs had risen to £690 million.
In January 2008, the National Offender Management Service began work on a rescoped program with an estimated lifetime cost of £513 million and a delivery date of March 2011.
Quoted by the BBC, Public Accounts Committee chair, Edward Leigh MP, called the project a "spectacular failure:"
"The central goal at the outset of the initiative was to have, by January 2008, a single, coordinated IT system for managing offenders across the prison and probation services," he said. "What they delivered was a master class in sloppy project management."
According to Computer Weekly, Leigh also spoke about "kindergarten [project management] mistakes:”
This Committee hears of troubled government projects all too frequently. But the litany of failings in this case are in a class of their own.
All of this mess could have been avoided if good practice in project management had been followed. A new project team has been brought in. They cannot afford to repeat these kindergarten mistakes.
The Mail reports that Prisons Minister, David Hanson, tried to put a good face on this ugly failure:
As soon as the extent of the projected costs and delays to the C-NOMIS project were recognized, we took immediate steps to halt the project and consider the most cost-effective way forward – which effectively preserved the work done to date. 'The revised program builds on this work, and steps have been taken to ensure the program remains on time and in budget.'
THE PROJECT FAILURES ANALYSIS
The NAO summary findings describe a broad set of project management and governance problems leading to this failure. The full report contains a wealth of detailed information for anyone interested in the relationship between project governance and IT failures.
The report enumerates key reasons for delays and cost increases:
There was inadequate oversight by senior management. Whilst the Project Board met at least once every two months, it did not actively monitor delivery of the project and was unaware of the full extent of delays or the implications of decisions it made upon project cost.
NOMS did not put the appropriate resources and structures in place to deliver such a complex project. The overall governance and resources applied were not adequate given the scale of the task. Roles and responsibilities were blurred, in particular financial accountability was unclear, and insufficient skilled resource was applied to the project.
Program management was poor in key aspects, including planning, financial monitoring and change. NOMS initial planning was overly optimistic in terms of both cost and timescales. For example there was no contingency, despite some recognition that the project carried a high level of risk. Budget monitoring was absent, with cost control focusing on monitoring the spending against the annual budget rather than matching cost against project deliverables. Change control was weak, and there was no process in place for assessing the cumulative impact of individual change requests on the project budget or delivery timetable.
NOMS significantly underestimated the technical complexity of the project. A single offender database is technically realizable, but NOMS did not adequately explore other potential solutions and underestimated the cost of customizing the software it had already selected for the Prison Service. [T]he estimated cost of developing the application rose from £99 million when the full business case was approved in June 2005 to £254 million by July 2007, primarily because of customization.
NOMS underestimated the need to invest in business change alongside the IT system. There are 42 probation areas in England and Wales, each with their own ways of working. NOMS, however, made no sustained effort to simplify and standardize business processes across prisons and probation areas, and did not allocate resource for this purpose. At the outset, NOMS treated C-NOMIS as an IT project rather than a major IT-enabled business change program.
NOMS’ contractual arrangements with its key suppliers were weak and its supplier management poor. Instead of tendering key project contracts, NOMS opted to use its current suppliers under existing framework agreements to develop and deliver the C-NOMIS application. NOMS allowed these contracts to go forward on a time and materials basis for longer than it should, which meant that there was insufficient pressure on suppliers to deliver to time and cost.
The report adds that overseers also managed financial controls poorly:
We have not been able to determine the full value of the waste and inefficiencies associated with the failure of the C-NOMIS project with certainty because of NOMS’ poor recording of costs.
The NOMS report also includes a table comparing the project against the official Office of Government Commerce (OGC) list of "Common Causes of Project failure." The chart summarizes how poor judgment and mismanagement contributed to this failure.
Click the image to see a full-size version:
In addition to all the other problems, Computer Weekly explains that project participants hid problems from responsible stakeholders for several years:
The project began in June 2004. But it was not until the summer of 2007 that senior officials and ministers in the newly-formed Ministry of Justice discovered that C-Nomis was running two years late and that costs had more than doubled....
It's a similar story to the Rural Payment Agency's IT-based Single Payment Scheme in which ministers were given reassuring reports on progress until it was too late to hide the seriousness of the project's difficulties.
In a remarkable instance of IT failure déjà vu, here's how I summarized the Rural Payments agency failure mentioned in the previous quote:
In general, the entire situation represents poor planning and project management taken to new heights of incompetence. [The system was] designed and executed based on poor practices, lack of experience, and world-class levels of bad planning.
My take. From financial irresponsibility to extensive software customization, this failure has it all. This case offers a textbook example illustrating multiple layers of mismanagement in fundamental areas such as procurement; governance and oversight; technical and functional analysis; business process transformation; and financial controlling.
Note: students of project governance should pay particular attention to the full project report, the Office of Government Commerce's project management website, and the NAO/OCG list of common causes of project failure. Many of the issues described in these resources are not IT-specific, and are therefore highly applicable to managing non-computer projects such as construction.
[Photo from Her Majesty's Prison Service.]