On a regular basis, I receive calls from reporters and researchers writing about IT failures. Most often, they want context on how often these problems occur and who is to blame. However, recently, I got a different kind of call.
Todd Wallack, a reporter for the Boston Globe, got in touch with a strange question. He had identified a Massachusetts IT project that has been running for 19 years and is still not complete. Wallack wanted to know whether this is a common situation.
Hearing about a 19-year old IT project, I was stunned. Durations of this magnitude sometimes occur in the federal government and on statewide projects in large states like California. But, it just seemed implausible here in Massachusetts, despite the state's history of IT failures.
Nonetheless, Wallack's research uncovered an IT failure unicorn -- one of those mythical, endless project beasts that sometimes turn out to be real.
Here are the specifics, taken from Wallack's front-page article in the Boston Globe:
MassCourts "is intended to link more than 100 courthouses across the Commonwealth, for the first time allowing court officials to look up information about cases anywhere in the state. The project started in 1996, when the state legislature approved a $75 million bond to fund the work. The state selected "Deloitte & Touche in 1997 to construct a single system to track data from the roughly 1 million cases filed a year throughout the state, replacing 14 separate systems used by various departments."
Although the initial budget was $75 million, the state spent most of that money during the project's early years. Virtually the entire amount has been spent or committed to external vendors. Despite this expenditure, the final deliverables will include a limited set of features relative to other comparable state systems.
In a departure from standard project practice, the state paid for project activities conducted by internal staff, such as software installation and training, out of the regular budget for salaried employees. Doing so understates the actual project costs and makes accurate cost accounting difficult. In addition, MassCourts used its operating budget to spend over $3 million on maintenance and support. Again, this obscures the real total by not identifying the amounts paid as part of a single, overall project budget.
Although the state justifies this form of financing, officials are unable to specify the total funds spent to date on MassCourts accurately. From my perspective, the lack of standard project accounting indicates poor governance. This view is consistent with a 2001 audit report from the Massachusetts legislature, stating the project did not conform to project planning and tracking requirements mandated by state regulations.
The Boston Globe article includes a link to comments from Deloitte, defending its work on the project:
the Court issued dozens of work orders to Deloitte; all of them were completed successfully
Deloitte claimed similar success on other Massachusetts state projects that observers called failures. During hearings on the failed unemployment system project where Deloitte was contractor, for example, the company declared, "We fully believe that the system is a success." I strongly disagree with Deloitte's self-assessment and testified at a subsequent Massachusetts legislature hearing: "On which planet can this project be considered a success?"
To understand how Deloitte can claim success on projects that are late, over-budget, and do not meet user needs, we must consider the original contract. Too often, these agreements specify billing based on process milestones rather than customer business outcomes. For example, a billing milestone might be "completion of software testing." If Deloitte completes the specified test, then it can declare success and move on. However, without clear links back to software usability and fitness for purpose, passing the test does not indicate whether the software will meet business needs.
These contracts decouple work and payment from practical success. In consequence, Deloitte can claim positive results from a contractual and legal point of view, despite obvious flaws in the business outcome. In fairness, the client has a significant role to play as well; however, this does not change the one-sided nature of such contracts.
In summary, MassCourts is a failed project with the following attributes:
- Ongoing for nineteen years and still not complete
- Actual project expenditures unclear
- Limited feature set
- Contractor claims project success
What do you think about this project?