OregonLive’s Cover Oregon: Health exchange mired in unexpected delays; what went wrong? tries to explain what went wrong with the Cover Oregon website launch. But really, you could do a postmortem on most government web projects and you’d have the same bullet list – changing specifications, out of control budgeting and a lack of real comprehension of real world implementation.The Oregon government is run by bureaucrats who project manage the actual work of implementing a tech project – just like the Federal government. Any developer can tell you (and for that matter, any government bureaucrat who is actually doing their job) that the bane of meeting goals are unexpected changes to requirements. Even with clearly defined goals and no changes to specifications, its hard enough to predict when a project will be accomplished. As Nick Bundick reports, the site was technically operational in January 2013, but then had to accommodate a laundry list of changes afterwards.
Changes cost extra. Always. Contractors know that clients always make changes along the way. But you cannot predict what those changes are going to be, so all changes are carefully metered, and changes then mean original goals are no longer feasible. When a change happens, it not only costs the actual cost of getting the work done, but also the cost of the bureaucracy around it and costs that can occur because of planned dependencies. California based Oracle Corporation – a company known for its extremely high cost services and database technologies – probably did its best, but its going to charge extra. And Oracle did honestly state that this is a 4 – 5 year project – but the State of Oregon wanted it delivered in half the time. Oracle is happy to accommodate a paying customer though, and doesn’t necessarily admit the honest truth that more money thrown at a project does not equate exactly to reducing the completion type by the same proportions – doubling your budget doesn’t necessarily mean its going to be done in half the time.
Oh yes – you can also be guaranteed that Oracle owned technologies were used (they are the largest database technology company in the world and also own Java), and will require regular maintenance and upgrades. Over the next few years after Cover Oregon is finally ready to go, you will have officials dealing with the uncomfortable truth that maintaining the system is going to be a lot more expensive than they originally thought.
Your government is also trying to tell you that it can move quickly from implementation to real world use in a short amount of time. It never works that way. Any technology project requires testing, and Cover Oregon is going to be a worst case example of this because the rules of insurance are going to be constantly changing. Quoted:
“Oregon had an overly optimistic expectation that exchange planning and the initiation of the corporation’s operations could be accomplished within the (funding) parameters of the original grant request,” officials wrote in an application. “The state underestimated the work, resources and expenditures needed to obtain federal certification in 2013.”
This was a complex project, but this quote could be applied to almost any government technology project of consequence. Government projects only get green-lit if there is a plan and a budget, who takes the fall when plans simply unrealistic and end up costing multiples of the original planned expenditure? Not the planners. Once a project begins and begins to snowball, you either cancel it (and then a planner takes a hit) or you let it go and hope money is found elsewhere. That money is public money, and Oracle is going to get paid for its work. Or another contractor comes in, spends more time learning the project and then completes it – at an even higher price.
But lets be realistic about this. People are already complaining about Cover Oregon being late. Any tech savvy person could have told you that. If they are a few weeks late – that’s not unpredictable even if everyone involved in the project is highly competent, so if Cover Oregon is only a few weeks late, that’s the best you can expect. But more than that, you can thank and blame the competency of the planners.