If you are looking for a case study on how to guarantee a project failure then you only need to look to the Obamacare website failures and the testimonies of the four contractors on Capitol Hill today.
I have a PMP (Project Management Professional) certification and have managed many successful (and not so successful projects) in my career and know what it takes to deliver a successful project. Allow me to provide you a list of top 5 things Project Managers (PM) can do to ensure their projects fail miserably and I’ll use examples from today’s testimony to hammer my points home.
1 – Lack of Testing
As I covered in a previous post, a primary rule in product/software development is to ‘Fail Fast’ which means that you have robust testing and toll gates in the initial design phase of the project to detect issues early instead of later when changes are much more costly.
That didn’t happen with the Obamacare website.
“In their first public remarks since the debut of the problem-ridden insurance exchange, executives of the main IT companies told members of the House Energy and Commerce Committee that full tests of the Web site that should have been carried out months in advance, but began just two weeks before its rollout.”
2 – Ignoring Project Team Members
It is disastrous for a project if a culture is created by the PM where team members are either discouraged from bringing up issues or are never asked for their input.
It appears the contractors’ opinions were not solicited and they knew the website would fail.
“Top Obamacare contractors said Thursday they never recommended that the Obama administration delay the Oct. 1 launch of HealthCare.gov — even though some of them harbored doubts about a website that would crash shortly after it went live.”
3 – Poor Communication
It appears that the PM of this debacle worked for the Centers for Medicare and Medicaid Services (CMS) and he or she has apparently never managed a project before.
The PM’s communication was so poor that nobody can identify who the actual PM was!
“It was not our decision to go live,” Campbell said. “It was [CMS’s] decision.”
“Republican requests for who exactly at CMS should be held accountable for the bungled rollout were mostly rebuffed. The contractors repeatedly said they couldn’t provide specific agency employees who made key decisions leading up the launch of a faulty website, though they eventually offered two names – Deputy CIO Henry Chao and COO Michelle Snyder.”
Even Jay Carney has no clue who was in charge of this project:
“Asked Thursday who was the “quarterback” for the mistakes, White House spokesman Jay Carney declined to name names. “This is on us. And that goes from the president on down,” he said.”
4 – Scope Creep
It is critical during the project initiation/planning phase to lock the design requirements and scope. It is all too tempting to add a little here and a little there during the project and before you know it the scope has been drastically altered which makes it impossible to meet the cost, functionality and schedule targets of the original project plan.
Not only does it appear the Obamacare website project suffered from scope creep but also what I’d call scope skyrocketing late in the project.
“An executive of a UnitedHealth Group Inc. unit said at the hearing that his company learned of one key change—disabling a feature allowing users to browse for plans anonymously—about 10 days before the website went live.”
“We don’t know who made the decision, when it was made, or why it was made,” said Andrew Slavitt, group executive vice president of UnitedHealth’s Optum unit.”
5 – Absence of a Team Culture
A large project like this requires many smaller teams to work on smaller pieces of the project but they must still view themselves as part of one big team. They must not work in a bubble but instead be cognizant of how their pieces of the project fit with the rest of the project team members’ pieces.
Of course it’s the job of the PM to set this culture at the time the project kicks off but this apparently didn’t happen.
“No one in the government made sure the many complex parts of the federal health-insurance website worked together properly..”
“The contractors said each of their pieces worked more or less as intended, but the HealthCare.gov website was nearly paralyzed when they were bolted together. The federal agency overseeing the site also took on the job of integrating the many parts of the system—an unusual arrangement for such a complex project.”
I have no doubt that there are excellent PM’s that work for the Federal Government in such departments as NASA and the Department of Transportation since they have clearly demonstrated they can deliver on project targets/scope. But with the HHS and the CMS it appears there is nobody there who has experience running projects on the scale required for the Obamacare website.
And if they are incapable of running a project that builds a website (which is done successfully every day in the private sector) what makes you think they’ll be better at making decisions for your healthcare?
Oh, one more thing. How much did we pay for this failure? About $483 million.
“Contractors also offered updated price tags for their parts of the project. According to the executives, CGI has been paid $112 million, a sum that would likely go up to $196 million by year’s end. QSSI’s bill would be close to $85 million by year’s end. Equifax has so far collected $2 million, and Serco’s contract was for $200 million, with about $30 million already collected.”
Isn’t it easy to spend money not your own and make mistakes at the tax payers expense? Just sayin’
Maybe Obama can call for a tech version of the Bowles-Simpson committee, and NOT listen to their recommendations either.
Not usually a proponent of conspiracy theories but it almost seems like this website was meant to fail. How could anyone at CMS and the Obama administration be this incompetent? Actually, there would need to be MANY who were incompetent.
It is not at all unusual for a large organization (corporation) to engage a large contractor to implement a huge software system without actually getting closer to the implementation than the weekly or monthly status reports delivered by contractor managers. These managers are seldom very close to the implementation either. Testing is a finicky engineering operation and has to be conducted with cooperation among the various development teams. Test plans are part of the early stages of design. I am pretty sure nothing of the sort was part of this project.
The teams were probably aware of the problems in integrating quite early in the testing. But who listens to technical people? I am reminded again of the Challenger disaster and the engineer who tried to warn management, in vain. http://www.nasa.gov/centers/langley/news/researchernews/rn_Colloquium1012.html
And why the hell is there not a professional IT department in the government?
Yep plenty of fail on this one. But the Government does have an IT department….It’s called the NSA!
Very funny… I wonder who is their outsourcing contractor? Bet it isn’t a bunch of guys from India. The United States can afford to hire its own development teams, and supplement them with consultants when absolutely necessary.
Apropos of nothing, I started my IT career at Sperry, which was the contractor for Polaris/Poseidon. I only did accounting software, so I had that low level of security clearance. The software guys for the guidance systems were classic geeks, shorts and sandals, and they were in their own well-guarded part of the building in Lake Success.
In addition to these developers, who were responsible for the product that was being prepared for the U.S. Navy (who were also in the building, fully armed), we had an IT department, the one I belonged to. When I was hired, I replaced a really terrible consultant. I know because I had to fix his code.
I am not a big fan of social networks. I have a Facebook account in my high school name, Sandy Perlmutter, which is almost entirely political. I don’t collect “friends”. My professional name is Sandra Greer, which is what I use on LinkedIn. I use my Facebook login to write on the occasional medical, technical, or political site, such as The Reg. Mostly I read articles. I have never been very interested in Twitter.