I swear, if I am ever called to Capital Hill – as some of my Lockton teammates have been to get our insight into the employer impacts of Healthcare Reform– I am soooo calling in sick. I may be a big, badass blogger filled with fire and fury, but I’m not sure I have the chutzpah to tell the United States of America the failed Open Enrollment go-live was all their fault! Can’t they still hang folks in the public square?
Well, if you weren’t invited, let me fill you in on some great conflict between the contractors and our congress that happened Thursday. Basically, the contractors laid the blame at the feet of the government, particularly the Centers for Medicaid and Medicare Services (CMS). If you’d like to learn more about the testimony and requisite partisan bickering, check out some of the news coverage. For anyone in our industry though, it’s imperative to learn the vendor management lessons way too many of our tax dollars have laid bare. We’ll expand on these topics in future postings, but here are some highlights:
1) Vendor Management is Vital
There’s a saying in football from Skip Bayless: “When you have two quarterbacks, you have none.” Our version of this should to be: When you have multiple vendors owning the implementation, you have NO vendor owning the implementation.
There were several separate companies all working together to erect the website for enrollment into Obamacare. Of course, each said their company and their portion of the installation was well done. But somehow when the parts became a whole, the whole thing fell apart. (Side note: that last sentence has to be the best I have ever written! I am so calling Mizzou’s J-School to brag.)
We believe strongly that “go-lives” work just like building a house. There can be multiple contractors, but there has to be only one, truly accountable General Contractor, who then manages the multiple Sub-Contractors.
We manage our HR Technology practice using a project management tool called a “RACI.” It stands for Responsible, Accountable, Consult and Inform. One of the rules we follow is that only one person on our team can be accountable – otherwise, there’s no accountability as pointing fingers and he said/she said immediately become options.
2) Limit Last Minute Meddling
If you’re a betting person, I have a “tell” to watch for that will lead to 100% accuracy as to which implementations fail: Last minute changes.
It sounds as if CMS ordered a change to how our fellow citizens registered to the site within two weeks of the go-live launch! (!^@#$@%@%) There has to be a sign-off line past which no changes are made. That gives the vendor time and space to do their work. The client can always add changes to later iterations – actually this is becoming the norm in the software development industry – but not trying to take out the chocolate chips from the cookies right before they’re supposed to come out of the oven!
There’s an arrogance on display here. It’s easy for bureaucrats – and truth be told, business leaders like me are also super susceptible to this as well – to order changes when we don’t have to feel the impact of our requests. I talk to HR leaders every week who confide in me how their executives have no idea how difficult it is to produce the data they request. Our profession masks effort from those who request it, which teaches them that changes are possible at any time!
3) Practice as you’ll Perform
It appears – according to testimony by at least two of the contractors – that the testing was performed with no where close to the volume that the production system was forced to handle on the first day. It’s so crucial to build (and test) systems as they’ll actually work in the real world – not some sanitized, safe sandbox where nothing goes bump in the night.
Chad Murray, the poor soul who’s trying to teach me how to play the piano, is always pushing me to practice my lessons at – or past – the speed at which I’ll eventually perform them at. His logic – which fully transfers to our industry – is that if I can handle the rhythm, notes and fingerings at 120%, I can always dial it back down safely.
I have no idea if someone didn’t estimate the response to the Affordable Care Act – but knowing that it offered health insurance to 47 million Americans or so who had not had access to it in the past – it is incompetent to say that 500,000 folks hitting the site in the first couple weeks is “higher volume than expected.” You’re telling me that someone didn’t think that folks with a pre-existing condition who’ve had to face medical bills naked for the last however many years are going to decide on their own to peacefully wait a couple weeks before signing up for budget-saving insurance. Sure we all know the plans don’t go into effect until the next plan year, but no one’s going to be thinking – “hey, I’ll wait a while before signing up.”
Volumes have to be estimated well in advance. If you’re picking implementers for your project, please learn a lesson from the government – find the most conservative, pessimistic, realistic, tried-and-true implementation experts. This is a business where experience matters. Take whatever volumes you estimate and double them for testing! Failure of these mission-critical platforms is too risky and too public to leave to chance or second-rate implementation help.
Ok, I’ll wrap this morning’s missive up. I’ve already worn off the ink on my keyboard keys trying to type fast enough to make my 9:30 am staff meeting. We’ll have some more of our implementation insights to share with you early next week. This will not be resolved quickly – or peacefully. Core support systems such as payroll or benefits administration are so complex by their very nature and importance. Implementations must be protected and nurtured with the best possible resources brought to bear when it’s an employer’s system – or otherwise – it will certainly not be CMS that faces the backlash.
Anything else you‘ve noticed from Healthcare.gov’s debacle?