Oct 21 2013
There is a rule of thumb in SW projects, which becomes a scientific law of physics when it is applied to large SW projects for the government. President Obama is about to repeat a huge mistake made by HHS/CMS on the Obamacare Website leading up to its disastrous debut.
Obama is going to throw a lot of bodies at a huge problem – and make matters ten times worse:
The Obama administration Sunday said it’s called on “the best and brightest” tech experts from both government and the private sector to help fix the troubled website at the root of the Obamacare enrollment problems.
Emphasis mine. The HHS statement is quite enlightening to those who know how the4se kinds of announcements are crafted:
Our team is bringing in some of the best and brightest from both inside and outside government to scrub in with the team and help improve HealthCare.gov. We’re also putting in place tools and processes to aggressively monitor and identify parts of HealthCare.gov where individuals are encountering errors or having difficulty using the site, so we can prioritize and fix them. We are also defining new test processes to prevent new issues from cropping up as we improve the overall service and deploying fixes to the site during off-peak hours on a regular basis.
Oh my. The problems must be enormous., This proves the in-house CMS team that pretended to be the system integrator is still in over their head, and this will takes months to fix now that they had to bring outsiders. [Hint: the best place to figure out what went wrong and how to fix it is to find all those disillusioned experts who left or were pushed out for being 'too pessimistic' prior to the debut - they were right you know]
This is all desperation by the administration. It is not the first time this stunt has been tried, but the result is ALWAYS the same. More delays. Bringing in an army of outsiders just added weeks and weeks to Obamacare fixes. The rule of thumb is called Brook’s Law:
Brooks’s law is a principle in software development which says that “adding manpower to a late software project makes it later”.
The same bush-league mistake was made by CMS during the end of the build up to the Healthcare.gov debut:
A Reuters review of government documents shows that the contract to build the federal Healthcare.gov online insurance website – key to President Barack Obama’s signature healthcare reform – tripled in potential total value to nearly $292 million as new money was assigned to the work beginning in April this year.
The increase coincided with warnings from federal and state officials that the information technology underlying the online marketplaces, or exchanges, where people could buy Obamacare health insurance was in trouble.
Did not help then, won’t help now. But one thing is for sure – this sign of desperation means Obamacare is in serious technical trouble. It is clear the incumbents (government and contractor) have not figured out what is wrong. I said days ago this would happen. The team that was so clueless as to implement this mess was never going to figure out how to fix it.
One of the most glaring indicators this is true is this line in the HHS statement:
We’re also putting in place tools and processes to aggressively monitor and identify parts of HealthCare.gov…
So, they are going to instrument the SW with logs to figure out where the problems are. I guess they don’t have run-time debug flags to capture performance data and interface data to decipher where the issues are. The fact we are weeks into this and they need to add monitoring tools (or points) is another bad sign of inexperience.
This monitoring capability alone will take weeks to put in place (and test), a week to capture data, and then weeks to analyze the logs to figure out where the problems are. That is 6 weeks easy.
Then add weeks to design some fixes, then more weeks to implement and test. Then a day or two to capture more data.
If all goes as these things always do, the first cycle will be a bust and it will be repeated. I envision two months to get the first concrete corrections in place. But this will not be enough to get Healthcare.gov where it should be. This will address the first order ‘tall poles’, but then there will be the second order, third order, etc. Something this bad should need 2-4 rounds of corrective action (each 3-4 months) to get to a decent baseline.
And all of this is before we add all these newbie ‘experts’. As the logging instrumentation is being put in place, the experts will be reviewing technical documentation, collating findings, debating the issues and solutions, and then attempting to clean up the technical documentation to guide the new fixes. These outside experts need months of work just to get up to speed. And all the time this is happening, the government and industry incumbents will be pointing fingers and doing CYA. The outsiders will be presented a very confusing picture, all the while waiting for real log data. It will be a mess.
I have not even accounted for establishing the test capability mentioned by HHS (the ‘new’ test processes – in itself a major development effort). If they try and fix this by going ‘live’ each time they make a fix, they will consume any good will they have left and Obamacare will be DOA. So they need to set up a way to realistically test it off line.
Simplistic, naive – and why they had no clue they were in deep trouble. Worse yet, I understand Healthcare.gov is actually copied and distributed to 36+ sites, and is not really a single integrated system. Each state exchange will therefore exhibit different issues and problems, so trying to unravel why something breaks in one state and then works in another will delay solutions in general.
This announcement is the last sign I need to predict Obamacare will not meet its scheduled capabilities this calendar year. Which means Obamacare will have to be delayed before it goes into a death spiral.
This has now become Extreme Amateur Hour.
Update: Too funny:
“These aren’t glitches,” Klein said on Morning Joe on Monday. “The website, to a first approximation, simply isn’t working.”
He explained that the more fixable problem facing the website was dealing with traffic. The bigger problem, according to Klein, was whether the right information was being sent to insurers after users enroll, including messages that users hadn’t yet enrolled even if they had.
Update: WaPo poll shows obvious repercussions of failure on the Obamacare Website (Healthcare.gov) – Concern over medical services:
Fifty-six percent of Americans say the Web site problems are part of a broader problem with the law’s implementation while just 40 percent see the Web site problems as an isolated incident.
Yeah, well if you can’t build a website, why would we trust you to do something like diagnose a medical condition, administer medicine or do life critical surgery?