I’m sure every person who is involved with development of commercial Web application knows about this huge failure – release of the healthcare.gov. I’ve been following this story too, because it’s about the software development – my bread and butter.
I don’t have experience of building Web applications that have to serve tens of millions of customers. I’m just a co-founder of two IT companies: one created software product that’s by more than 100K insurance agents, and the other one is an IT consultancy that helps customers in creating large online stores. But when it was reported that people can’t even login to the system, it was clear to me that the application would require some serious redesign. It’s not just about applying some patches here and there.
People who developed this site made their mistakes, but let’s not simply badmouth them – based on bits and pieces of information we’re getting, the roots of the problem are in the product owner – the government. The product requirements kept changing as recent as six months ago. Initially, the healthcare.gov was supposed to allow browsing insurance marketplace anonymously. WSJ wrote that the requirement to force people login just to enter the application was given only a month(!) ago, which required a very serious change in the application architecture. Now the IT contractors were tasked to introduce the infrastructure of authentication servers capable of processing millions of people.
The fact that this was a pet project of Barak Obama put tremendous pressure on people who were building this system. All these fanfares about October 1st opening were not credible as many other speeches by the President of the USA. CNN reports that the Web site was crushing during the tests with several hundred users, but the administration decided to go live anyway. WAT? Did they expect a miracle? A couple of years ago a prospective client called us for help stating that they’ve developed an online casino, which was supposed to go live in a month, but worked fine only if… there was a single user.
The statement that “our team is bringing some the best” and the brightest minds to fix the issue is promising, but it’ll take time to not fix, but redesign the system. I’m not the best and brightest mind, but I know how the best and brightest started working on this issue. They need time to learn the current version of the system, which some people say is a half-million lines of code. Three weeks after healthcare.gov went live it’s still not working. If a system can’t be fixed in three weeks, it has very serious design flaws. Sending navy seals won’t help here.
It’s funny to hear from people who were developing different component of the system statements like “We’ve developed only the UI portion of the system. After the user hits the button Apply Now, it’s not our code”. Other people happily report that the data hub is operational.
Forty years there was a popular Russian-speaking comedian Arkady Raikin. He had a sketch appearing in a very poorly crafted suit. He came to this made-to-measure suit store asking, “Who crafted this suit?” One of the craftsmen shows up asking “Do you have problems with the buttons?” “No, the buttons are sewn really well, but who crafted this suit?”
Computer World magazine wrote an article stating that healthcare.gov “didn’t have a chance in hell“. 94% of projects that cost $10M or more fail(!). The author of this article writes, “The healthcare.gov contractor was initially awarded more than $93 million for the project, but costs have been soaring above that.” OK. Was this a typical situation when the salesman of this IT firm was tasked to win this project at any cost, and later on “our guys would figure out how to deliver”? Or the IT firm signed off on a project scope that was substantially increased later on? This is a super high-visibility project, so there is a hope that we’ll find out who did the initial estimate of the job, and who changed the scope. Interestingly enough, at this point nobody remembers who and when introduced the requirement to force people to create an account first and then browse the insurance marketplace? Was it even in writing? Maybe an email? No? Nada? Nyet?
Recently Barak Obama said that the good part of being a president is that people always return your phone calls. There is another advantage of being president. When your subordinates deliver really bad software, you get to simply say “Nobody’s madder than me about the website not working as well as it should”. What do you thing would happen if I’d deliver to the client a non-working Web application and said, “Nobody’s madder than me?” Who am I. Definitely not the president of the USA. Sometime people don’t return my calls too. Oh well.
Actually it’s not that bad. You need to give credit to the administration that was able to quickly put together a team of people who take people’s insurance applications over the phone, and the wait time is not more than one minute. Another help is on the way too. According to a reputable online publication The Onion, the new and improved Obamacare program released on 35 floppy disks.
Something’s gotta give. But what’s going to happen with this nice looking girl from the Web site? Is she even married?
Update. I was trying to give a benefit of a doubt to the team developed healtcare.gov site, but when a congressman said this morning that he couldn’t enter the date because the Web site was constantly complaining about the wrong date format, it clearly shows me that healthcare.gov was developed by a bunch of rookies that either have no clue (or don’t care) about how to do a basic validation. In this case the Web site should be redesigned from scratch.
i can not even laugh about it… but it must be George Bush’s fault, he definitively had someone republican to sabotage the project
For a more realistic look at the problems, see this piece by Tim Bray: https://www.tbray.org/ongoing/When/201x/2013/10/21/Obamacare-website.
As someone who used to try to cajole contractors to produce some useful working code for the federal government, I’m not surprised there were problems at the rollout. On top of all the legitimate issues Bray highlights, there is the fact that the “best and the brightest” probably aren’t working for the typical huge bureaucratic corporations that sucks on the federal teat.
Plus, I remember being lied to a lot by developers.
Thanks for the link to Tim Bray’s article. I’ve enjoyed the reading. The federal government is inefficient and bureaucratic. The fact that even the states’ healthcare marketplaces are functioning is a proof for this.
I don’t think it bothers a large percentage of Americans if it never works. My two cents.
The failure of this web site is a lot more than a failure of this web site. It’s an indication that our government has to be dismissed… I’ll blog about it soon.
Verizon will fix it (especially if they outsource it overseas to make higher profit 😉 ). But I also heard that they were already working on it long before Oct. 1.
ObamaCare 2.0 coming?
Watch this carefully!
http://patriotaction.net/forum/topic/show?id=2600775%3ATopic%3A6455331&xgs=1&xg_source=msg_share_topic