The project hasn’t failed, and there’s no fair comparison between the failures of FEMA and the rollout of an online application. Otherwise, I agree with you.
The IT Project which had a goal of getting HealthCare.gov launched by 10/1 absolutely failed. There’s not even really a debate about that, they launched a bug ridden, non-functional web application.
That doesn’t mean it’s the end of the world, or that healthcare reform failed or whatever–but to deny the IT project was a failure is simply incorrect.
If that was the measurement of software project success there would only be a handful of successes . Missing a goal doesn’t make a failure of an entire project.
It makes the launch a failure.
If they had said “we’re not ready”, then it’s a delay and failure to meet the date, but not a launch failure.
I don’t see the purpose of such categorization. Success or failure is measured in the long run. If they get things working reasonably by the end of the year it will be forgotten, because it has no significance. There were plenty of things done wrong here, but it doesn’t matter if they can be corrected.
It’s only a failure if it’s never going to deliver (or if it was terminated due to failure to make objectives). You’re simply wrong. Missing a date does not automatically qualify as a project failure unless it’s specified that the project must deliver on a certain date or be terminated. It’s very commonplace for projects to deliver late, but ultimately deliver enough value for years and years that it’s laughable to label the overall project a failure.
There was a failure to meet a milestone. Someone screwed up and should be held accountable. But nobody with any project management familiarity considers this a whole project failure, not unless they have political incentive to believe it failed before it even started.
It has no significance? I guess the millions of man-hours wasted by the people who tried to do what the government told them to do doesn’t count for anything?
Also, health care providers have been sending out millions of cancellation notices for existing plans, because they were told that people would be able to sign up for the news ones after Oct 1. They stuck to their plans, the government failed. Now what happens if people can’t buy the coverage they need in time to meet deadlines? What happens to the people who get sick or injured while in limbo between cancellation of their existing plan and being able to purchase a new one?
Also, there is a serious risk here that the people who tried to buy health insurance fall into two categories - people who really, really need it because they are in poor health, and people who don’t really need it but were willing to give it a try. i.e. all those young people out there who are already at risk of not buying insurance at all. The people who really need it will keep trying, but how many of the fence-sitters now have a ready excuse for not buying what they didn’t want to buy in the first place, and will stay out of the system? Obamacare will fail due to adverse selection if they don’t get about 2.7 million young people to sign up and pay for the insurance. That’s according to the CBO, who also identified this as a risk to the whole program. A risk that’s just been made worse by this debacle.
There is no doubt about it - this was a failure. It has damaged Obama’s ‘brand’ and the trust in government. It has made Obamacare incrementally more likely to fail.
Yes, if they get this running in a couple of weeks and convince everyone of that and get everyone to come back, most of the damage may have been mitigated. Likewise poor, buggy software releases in the private sector sometimes get fixed up quickly and the companies behind them salvage their tarnished reputations. That doesn’t stop them from firing the people who botched the first release.
Seriously - releasing something to the public while knowing that it has no chance of working isn’t just a coding error - it’s a violation of business and engineering ethics and would get a private company in very hot water if the users could prove damages. This isn’t just software with a few bugs - it’s a major portion of the biggest change to healthcare since the 1960’s, and it doesn’t work. At all. And the people who foist it on the public knew it wouldn’t work, and rolled it out anyway. It’s unbelievable.
Not only that, but their temporary ‘fix’ was to put very misleading insurance data online, making customers believe that the insurance they want is much cheaper than it will be. They put up a disclaimer that the insurance might actually be even cheaper, but said nothing about it possibly being more expensive even though they gamed the data in a way that makes it almost certain that it will be more expensive than what they are claiming for anyone who doesn’t qualify for a subsidy. That’s fraud.
Anyone who cancels their insurance based on that information only to find that the real number for new insurance is higher would have an actionable claim against a private company. And it’s the kind of fraud that courts like to grant massive punitive fines for. But I suppose this government will get a pass on that too.
No, there wasn’t a ‘failure to meet a milestone’. There was a failure to meet the conditions required to release the software to the public, and they released it anyway. If you are familiar with project management and you think this is acceptable, please tell me where you work so I can be sure to never buy anything you have to sell.
By the way, the people you’re arguing against in this discussion ALL have ‘project management familiarity’. We all understand that software sometimes (most of the time) does not get finished on schedule. And we also understand that when that’s the case, you do the right thing and you don’t ship it. Another thing you don’t do if your schedule starts to blow out - you don’t decide to pull it back in by cancelling your testing phase. That’s the kind of thing that can (and should) get a project manager fired.
We also understand that usually a project manager doesn’t get to this point because the people above him should be keeping watch on the project and should be calling him for answers and mitigation strategies when the schedule starts to slip. And in a properly run organization, one of the checks on the system will be the QA manager who will have developed a series of acceptance tests, and who will simply not sign off on a release until he is satisfied that the quality is there and that the acceptance tests have passed. In my organization, the QA lead has absolute authority to stop any release if it doesn’t meet his standards. He’ll have to justify his decision, but he has the authority. I don’t even know if there WAS a QA lead on this project. I don’t think so.
In this case, it appears that what happened was that the warnings of engineers were met with responses along the lines of, “Just get it done. There is no other option. We’re launching on Oct 1, so it had better be ready.”
Interesting diatribe. I like that kind of thing. But it’s just full of political posturing that has nothing to do with reality.
Let’s start with the millions of man hours. Even the ridiculous figures for cost put out by the opponents don’t account for that much time.
I don’t like to play the cite game but do you have any evidence that any health care providers are cancelling anything? Why would they do that, because they don’t like making money?
Continuing, who cares about Obama’s brand? This is about finding a way for people to afford healthcare, if it does that then it is a success.
And once again, releasing things to the public before they are ready is the standard of business ethics, and government ethics.
Why aren’t you and all the other anti-Obamists complaining about the price tag? Working or not that is the real scandal here. There is nothing this software could have done to justify the cost.
Well, Sam, not going to go for the line by line response. For economy’s sake, I’ll just leave off all the points wherein you think that my liberal hypocrisy is somehow pertinent. Happily, that’s a savings of fifty percent right there.
As well, I’m a bit stumped as to how I might express my wonder and awe at the wide range of your knowledge and information, which entitles you to vast and sweeping judgement with but a paltry investment of fact. (I may need to start wearing sunglasses when reading your posts.) It is hard to even mildly express skepticism without offense to dignity, one does not meddle in the affairs of Canadians, they are subtle, and quick to anger.
Is it pretty fucked up? Can it be fixed? Dunno, way over my pay grade, but most things can be. Especially if they are built up from modules of working software that are proven to be sound, so at least you know where the bugs aren’t. They aren’t in the parts that worked before you got your mitts on them, and if you didn’t change anything, they aren’t in there now.
So. broken, needs to be fixed. Is that the Mambo Contest for the Zombie Apocalypse? A killer asteroid is unleashed upon us, a dread plague? No one who was uninsured yesterday is any more uninsured, not even supposed to take effect for a couple months. The hysteria on display wildly exceeds the damage done. I expect it can all be rectified with less than a tenth of the money the Insane Clown Posse wiped their collective asses with and flushed down the toilet. What do you think, two, three billion dollars ought to do it?
As far as blaming Obama for picking experts, well, yeah. In your entire career, you never once met a thoroughly credentialed idjit? Someone who looks like the bee’s knees on their c.v. but doesn’t know endless loop from Endless Love? Guessing you have, guessing you sussed them out right quick. But how would I know the difference? Say its you and him, what are you going to say to me that will show me that you know your shit and he don’t? When I have no idea what either of you are talking about. That being why I am hiring an expert.
Obama is law, history and politics, may not know a binary from a bipolar. He trusted the wrong expert? Who hasn’t?
No, success or failure isn’t measured only in the long run.
If you create chaos and waste a bunch of people’s time and money, even if you fix it in the long run, you still created chaos and wasted a bunch of people’s time and money.
Which is exactly what has happened here.
Yes, except for the part about success or failure not being measured in the long run.
Has a single massive web-based service ever succeeded on launch day? I’m from the gaming world, and it’s basically taken as a given nowadays that any game that runs off of a server will be completely broken for a couple months. Even if it’s perfectly well programmed and they have good hardware, it’s never cost effective to create a service for launch conditions, because then you’ve overpaid drastically for infrastructure that will be useless after the initial rush. There’s a pretty good reason why Google has “invite only” betas to all their new services – so they can develop their programs incrementally rather than having to deal with Google’s entire userbase crushing the new service at once.
Okay, this failure had legitimate bugs and not just scalability issues, and I’m not excusing that. But it seems to me to be silly to expect this to succeed where, to my knowledge, almost everything of a similar scope has failed in a related fashion. If the Government had overpaid for tons of hardware and software to make everything smooth at launch, people would probably freak at how much it cost when that stuff largely wouldn’t be needed by December. It would be considered a textbook case of government overspending.
I’m not saying they didn’t drop the ball, they kind of did, but I don’t think they failed quite as hard as everyone seems to think they did.
As for the “tech surge”, it seems like a stupid, stupid idea to me. Hiring new people means they have to get used to the software and infrastructure, they’re not at full efficiency (and, in fact, tend to cause more bugs than they solve as they get used to it). Further, throwing 100 people at a problem is not even close to a 100x performance increase, there are heavy diminishing returns because things frequently cannot be done in parallel. It also introduces more room for human error because the more contractors you throw at a problem, the more these people have to communicate, and you risk more and more miscommunication the more people you have.
Hire a couple more independent experts, maybe. Sure. But I think this tech surge is a political show to assure people Something™ is being done. I really think the most effective solution would have been to just leave it alone and in a month or two it would have fixed itself anyway. I highly doubt throwing more experts at it is going to speed this up.
Edit: Okay, I finally found an article stating that the “tech surge” might only consist of 8 experts. That’s not as bad as everywhere else was making it sound.
I wrote “success or failure isn’t measured ONLY in the long run”.
The failure of this system today is completely independent of any future long term success it may possibly have.
Today’s failure is a fact and a done deal, no amount of message boarding will change that.
So we’re done here?
They had to cancel those policies because they are not ACA-compliant. The idea was to cancel those policies and direct policyholders to sign up for new ACA-compliant policies being sold on the exchanges.
[Note: this is not to get involved - one way or the other - in the point being made about the implications of this, but just a factual note.]
I don’t know that there even was a single person taking ownership of this project, but I can assure you at a private software company a launch of this type would be considered a project failure. Maybe if there was a really good reason and it was outside the PM’s control they wouldn’t lose their jobs, but if it was a highly public product and the launch was a highly public failure it’s likely the PM’s head would roll even if there were very understandable reasons they failed to meet the criteria for their deliverables on time.
Of course it’s hard to imagine a failure of this magnitude for a private software project. Take a product like Office 365, most people probably don’t even know when it officially launched, because it had been in development for years and the time it went “public” Microsoft had already opened it and had it opened for awhile to MSDN subscribers etc. Whoever was running the Office 365 project at Microsoft, if their timeline including doing integration testing a week before it was released to the public, and there wasn’t months or even more of alpha/beta testing in the project timeline that PM would be fired for being incompetent and they’d probably look into who hired them to try and figure out how someone so ignorant of standard project planning/management even had a job at Microsoft and how they were assigned to such an important project.
Even the Microsoft launches that get a lot of negative publicity, like Vista for example, were really shining beacons of success compared to HealthCare.gov. The day of launch Vista was a functional, working computer operating system. It had features (especially in the realm of all the UAC) that annoyed and frustrated users, but it wasn’t non-functional.
Probably the closest private software projects I can remember recently to be this bad would be the GTA Online launch and the newest SimCity from EA. In the case of GTA Online though, the standalone, single player version of the game shipped several weeks before they turned on the online functionality. The single player game by and large received immensely positive reviews, so one could argue GTA V itself had a successful launch but their online feature launched on 10/1 was not done well to start. The newest SimCity on the other hand was an unmitigated nightmare that required always-on play for no discernable reason and yet EA’s servers had nowhere near the capacity for the players that were actually playing (and demand for that game wasn’t exactly immense–suggesting EA just dropped the ball.)
What I’ve seen for a new system launch just as an example would be categorizing the current project status with a simple “red, yellow, green” color code. Red would be the worst. Generally there is a distinction between the project to get a new system to launch and then ongoing system maintenance/improvements. Typically once the project is handed off to an application support team the project is finished.
So in a private company, this project would have been red for some time as they obviously were not going to get any testing done before the 10/1 launch date. This project would have had a list of requirements, and just actually turning it on 10/1 would not be enough for it to be considered a success. A private company would consider this outcome to be a failure for the project itself, but that doesn’t mean the product/system would be cancelled.
When I’ve seen stuff like this happen, I’ve actually seen the initial project basically scrapped and a new “remediation” project begun with a new PM, and it might be that project that actually satisfies all of the requirements from the original project requirements list that were not met by the original project team.
We can argue about whether a situation like that is a “project failure” or whatever, but that’s just semantics. A PM that missed their deadline, failed to get any significant testing done, failed to even schedule various standard testing processes and etc would be considered to have failed. The company may (and probably would not) not give up on the product/system itself, but that PM would be considered to have failed at their job.
I actually predicted with 100% certainty that this would be mentioned. But I’ll note that running a political campaign isn’t the same as running government. Project ORCA was actually run by the campaign’s Political Director and the Director of Voter Contact, it wasn’t ran by the campaign’s normal IT division. It also represented a small portion of the campaign’s total spending on IT. So what really happened is guys Romney picked to run the politics and voter outreach cooked up an IT project. It should have been handled by a different part of the campaign, and unsurprisingly an IT project ran by people with no IT project experience ended up failing to work on election day, and they had several of the same problems you’ve seen with HealthCare.gov (no beta testing being a big one.)
It ended up being a small project that did have a measurable negative impact on the campaign, but because it was small relative to other IT spending and not being ran by the IT staff it’s quite possible it was never on Romney’s radar–I wouldn’t be blaming Obama if a new version of say, the Census.gov query tool was released buggy. I wouldn’t expect him to even know that project was going on. But HealthCare.gov is a core part of a really the Presiden’ts only domestic achievement and probably the only thing historians will even hope to be able to tout about his mediocre, Carteresque Presidency. I don’t believe Romney would have been kept in the dark about this project until after 10/1, given his managerial experience. Nor do I think he’d have hired a team of mostly management-inexperienced policy wonks to run government agencies.
I can assure you that you are incorrect. I’m not defending this project, it’s a mess, enormous amounts of money were wasted, and it was totally mismanaged. But private software companies run into these same issues all the time, and particularly when they are consulting companies that have no stake in the final product and are happy to have projects run late and over budget because they make more money. But success or failure will still get measured in the long run and natal problems get forgotten if they are cleared up in a reasonable time. I’m not even saying that will happen here, but it’s too soon to be declaring success or failure of this project in terms of functionality.