Credentials: 25 years working with computers, over 20 in IT, have direct experience in the front-end and back-end challenges of integrating multiple diverse technologies with multiple vendors in systems that can suddenly spike in load, must be available,must have perfect data, meet all regulatory and legal requirements and still are capable of performing whatever function they were designed for. I've actually done nearly every kind of business process and software development methodology buzzword that you've ever heard of and have seen a lot of project and program managers wrestle with complex systems. I have not done government procurement.
The truth about methodologies and approaches is that nothing works for everything. All of them are useful under some circumstances. The same is true for hardware architecture, software packages and the prejudices of individual programmers, software engineers and program managers. Whatever easy way you think that Healthcare.gov might have used to be perfect on launch day, you're wrong. Your experience in doing something easy doesn't count.
There was a lot done wrong to make the project nearly impossible, hostile stakeholders, little action till after 2012 election, fixed launch date, large list of regulations and laws to comply to etc. Nothing you know how to do could fix all that. But at the moment the people working with Healthcare.gov are probably the only people in the world with anything much to say about what could have made things better.
Follow me below the fold for a lesson in process engineering and why I think Healthcare.gov is something new, that makes all of us outside it trying to match our experiences to it probably wrong.
I'm going to step away from software for a second and switch to something everybody understands.
Consider the problem of moving 5,000 people 20 miles. There are a lot of possible approaches.
If you are a Roman Legion, that is only a day march. No big deal. Your investment in the Roman Legionairy Training, Roman Road Infrastructure and centuries of experience supplying legion-sized units pays off in a big way. Hell, you can quick march 30 miles pretty routinely, as long as you don't have to do it several days in a row. This speed assumes good weather, good road and sufficient clean water and food either carried with the Legion or available at the destination.
Mongols don't need the good road or the food supplies, but they do need grass or equivalent to feed their many horses.
A WWII infantry army would organize it differently and have a different idea of how far they can go and what logistic support they need. An American army in WWII would have a lot of trucks and gasoline. Most other armies still used horse carts and feet.
An Airborne unit of infantry can pretty much ignore all terrain except for wherever they intend to land. This assumes command of the air and decent weather.
If you are a bunch of American football fans going to a game, that's a short drive. You need only a place to sit when you arrive, parking and sufficient restrooms and snack options to satisfy the needs of a several hour stay. This assumes either private or public transit and a lot of other infrastructure, but needs little training of the individuals involved.
If you are trying to evacuate a small town from a chemical spill, you lack the advantages of the Legion or the voluntary movement of football fans. The authorities hopefully have some emergency contingency planning and will help those without transport, as well as the old, disabled or underage move fast enough. If the authorities are what the citizens are fleeing from (or from a battle, say) things get a lot worse. Worst still is a disorganized mob of refugees under attack the whole way.
Each of the above groups is doing the same task, but their training, technology and approaches are completely different.
Romans traveling 40 miles will build an entire fortification from scratch routinely halfway and sleep the night. A WWII soldier would only do something similar in a hot zone, and the "digging in" is much more rudimentary. But the WWII soldier has much better scouting resources and communication than Romans did, so the odds of surprise attack are a lot less, and "safe" zones really are safe. Even in the heart of Italy, Romans had to deal with highwaymen and pretty large organized groups of hostile armed locals from time to time. Not to mention the odd slave revolt.
Training, techniques and approaches evolve from experience. The "travel 20 miles" problem goes back to the dawn of time. Hunter gatherers just didn't go places that weren't likely to provide food/water on the way. Most agricultural peoples do not travel further than they can go in about half a day's travel, because food/shelter/water are iffy when away from home. Armies of all types have to travel away from home, so they all evolve practices to aid this. Often they work like hunter-gatherers, looting what they need as they go. Aside from the obvious problems with such a strategy, it means you can't go back where you came from and you are limited in where you can go. So baggage trains were invented...although animals or people moving stuff for your army also take food/water. Railroads helped a ton except where there are no tracks. Ditto trucks/aircraft/etc. All solutions to the "move a bunch of people 20 miles" involve both infrastructure and extra training.
If the problem changes to "move people 20 miles for 30 days straight" or "move twice as many people 20 miles" or "20 miles through that swamp" or "20 miles of water" or "see how fast we can move people 20 miles" or "move them under fire" you have to invent entirely new solutions, or adapt old solutions to the new challenges. So far the most extreme in difficulty is "move 20 miles straight up", although from a logistics perspective that's a smaller problem than "move 50,00 troops out of Afghanistan without exposing the other 30,000 troops to unreasonable danger".
Healthcare.gov has a web page with products, but the task is not "make a web page showing products". This is "show products from a large number of individual vendors that qualify for a single individual given a bunch of rules we barely had written in time for the launch". This is like having a bunch of troops from 15 different nationalities all marching on the road at the same time, and at least a few of the military units actively hostile to the idea of marching 20 miles.
Healthcare.gov has you create an account, but the information needed for proof of identity is much more severe than is typical and the data they store is far more sensitive than anything I've seen on the net. It's extremely rich identity information and credit information. You really don't want to lose track of that. This is like trying out a Tsunami evacuation plan on no notice rather than just having 5000 people show up to the game before it starts. There probably is a right way to do it but nobody knows what it is because Tsunami's don't happen often enough to gain any real expertise. If the same town gets hit by several, the third or fourth evacuation will go pretty well. (see setting up account issues, improving rapidly)
Healthcare.gov connects to multiple data sources aside from the vendor information. But it isn't like your typical ordering system, where most of the data sources are owned by a single entity. The required authentication and data collection systems are scattered across many government agencies, some of which probably have never done much data sharing before. Add a bunch of civilians, camp followers and a few marching bands in the middle of those 15 different types of troops mentioned above. Some of them have cars and trucks, some are pushing wheelbarrows, some are trying to walk while carrying an elderly parent on their back and herding 5 children.
All of the road we're supposed to travel on was built in a hurry, by a bunch of different people. They all claim it is ready, but the stretches are of wildly different quality and a bridge or two is entirely missing. At least one road goes the wrong way and another doesn't exist, forcing a cross-country hike.
Oh by the way, we need to move 50,000 people, not 5000 people. That was the situation for the first few days.
Before the event you can try to imagine moving the 5000 people 20 miles. You can recognize the difficulties of multiple types of training, multiple types of transport, multiple hands building roads and bridges, you can look at the weather and try your best to manage all that.
But in reality, you will probably fall back on whatever approaches you know best, even if it suits a Roman Legion on the march better than the giant mix of people, or Rush Hour traffic, or even herding animals (the Mongols used those techniques when forcing refugees where they wanted them to go). And because it is such a complex mess, and because it is also a brand new kind of complex mess, I can guarantee as sure as the sun will rise that whatever your time estimate for the 5000 people arriving 20 miles away, it will be optimistic.
All the people won't arrive together. There will be a trickle, then a stream, then a flood, then it'll ease off and end with a few stragglers.
We're moving into the "stream" period about 10% of the way into the process (3 weeks in, 26 weeks to sign up...). The goal is 7 million signed up by March.
Be patient and be kind. Unlike my analogy, the IT people are able to fix roads and build bridges during the journey, ahead of a lot of the walkers. They're able to come up with better approaches to march them through the swamp, maybe even deploy some airlift for disabled and sick, and provide better instructions along the way.
When all of this is over it will be much more possible to build a new system equally complex. A tremendous number of lessons will be learned, because somebody will have done it once. Right now, it is merely superficially similar to what already exists. Do not let appearances fool you.
Finally...nobody is slacking. There is the maximum imaginable pressure already exerted to get the problems fixed. Pointing fingers and assigning blame may help improve a future project but doesn't actually help get the existing one working. Cut the IT people some slack. They did the best they could in a horrible situation, and are not going to be seen by their families this holiday season. If you want to be mad at somebody, focus on the folks who tried to plan this, but even then, well. I'm not normally one to quote Jesus but none of us who do that kind of thing for a living are without sin. Don't throw stones because you know, deep inside, you probably would not have done any better and most of us probably would have done worse.
I save my blame for the sales pukes who convinced the government that their contractors could do the job in no time and promptly ran up several times the original bill. But that may be a bias on my part. This is new, perhaps their methodology for estimating the challenge were wildly thrown off too.
It's big. It is new. It is extremely complex. It has legal, security and regulatory restrictions all through it. It has actively hostile oversight and some of the so-called partners may be actively trying to subvert it.
And the project actually started coding in Spring, with a legally mandated Oct 1 release date.
It is a new thing under the sun and it is amazing that it works at all.