Can Analytics Save Obamacare?

I have had a few intelligent conversations (and a few not so intelligent) about what is wrong with the ACA roll-out and what can be done to fixed it.  I'm not talking about policy or politics, simply the actual ACA website, www.healthcare.gov.  In my opinion this is where they went wrong:

1. Outdated project management methodology:  In my project management classes, it would be typical to talk about how NOT to implement a project.  The common example was the DOD (Department of Defense) using a "Waterfall model".  This was the common methodology used by the government for all IT projects in the 80's and 90's...and also a reason why many of these projects failed.  An estimate is that 75% of all of these projects failed or were never used!  And you may have guessed it, ACA used this waterfall model for development instead of a more modern approach such as lean or agile methodology
2. The best and the brightest?  It is questionable whether or not the best computer programmers and IT project managers were tasked to this project.  Working for the government requires many hoops to be jumped, one of which is American Citizenship.  Also, government contracts are often awarded by low cost as opposed to merit.

The government is now doing what they can to make bug fixes.  Although the political issues will probably always be there when discussing the ACA, it looks like these technological issues need to be resolved first.  Bug fixes at this point are the proverbial "lipstick on the pig" because the framework that the ACA is built on is not built for optimization and growth.  

Where do analytics belong in this equation?  Analytics can help define the desired outcomes and help optimize the system once these bugs are resolved.  Even after these bugs are taken care of, the system will need to be continually revised and updated to provide the best patient outcomes.  This is where I feel analytics can come in.