Skip to Content

ObamaCare: IT in the spotlights, almost never a good thing

Sogeti Labs
October 25, 2013

461_30_Barack-Obama-using-his-Mac-and-BlackberrySince October first, thanks to Obama’s Affordable Care Act, Americans can finally sign up for reasonably priced healthcare insurance even when traditional insurers would not cover them (for example because they had a history of illnesses). The central point of this step is that there is now one website on which people can sign up. The opening weeks of this new site have been dramatic: it does not work as planned. I cringe when I think about this, because I can imagine all the things that could have gone wrong: the difficult communication with the customer, the short time lines, perhaps some coordination with an offshore team, testing was held off till the end, complexity in the underlying systems, integration challenges, peak loads are much different from what was anticipated… etc. When reading this great (long) post by Yuval Levin, three thoughts popped into my mind – which I provide merely as inspiration, as they may or may not be close to what really happened inside this project :-):

  1. The IT provider is to blame. No, this is not a childish attempt to just talk badly about a competitor: I believe strongly that whenever a service provider takes on an important and highly visible project like this, they should take complete ownership of its success, regardless of how the liabilities are defined in the contract. This means that if problems arise, you solve them creatively (of course), or you escalate until you reach a decision maker who can make decisions that solve your problem. In this case, up to president Obama if you have to. If requirements don’t come in on time, you go get them (you haunt the people who decide on requirements to extract the right information). If last-minute changes pop up that threaten the success of the whole project, you refuse(!) to just implement them. Extraordinary projects require extraordinary dedication to success.
  2. User Adoption is the most important thing. Didn’t we learn anything from Amazon, web startups and kickstarter companies? How do you drive adoption? Start early, allow sign-ups before your product is ready to get people on the ‘beta’ list, show what you have to offer in clear picture, a video, or better yet: allow people to navigate right to the end of the ordering process before needing to register. In this case, users have to sign up (through a very complicated process) before they can see ANYTHING. What they could have done instead was create a simple site, that shows you the type of choices you can make with a range of possible fees attached (“anywhere between $100 and $500 per month, sign up here to receive a personal quote”). I am astounded that it seems nobody with any Silicon Valley expertise helped think about how to ‘drive adoption’ for this site. (And it’s not just about ‘the site’, as Yuval rightly argues: if adoption is unbalanced, it could mean great risk or cost for the insurers behind this site!)
  3. Simplicity isn’t easy. I know, I know, I’m an armchair analyst. I was not involved in this project (alas), so what do I know, really? It was not an easy project, I’m sure. The simplicity that is needed on the users side is not there on the backend: there, the system talks to many different parties and the number of people, states, companies and variations involved must be enormous. Still, all this was known on day one that the project started, so I wonder what the approach has been. Did the site try to solve all complexities in some middle layer? Were all connecting systems challenged to change to a new, simple, interface? Were products aligned to make things simple, or were product complexities fed into the site? I just hope there was some kind of agile approach, where big and important decisions were taken early on, and along the way the decisions got smaller and smaller. The big decisions from the start should represent the simplicity that resonates through an entire project like this.
Of course it hurts my professional pride, when ‘IT’ fails in such a visible, important way. But then again… this too will be solved, it will work, they’ll fix it in the next update, or the next. Users will forget the beginning and use the healthcare exchange for how it was intended. A year, two years from now, people won’t even remember. More people will be covered, leading to healthier happier people. In the end, IT does help to create success; I just wish the process wasn’t so painful so often.

About the author

SogetiLabs gathers distinguished technology leaders from around the Sogeti world. It is an initiative explaining not how IT works, but what IT means for business.


    One thought on “ObamaCare: IT in the spotlights, almost never a good thing

    1. Agree with your basic tenets, but would like to point out a couple of caveats. Take this as coming from someone who has been in the trenches “taking ownership”. Important to remember two key facts that are in fact classics on the “why projects fail” list:
      – Scope of this important project was never fixed, yet the deadline was fixed. Remember that states kept opting out of having to set up their own exchanges and this then was added to the scope of the federal exchange. Each exchange represents potentially hundreds of integration points to insurers and data exchanges, additional state rules etc. that all need to be developed and carefully tested. So then, why did the contractor not just add resources to make the timelines?
      – Funding for the project was not approved and not released by Congress until WAY late in the process. Congress used withholding of funds as an important means to thwart Obamacare. Plus, Congress hoped that the Supreme court would throw out the law entirely, so no need to release funds. In the mean time the department of Health and Human Services scrambled to come up with means to finance the effort.
      This is quite a deadly combination of circumstances for any critical project. Take that in combination with some questionable architecture and workflow decisions and you have the failed launch we are experiencing.

    Leave a Reply

    Your email address will not be published. Required fields are marked *

    Slide to submit