Go to Top

Five Data-Driven Tips for Fixing the Obamacare Website

No army in the world can fight an idea whose time has come. So said prolific visionary Victor Hugo. One could certainly argue, especially in the age of Big Information, that healthcare exchanges represent one such idea. However, just because an idea is good, doesn’t mean that any implementation plan will guarantee success. And all too often, huge changes to any large system can do more harm than good. (Remember the Hippocratic Oath? First, do no harm!)

View archived material from the webcast “What Went Wrong with the Obamacare Website, and How Can It Be Fixed?”

All is certainly not lost, though. Today is the first day of the rest of Obamacare’s life. So, looking on the optimistic side, here are some data-driven considerations for fixing the centerpiece of the so-named Affordable Care Act:

1    Got Splunk?
Credit Mark Smith of Ventana Research for pointing out this obvious course of action. At the just-finishing Teradata Partners Conference in Dallas, TX, the master of research methodology noted that a healthy dose of Splunk’s trouble-shooting functionality — which leverages powerful indexes of machine-generated data — could do wonders for solving the “glitches” that persist. Another vendor that offers similar capabilities is Glassbeam, which goes beyond indexes and more into dynamic discovery. These tools can work magic for solving highly complex systems management issues. In the hands of professionals who can make them sing, such applications can expedite troubleshooting by orders of orders of magnitude. (And there are many other useful tools for troubleshooting, many of which we showcased in our Operational Intelligence Market directory, including ExtraHop Networks, SQLStream and Vitria: http://insideanalysis.com/markets/operational-intelligence/)

2    A Stitch in Time
At the core of our data management industry, the Apache Foundation continues to innovate in game-changing ways. The new YARN framework promises to revolutionize the already revolutionary Hadoop ecosystem by embedding a sophisticated scheduler above the Highly Distributed File System. The immediate value is an alternative to MapReduce, which Java programmers know is a prodigious yet complicated protocol that few developers fully understand, and only a handful have mastered. Is the healthcare.gov architecture employing any of Apache’s open-source projects? Or have decisions been made to stick with expensive closed-source apps? Open-source would seem to be a no-brainer here,  partly due to the mantra of that paradigm: “Bad code goes away!”

3    Focus on Strengths
We live in a multinational, globally competitive world more than ever, especially when it comes to software development. A trend now for at least a decade, off-shore code jockeys are involved in every aspect of enterprise software design, testing and delivery. Many of them are very good at what they do, and are far less expensive than stateside developers. Trouble is, too many are not on the approved General Services Administration (GSA) lists; as is the case with a slew of the most cutting-edge applications available today (which roll out almost daily, and in legitimately mind-blowing fashion). All of which speaks to the other elephant in this room: mission-critical systems for the provisioning of health insurance should not be saddled with the old-world constraints of federal purchasing procedures, especially when serious punishments await citizens who do not comply. Politicians should be good at politicking; but let’s leave the enterprise-grade development to business.

4    Show Us the Data
On a recent episode of The Daily Show, the Secretary of Health and Human Services, Kathleen Sebelius, told Host Jon Stewart that she didn’t have the data as to how many people had signed up. This is preposterous whether or not her statement was true. Knowing how many people successfully enroll should clearly have been a critical design point of this system. Besides, sharing data only helps us overcome challenges; circling wagons invites skepticism while protracting serious problems. And once the public knows what healthcare plans cost in other states, that means health insurance providers will need to  achieve competitive advantage by making their operations more efficient, via streamlined processes, more effective software systems, and better customer service. Across the board, we all want more transparency, but definition 1, not definition 4: http://dictionary.reference.com/browse/transparent?s=t

5    Easy Does It
Rome wasn’t built in a day, and neither should anyone expect a system such as healthcare.gov to become seamless too quickly. Truth be told, the stories of major Information Technology (IT) projects collapsing despite multi-million-dollar price tags are many, and there are some especially egregious examples in our own government. One big difference with this disastrous launch is the amount of hype that surrounded it, and the fact that it’s decidedly public-facing (oh, and those penalties, too). Careful is as careful does. Many in the media and elsewhere have continuously injected the argument that “politics” got in the way. This is nonsense because “politics” is an abstraction; only people can get in the way, not amorphous Platonic ideals. (And remember, it takes two to tango.) Ideas can only inspire. Real people — and machines using data — must get things done. If the site is not running smoothly by the time people are required to have insurance, the penalties must be delayed, lest we transition from byzantine to draconian.

Of course, we all know that critics are a dime a dozen. Actions speak louder than words. To that end, if anyone in the administration is reading this, please know that your humble correspondent will gladly offer hours and hours of free consultation to help healthcare.gov get up and running. And here at InsideAnalysis, we know, quite literally, the best and brightest in the world of enterprise software. Rest assured, many of them would be happy to help as well. Holla back!

Eric Kavanagh

About Eric Kavanagh

Eric has more than 20 years of experience as a career journalist with a keen focus on enterprise technologies. He designs and moderates a variety of New Media programs, including The Briefing Room, Information Management’s DM Radio and Espresso Series, as well as GARP’s Leadership and Research Webcasts. His mission is to help people leverage the power of software, methodologies and politics in order to get things done.

Eric Kavanagh

About Eric Kavanagh

Eric has more than 20 years of experience as a career journalist with a keen focus on enterprise technologies. He designs and moderates a variety of New Media programs, including The Briefing Room, Information Management’s DM Radio and Espresso Series, as well as GARP’s Leadership and Research Webcasts. His mission is to help people leverage the power of software, methodologies and politics in order to get things done.

3 Responses to "Five Data-Driven Tips for Fixing the Obamacare Website"

  • Geoffrey Malafsky
    October 23, 2013 - 8:11 am Reply

    Great stuff. I have a slightly different view having seen many Govt acquired IT projects from the inside. First, this outcome was preordained by the Fed Govt acquisition process which has real goals of propagating the infrastructure of acquisition as opposed to the acquiring of effective, efficient product and services. Note that I did not put cost in that description as Govt IT is so almost completely dysfunctional that getting something that works should be worth any cost. The process starts with many people talking about what is needed, then switching to what faddish IT thing can be used drawn from media and Gartner quadrant. Then it moves to the acquisition professionals who gather requirements over many months for 1 of 2 formal methods: 1) std linear requirements analysis, 2) performance based contracting. The former is well known to be untenable for IT since most real technical issues arise as the work progresses. The second is a very good approach but requires continuous monitoring and adjudication, i.e. agile development, which is anathema to the Govt organizational philosophy of making decisions solely to promote one’s own agency office without regard for other issues. Hence, the actual work bogs down and becomes a slave to slow moving non-decisional change request processes. Here is where the contractor has been set up for a fall: they cannot move quickly with ideas that might work, but might now, to build and test but are guaranteed to be blamed and have their contract impacted when the inevitable performance problems occur.

    One obvious question to ask of the design strategy is why the Govt wanted to build their own high capacity site knowing from all the huge volume ecommerce, social media, and search sites that the user experience relies on enormous backends and network management that the Govt (other than NSA) have never been able to get right (anyone from Navy NMCI reading this?) Thus, a 10000X cheaper and faster path would have been to build just an information presentation and brokering site which pushed potential enrollees to insurance company web sites as step 2. Since they have to process the enrollment, let them fold the new traffic into their existing web sites and process, The Govt can then publish a required standardize process and data entry form, i.e promote standards.

  • Nils Davis
    October 25, 2013 - 7:01 pm Reply

    I would just point out that there are many many private sector IT projects that fail, that never launch, that launch far beyond the original schedule, or that launch and then are retracted. The government does not have a monopoly on this. And in fact, the exchanges, as broken as they were , did open on October 1st, and some people – at any rate, thousands – were able to sign up the first week. There are hardly any private sector IT projects of significant scale that achieve that.

    Am I unhappy that the exchanges didn’t start up with only minor glitches? Of course. Am I surprised that it actually launched on the day it was supposed to? Very!

  • Joe Cusack
    October 28, 2013 - 6:41 am Reply

    Regarding Point 4, it IS preposterous that the Federal government is content to have monthly metrics. Where can I find such lenient clients? I worked in an era when private businesses launched satellites for Federal and commercial customers at the appointed hour with nearly zero defects and down time.
    Everyone understood you couldn’t fix them after they left the Earth. How has American delivery become so sloppy?

Leave a Reply

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

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>