Understanding Healthcare.gov’s Rocky Rollout Infographic

Galorath Inc. (the SEER Cost, Schedule, Risk Model Developers) watched the healthcare.gov rollout difficulties, the outcries and finger pointing and decided to take a more analytical look. While it is easy to throw stones at stakeholders, this was a huge IT project and there were bound to be challenges. Could it have gone better? Sure. Were there adequate resources? Seems so. Should testing and quality assurance been more rigorous? Yes, but there didn’t appear to be adequate time. Were the requirements firmed up in advance? That could have been a significant contributor.

We are confident that healthcare.gov will recover and this will go down in history as another IT lesson learned. Using our SEER models up front could have shown the minimum possible schedule as well as costs and risks. This foresight could have helped the government and suppliers to do better to plan for the inevitable defects.  Tracking progress with SEER could have also provided an early warning indicator...  Probably early enough that corrective actions could have helped.

Scroll down to the bottom of this post and look for the heading, "Embed Our Infographic On Your Site!" and you can use that code to embed this infographic on your site. Please link back to http://www.galorath.com if you use this infographic.  We have also provided a PDF document you may distribute freely with link credit to Galorath.com.  The PDF is 11 megs so be patient.




Infographic Transcript:

Why did healthcare.gov have rollout problems?

  • Extreme number of legacy systems, inside Government and among 3rd parties
  • Citizenship certification and income verification
  • Numerous, powerful stakeholders: the President, insurance companies, states, Congress
  • Complex eligibility rules
  • Different rules for different states
  • A volatile environment with numerous starts and stops: Supreme Court, Presidential election, extreme political disagreement
  • Registration requirement added very late in development
  • Not enough time

Not just another eCommerce site!How much has been spent on Healthcare.gov
How much does $150M buy?

  • 10-30 major systems
  • $18K average monthly salary (government rates)

What kind of quality is delivered on delivery day?

How much software can you buy for $150M?
500,000 to 600,000 lines of code Or about 24,000 unadjusted function points. In other words…

Half of Android 4.0 (2011)

One tenth of Quickbooks (2012)

A teeny fraction of Facebook

Why did some states develop their own exchanges?

  • Historical insurance sovereignty within states
  • State-specific insurance markets

Which states developed their own exchanges?
Map of states implementing there own exchanges

What were the states’ web site experiences?

  • 1,700 individual rules affect eligibility for health insurance subsidies in Oregon
  • In Oregon, writing the eligibility rules engine took 12 people nine months.
  • Estimated cost of $54M
  • Paper only on Oct 1, site operational in mid October


  • Given $990M
  • Operational on time


  • Operational on time


  • Given $150M
  • Operational slightly late


  • Operational on time

New York

  • Operational on time


  • Given $90M
  • System due early 2014

Minnesota, Nevada and Rhode Island

  • Operational on time, but issues connecting with Federal data hub


Sources: http://www.nytimes.com/2013/10/09/us/politics/uninsured-find-more-success-via-health-exchanges-run-by-states

Why build a Government insurance exchange if there already are private ones?

However, private exchanges can now sell ACA-compliant plans WITHOUT subsidies. 34 were added in 2013.
Source: http://www.cnbc.com/id/100925732
Stumbling block: private exchanges also can’t access data hub to determine subsidies.
Source: http://www.cnbc.com/id/101153131

Private exchanges will later be able to offer subsidized plans.

How does the Government perform on other IT projects?
An analysis of Government sector projects indicates that they generally performed better than the non-Government sector projects. The highlights are as follows:

  • The Government projects were 8% more productive than non-government projects. By that we mean that more functionality was delivered per developer hour.
  • The speed of delivery of Government projects is slightly better, (measured by the number of function points delivered in a month)
  • On average, Government projects are being delivered 37% later than their delivery estimate and 22% of projects exceed their cost estimate.

Are there recent, successful Government Web IT projects?


  • How’s My Waterway - Environmental Protection Agency
  • Self-Service Logon Remote Proofing - Defense Manpower Data Center
  • Medicare.gov Responsive Design - Centers for Medicare and Medicaid Services


  • National Broadband Map - Federal Communications Commission/National Telecommunication and Information Administration
  • Supertracker - USDA Food Nutrition and Consumer Services
  • Arlington National Cemetery Explorer  - Arlington National Cemetery

When do Government projects succeed?

  • Incremental delivery
  • Expectations management
  • Team empowerment
  • Strong leadership
  • Accurate upfront estimates