HELP! I need somebody…Why the ACA Rollout team(s) should have listened to the Beatles


Henry Chao demonstrated that sharing salient information expediently, specifying your preferred timeline, personnel requirements, and needs, and calling attention to trade-offs when new priorities arise can play a role in identifying a failing or failed project.

It’s easy to see why sharing (the right) information to the right people (equipped with authority or access to other folks with authority) can help reverse a project careening toward failure. In general, information sharing can mitigate poor developments from becoming a disaster. (At least that is my hypothesis!). Case studies from Eric Rosenbach’s course on cybersecurity echo this sentiment and, in part, validate the need for sharing relevant information. Consider the case of the Office of Personnel Management (OPM) hack. The Director (below) and her leadership team roundly failed to keep Congress accurately apprised of the data breach and the steps (or lack thereof) the OPM took to limit its severity.

Caption: You might assume this is a picture of OPM Director Archuleta pledging to tell the whole truth. And, though that’s definitely what she pledged to do, a more accurate description of this shot would be the Director agreeing to tell Congress very little about the very few things she had done to prevent the breach.

By limiting information sharing, Director Archuleta and others gave Congress a false sense of security. If Director Archuleta had been more upfront it is likely that Congress and other regulators could have taken or enforced more significant responses to the fixable problems that facilitated the breach. This thinking receives support from Janis’ study of groupthink. He makes clear that insufficient information search (and, by extension, insufficient reception of information) can lead to all the pitfalls tied to groupthink. What’s left to be tested in each situation is how to appropriately format the information, the correct timing of delivering that information, and sharing that information with the right recipient. Sadly, Henry tried to share information with Todd Parks. But Parks seemed to be the wrong recipient — according to Henry, he never delivered the relevant information to the President.

Henry may have found it easier to share information if he had distributed his preferred timeline and desired resources with the right people — meaning people with authority and the power/will to act on that authority — from the start. The timeline Henry presented to class showed revealed specific dates and staffing details that, if distributed to and endorsed by leadership across the involved agencies, could have given a stronger basis for calling for delays in the release. For example, he called for 95 percent of staffing to be completed months before that threshold was actually crossed. If his timeline had been known and adopted, then he could have more fervently argued for an extension based on insufficient staffing.

Finally, Henry made clear that the serial introduction of new priorities — such as the roll-out of a communications preference site — frustrated attempts to stay on target with the actual final product. Staffers on future projects should more actively flag when new “priorities” are outside the scope of the main project. By raising more awareness about the intensiveness of trying to reach new goals, staff will again be better situated to compel leadership to either stick to the original goals or grant an extension to accommodate yet another new priority project. This hypothesis could be tested at low cost and with ease. Staff could keep a list of top priorities in a visible place and require project leads to physically alter those priorities if anything shifted. This somewhat clunky process would allow leaders to see just how another priority would impact other efforts.


The procurement team bears a huge responsibility for the outcome. As evidenced by the screenshot below of the GAO’s report, the contract contained major errors.

For instance, the contract called for services without having a clear understanding of how it would assess those services. Additionally, the contract provided poor incentives and introduced methodologies unfamiliar to CMS staff. As pointed out in class, specificity in contracting can come with blessing and curses. Dave pointed out that having a real North Star for the product should guide how you get to that outcome rather than vice versa. So the procurement team should not have all of the blame…Lynsey’s example of the interaction between the budget and procurement teams illustrates that there’s an entire infrastructure behind poor outcomes.


To alert other members of Health and Human Services (HHS) I would have encouraged key stakeholders to host small, regular tests of the portal with HHS leadership and project managers. This testing process would have made HHS leaders more bought into the results of the project as well as more aware of the current status of the project. This additional attention paid to the “customer” aligns with Agile thinking.

Notably, though, the benefits of pursuing Agile depends on customers not only being brought into the process but actively engaging with it. I am not convinced that HHS top officials would have understood the implications of failed tests and, even if they did, that they’d be willing to spend additional resources to remedy the issues.

So I would have conducted a small test of the portal with a group of constituents in swing states like Ohio and recorded the results. Then I would have invited both White House (WH) representatives and HHS leadership to watch the consumers struggle. By placing these tests in political terms, the WH and HHS would have been more likely to understand that releasing a poorly functioning portal would spell more trouble than the alternative.

Student at the Harvard Kennedy School and UC Berkeley School of Law. Apolitical Contributing Author.