The three mistake-ers : 3 things I learned from Henry Chao and healthcare[dot]gov
Updated: Feb 14, 2019
Introduction
There are several things that stood out to me in the analysis of healthcare[dot]gov over the last few classes from the readings, from discussions in class and from Henry Chao’s interaction with us. I think broadly they fall under the brackets of policy, power and passion.
Not being familiar with the inner workings of large scale IT implementations in general or the US healthcare system in particular, I automatically assumed that this failure was a mixture of incompetent technical staff combined with textbook public-sector insouciance.
But on closer examination, the issues are much more complex and a part of me is surprised that the team was able to implement, even if almost barely, an obscure legislation while battling a massive industry.
In the following section I've attempted to analyse the reasons for failure with possible ways to fix it. I've also included ELI5 sections that stands for 'Explain Like I'm 5' which is an extremely popular subreddit.
Lessons
1. Policy: Build the entire minimum-viable product through a constant process of creating, testing, iterating, expanding and testing again. If any member of the team notices that sections are being built with a “to-be-tested-later” attitude, they must be trained to immediately call it out and have structures readily available to signal this. I myself am more aware now of projects that are iterative vs those that are waterfall-like.
There must be a policy/law requiring vendors to report more frequently and have procurement policies that are conducive to this method of building instead of the regular procurement process.
Each individual team needed to go through a rapidly iterative process and gradually equip itself to handle scale, stably. This could mean a roll-out that is region-wise or state-wise, incrementally.
This doesn’t mean doing it state-by-state, but definitely testing the first 10,1000, 100000 users before rolling it out.
I have seen great frameworks in design thinking, ADDIE and human centric design that stress on iteration, testing and evaluation before scaling. We studied them extensively in David Dockterman’s class and Bill Wisser’s class at HGSE.
ELI5: Imagine you start a cookie factory. you need to test the first few cookies before baking a batch and test a few batches before going into large scale production. Even then, test batches periodically and look for ways to improve – constantly. Ideally this process of testing and iterating should be made a part of your cookie factory's DNA. If you ever see lots of cookies going out without undergoing testing, that's a signal to change course!
2. Power: It’s important to distinguish the people in the team who are in leadership positions from those who have the power to make decisions. At multiple stages in the healthcare[dot]gov saga it’s clear that the ones who knew what was going wrong (Eg: Henry Chao) were raising red flags to those who held leadership levels (Eg: Todd Park). These folks however, either didn’t have the heart to pass the information up, or weren’t in a position to make decisions. They were acting as postmen at best. And if these people in leadership positions ever felt their power threatened in any way, they would start to exclude naysayers from critical meetings. This leads to a greater gap between decision makers and important opinions.
It is of immense importance that people who see things going wrong be able to present their opinions objectively to decision-makers without playing complex games of power. Manoeuvring power is time consuming and shifts focus away from the core.
The best way to solve this was to have a dashboard / concern system where people could post their concerns that was visible to a decision maker.
ELI5: If you or someone on the team senses that the cookies may go bad after production, it’s important that you know who to address your concern to. One should know whether to fix the oven or speak to the CEO. If you don't know who to talk to when something may be going wrong, that's another signal that you need to change course!
3. Passion: The energy that is needed for large-scale projects is immense. Often this energy stems from spectacularly passionate individuals who channel it out from the system. Along with this passion also comes an incredible amount of emotional attachment to the project. The larger the scale and potential impact, the more increased are expectations, the greater the media attention and more intense the scrutiny from allies and enemies. In this electric atmosphere, teams that are led by charismatic leaders can often lose track of reality in the pursuit of doing something incredible. Once formed, members of these groups can go to great extents to protect the identity of the group.
In the case of healthcare[dot]gov, the entire Affordable Care Act had so much energy, debate and politics surrounding it that the promoters of the act and the portal believed that they had to and could bring the project to life at any cost. Even if the cost was to turn a deaf ear to real-world limitations of technology and deadlines. (Imagine how bad it would look if the Republicans found out there were delays in the project. Right?)
Perhaps the way to tackle this was literally to sit down with the president himself and explain possible consequences of the roll-out and alternatives.
ELI5: You think your yogurt-ketchup topped cookie is the best one in the world. When someone tells you that the cookie tastes good, but the yogurt-ketchup combination is disgusting – you stop talking to them. Listen to this feedback! It’s incredibly important. If you're avoiding people who are telling you how to become better, this is another signal to change course!
Bibliography
Janis, I. L. (1972). Victims of groupthink: A psychological study of foreign-policy decisions and fiascoes. Oxford, England: Houghton Mifflin.
Brill, S. (2015). America's bitter pill: money, politics, backroom deals, and the fight to fix our broken healthcare system. First edition. New York: Random House.
Levinson, D. R. (2016). HealthCare[dot]gov: Case Study of CMS Management of the Federal Marketplace.