Why the Commercial Sector Should Learn From Government IT (And Not Vice Versa)
Before I joined AWS several months ago, I was the CIO of US Citizenship and Immigration Services (USCIS), one of the 15 operational component agencies of the Department of Homeland Security (DHS). It was my first and only foray into government — prior to that role, I had been a CIO and CEO of commercial-sector companies. How did I suddenly find myself in the public sector? One day, when I was trying to decide where to go next with my career, I happened to read an article about the IT challenges Homeland Security was facing, given its nature as a merger of formerly independent agencies. Something clicked, and it engaged the problem solver in me. I wanted to see if I could help fix government IT. I like hard problems.
It was, actually, a hard problem. Homeland Security had inherited all the challenges of federal bureaucracy, and had added to them layers of policy and process meant to control the sub-agencies that operated under its umbrella. The result was a kind of organizational sludge that resisted change and (perhaps inadvertently) mandated outdated approaches to IT.
Nevertheless, over the course of my time there, we were able to move USCIS to the cloud and to DevOps practices, and to institute a user-centered design approach where we worked backwards from customer needs to capability design. We went from releasing code to users every 6–12 months, to deploying new functionality several times a week, or as much as several times a day. In the process, we reduced our infrastructure costs 75% by moving to the cloud, transformed the culture, and developed new skills for our employees.
To do all this, we had to innovate on many fronts, creating new approaches to QA, security, project oversight, and procurement. We even unleashed the Netflix Chaos Monkey in the DHS production environment to make sure our systems were always resilient.
The more I work with AWS customers, and the more I present at conferences and speak to audience members, the more I realize that the challenges we faced in transforming a US government agency are similar — or identical — to those faced by every private-sector organization trying to pull off an IT transformation. Indeed, the very nature of a transformation means that it will face organizational impediments. Change agents in both the government and commercial sector need to deal with bureaucracy, organizational politics, conflicting demands from stakeholders, oversight mechanisms that were set up for a different world, budget constraints, constraining compliance requirements, general inertia, lack of skills in the new technologies, and the need for cultural change where employees are fearful and perhaps confused.
Think of the government agency as a laboratory where all these problems exist in their extreme form. Then add the fact that everything one does is in the public eye — IT change leaders are second-guessed by the press, Congress, GAO (Government Accountability Office), OMB (Office of Management and Budget), various inspectors general, ombudspeople, … and, well, ultimately the public. For problem solvers, this is, in many ways, the quintessential challenging environment: any solution you devise for cultural transformation must be virtually foolproof, withstand scrutiny, and overcome the mountains of constraints. On the other hand, any impact you make will be magnified and obvious.
Put another way, if an idea works in the government, it is almost certainly widely applicable to commercial organizations. Government is a place to experiment, fail fast, and learn (only don’t describe it this way to Congress).
In several of my future blog posts, I’ll discuss some of what we learned, as a team, trying to transform The Big Bureaucracy. You can also read about some of these ideas in my two books: The Art of Business Value and A Seat at the Table: IT Leadership in the Age of Agility. For this post, though, I’ll just touch on a few high-level ideas for causing change in an environment where change is extremely hard.
Setting the vision. In an environment where change is hard, there will always be fits and starts, setbacks, and unfortunate compromises. One of the big dangers is that, as these setbacks accumulate, employees may lose their way — the transformation becomes just a jumbled set of incomplete initiatives. It’s critical for leaders to avoid this by setting and maintaining a strong, vivid, and compelling vision for where the transformation is heading. My choice of the word “maintaining” is deliberate: it’s not enough to set a vision once, at the beginning of the effort. As Stephen Orban, Global Head of Enterprise Strategy Strategy at AWS, points out in his post on leadership, vision must be reinforced as often as possible.
Moving incrementally. To balance that strong and uncompromising vision, the organization must drive for small, incremental wins. No matter how severe the obstacles, the transformation must start moving in the right direction on Day One. And moving in the right direction does not mean having meetings or preparing presentation decks. It means concrete changes that affect outcomes. We always asked, “What is the smallest thing we can do immediately that will move us toward the vision?” Big vision, small execution.
Provoke and observe. I love this term, which I borrowed from an article by Christopher Avery (he says he borrowed it from somewhere as well). Just as Agile approaches promote inspecting and adapting, transformation in a difficult environment requires provoking and observing. You don’t know all the impediments you’ll face until you try something and see what the reaction is. And you don’t know which changes will turn out to be easier than you expect until you try experiments. The important thing is to provoke in a considered, deliberate way — you want to provoke in a way that will yield the maximum learning.
In future posts, I’ll describe some techniques we developed for overcoming particular impediments. For now, I hope that these high-level techniques will help you get started on your transformation journey. They might very well — they’ve been tried and tested in the government!