Delivery-driven Government

Jennifer Pahlka
Code for America Blog
8 min readMay 30, 2018


Principles and Practices for Government in the Digital Age

Public trust in government is at historic lows in the United States. The most obvious culprit is partisan gridlock in Washington, but perhaps as important is government’s difficulties delivering on its promises. New politicians often arrive with new policies but still fail to get results. There is an increasing failure of implementation, which has only gotten worse as society has gone digital. Only sixteen percent of senior fellows at the National Academy of Public Administration believe government is any good at designing policies that can be implemented at all. (Bill Eggers, If We Can Put a Man on the Moon, 2009.) Too often dependent on unreliable surveys and snapshots of historical data, governments are unable to identify missteps until years after the fact, if at all. It’s like asking a pilot to fly a transcontinental flight with only after-the-fact, unreliable estimates of her airspeed, heading, and altitude instead of the panel of instruments with constantly updated data and tested checklists to reduce accidents and errors that modern pilots rely on.

There has to be a better way.

There is a better way.

In the digital age, we have the ability to give the “pilots” of our government programs the necessary instrumentation to see where they’re headed and course-correct along the way. This is the way modern technology platforms operate: measuring and using information about user satisfaction and achievement of intended results to modify and improve their services on an ongoing basis. Governments that don’t take advantage of these powerful new tools are sacrificing their relationship with the people who must use their services, their ability to achieve the goals of the program, and most importantly, breaking promises by failing to deliver.

Today, as Tom Loosemore, formerly of the U.K.’s Government Digital Service, has said, most “policy is educated guesswork with a feedback loop measured in years.” Our governments must rethink how they serve the public using the tech-enabled feedback and iterative improvement mechanisms that drive some of the most successful consumer and enterprise services today, but in a public sector framework that protects the privacy and agency of the people served. Developing policy around a deep understanding of the needs of its intended beneficiaries, continually informed by insights derived from the delivery of the services based on that policy, reduces the guesswork and increases ongoing situational awareness.

The movement to modernize government technology has been focused on the delivery of government services using modern technology and best practices. But that is only half the solution; now we must also learn to drive policy and operations around delivery and users, and complete the feedback circuit. Only then can we effectively achieve the goals government policies intend.

Delivery-Driven Government Principles

Understand and Meet User Needs

Delivery-driven government sees the institution and its processes through the eyes of its users, meaning the members of the public whom government services are meant to benefit (usually individuals and families, but also businesses or organizations). Government needs — compliance with various regulations, the need to operate within silos, the need to use complex or legal language — often conflict with user needs, and governments must find practical ways to promote a culture of understanding and meeting user needs within the constraints of the public sector context. When government prioritizes its own needs, its own silos, its own processes, we do not live up to the ideal of “government of the people, by the people, for the people” upon which this country was founded.

Real-Time User Data, Not Years-Old Estimates

Teams focused on delivery to users need actionable information about the services they operate so the implementation of policy can be adapted with situational awareness. Instrumenting delivery systems to provide that insight from users creates a technology, policy, and operations feedback loop that connects intention with implementation.

If there were a line of angry people out the door in a government office or City Hall, staff or leadership would be forced to address it. Yet similar angry users of government online services can easily go unnoticed. Can government teams see how many people are trying to use an online service, and can they tell when it’s broken? If users get to a particular question in an online form and then abandon it, that data signals a problem with the question, the service, or the policy that requires it. It is a signal from delivery that should drive improvement toward better outcomes. If we’re serious about improving government services, we must make the intangible tangible, the invisible visible.

Iteration, from Intention Through Implementation

Knitting together technology, user data, service design, operations, and policy into tight feedback loops creates the conditions to constantly adapt programs and improve outcomes. By developing low-level policy and regulation in iterative cycles alongside tech teams, rulemakers can see how their best guesses actually operate in the real world and adjust before initial rules are finalized. By modernizing and instrumenting the systems that deliver these services, governments can continually adjust the technology, operations, and low-level policies of a program as needed in service of the desired outcomes.

Practical Steps Toward Delivery-Driven Government

Delivery-Driven Government requires modernizing government technology to enable instrumentation of systems, which is easier said than done. But it also requires reorganizing teams and sometimes reporting structures, strengthening the skill sets in government around technology and design, and changing a set of practices in procurement and hiring. Here are some practical steps leaders can take to make modernization efforts easier.

Build Multidisciplinary Teams with Tech and Policy at the Table from the Start

The days of throwing policy, requirements, or specifications over the wall to the next team are gone (or need to be). As Clay Shirky has said, “the waterfall method amounts to a pledge by all parties not to learn anything while doing the actual work,” and today even the most thought out plan rarely survives contact with the real world. Creating policy that can be well-implemented requires technology and design teams to have a seat at the table as early as possible (even during the legislative phase), in order to conduct user research and inform the art of the possible. Successful teams represent the disciplines of technology, design, policy, subject matter expertise, communications, law, privacy, security, etc., and break down the silos between functions, departments, and even jurisdictions to truly work from the point of view of the user.

Organize for Ownership and Outcomes

In most government entities, technology is a functional silo, and technology-related functions sit under a Chief Information Officer, whose duties and required skill sets can be sometimes unhelpfully broad. These areas of responsibility can include: citizen-facing digital services, back office software (day-to-day core services like email, human resources management, and accounting), mission IT (applications that run the internal processes of departments and agencies), and technology infrastructure. Recently, some governments have added other technology leadership roles (Chief Technology Officers, Chief Innovation Officers, Chief Data Scientists, etc.), sometimes in the reporting line of the CIO, sometimes outside of it, often with the goal of bringing in new skill sets and approaches. In the end, there are many different structures that can work, but leadership roles and the requirements of those positions must map to clear responsibilities and goals, and the most successful delivery-driven government efforts tend to have technology leadership reporting to the highest levels of the organization, not buried within it.

Just as important, it is frequently incumbent on technology leadership to lead the way toward matrixed organizational structures that allow for multidisciplinary teams to have clear product leadership across silos. It’s not easy to establish the kind of ownership over program outcomes that is needed to enable continuous improvement, but it is possible, and it requires a rethinking of how government organizes its workforce. Silos work better when functions are clearly defined and skills and competencies are mature. Right now, we need partnerships and collaboration to build collective ownership and capability.

Recognize and Elevate the Talent You Have (Even as You Bring in New Skills)

Government executives and newly elected leaders are sometimes convinced that the way to foment digital progress is by bringing in a completely new team and buying big new technology solutions. The truth is, many of the biggest opportunities you have are already in-house but haven’t yet been activated. Your task is first to understand what capacity exists in current technology and processes, and then identify where latent talent in the organization can be scaled to impact. It’s common that there are dozens or even hundreds of civil servants who both know how the system works inside and out and have been trying to change it to improve outcomes for years. While it’s likely that building a delivery-driven government will require hiring new talent, failing to recognize, elevate and support the champions you already have (and who have the battle scars) as a new generation of leaders can set a change agenda back years, even decades.

Change Hiring and Human Capital Practices

While it’s important to elevate career public servants who can effectively support delivery-driven government, it’s a rare government entity that doesn’t need many more technology and design staff to achieve their goals. It’s equally rare to find recruitment, hiring and retention processes fit to the purpose of hiring and retaining the talent needed. In many cases, the appropriate job classifications and evaluation criteria don’t even exist, not to mention the pipeline processes needed to attract and effectively evaluate candidates. Executive-level investment in these areas is a surprisingly good return on investment, and frequently necessary to achieve the results that many leaders are looking for. There’s no need to completely reinvent the wheel; other cities, states, and federal agencies have been trying and succeeding in new models. Look to Code for America’s Talent Initiative and others to help you find the right peers to borrow from.

Change Procurement Practices

As with hiring and human capital practices, many legacy government procurement practices don’t support delivery-driven government and aren’t sufficiently geared toward rewarding outcomes. Procurement tactics must be adapted to fit technology strategy, not the other way around, and leadership will be needed at various levels to validate the procurement practices that support this. As with talent strategies, there’s plenty of inspiration and practical examples to be borrowed from other governments.

Don’t Innovate Where You Can Borrow and Adapt

Too often, leaders give areas where government is weak an “innovation” label in order to recognize that practices must change. This can be useful, but can also backfire. In some cases, there are tried and true playbooks and approaches that not only work well outside government, but have also been proven to adapt well to a government context. These playbooks may be more helpful to think of as a change in approach, in part because they allow those who don’t think of themselves as innovators (or aren’t seen that way by others) to participate in and even help drive the shift. In areas like performance management, there are a wealth of resources to draw on, including the What Works Cities initiative from Bloomberg Philanthropies. For digital services, there’s the CIO Playbook and others. True innovation is, of course, needed in government as we face never before seen challenges like regulating self-driving cars, but take care to demystify and democratize the solving of problems for which we already have pretty good tools.

For more on delivery-driven government, sign up for our newsletter at or email

Thank you to:

Mike Bracken, who inspired “delivery-driven government” with his blog post: “The Strategy is Delivery. Again.”

Jake Solomon, whose talk at the Code for America Summit in 2015 further cemented the importance of delivery.

David Eaves & Ben McGuire of the Harvard Kennedy School, for working with us on advice for incoming mayors and governors, which influenced this paper.



Jennifer Pahlka
Code for America Blog

Author of Recoding America: Why Government Is Failing in the Digital Age and How We Can Do Better, Fellow at the Federation of American Scientists