High-quality government software starts with efficacy, not efficiency

The emerging Department of Government Efficiency (DOGE) has sparked discussions about cutting spending and reducing the size of government to improve efficiency. But the fundamental problem hampering government software development and delivery isn’t efficiency, it’s efficacy. Cheaper and faster doesn’t matter if the product or service doesn’t do what you need. As Peter Drucker said, “There is nothing quite so useless as doing with great efficiency something that should not be done at all.”

Agencies can only achieve real efficiency by ensuring solutions deliver the desired mission impact. When it comes to software, it’s easy for large IT organizations to lose sight of this, but the mission impact is not automating infrastructure for maximum efficiency or determining how to manage Kubernetes clusters best. Ultimately, the goal in government is to continuously deliver valuable software that does things like win wars, save patients' lives, or provide critical services to citizens.

The question isn’t just how to deliver faster or cheaper, but how can government agencies deliver high-quality, secure software that creates measurable outcomes that drive mission impact?

Lessons Learned from Government Software Factories

The intent of government software factories was to enable the government to continuously deliver valuable software that users love. In theory, software factories promised speed, agility, and mission-critical outcomes for warfighters and end users across the public sector. In reality, software factories have struggled with the same issue plaguing government writ large: prioritizing efficiency over efficacy.

I co-founded the U.S. Air Force’s Kessel Run, the Department of Defense’s first software factory. Like many software factories, we started with the right priorities: shipping valuable outcomes to production quickly and continuously. Early wins proved that agile, iterative practices, paired with user-centered design, could deliver capabilities that directly improved mission success. But as the model gained popularity, many software factories lost focus to innovation theater—appearing productive on paper but failing to deliver measurable value to users and especially to the larger missions of which they are part​​. Instead of competing on outcomes in production, teams competed for resources with PowerPoint presentations and storytelling.

In addition to the lack of accountability and measurable outcomes across many software initiatives, the focus on “DIY” solutions versus commercially available platforms and applications has compounded the premature efficiency problem.  Government stakeholders must let go of the notion that every commercial solution leads to “lock-in.” While every decision has the potential for lock-in to some degree, the ultimate goal is flexibility and maximizing mission impact. Some organizations consume entire budgets on platform activities, which are, by definition, below the mission value line. DOD software factories aren’t in the platform business, they’re in the warfighting business. DOD should only build custom for the capabilities that differentiate it from non-warfighting organizations. Unfortunately, DOD builds its own finance apps and Kubernetes platforms while trying to buy “commercial” warfighting apps–the exact opposite!

The software factory model works but success requires a renewed emphasis on measuring outcomes over effort and incentivizing effectiveness. Metrics must measure real outcomes, not outputs. Focusing on the efficacy of continuously delivering valuable software users love will also drive efficiency. This would also naturally drive towards commercial solutions below the value line, but it’s worth emphasizing that this needs to happen, and fast.

Competing for Efficacy: A New Approach

Fostering competition between contracts, rather than for contracts, is a more practical approach to driving efficacy and efficiency in government software development. Today’s proposal processes fall short, failing to deliver the best products or capabilities with the best value because they prioritize silly FAR compliance, non-value added box checking, and storytelling (lying) over real-world outcomes.

Imagine two government program offices, awarded contracts to develop the same solution, racing to deliver the best capability where success is measured in real mission outcomes: fewer mission failures, reduced costs to the field, or improved operator decision-making. Evaluation would focus on who is delivering, how well they’re delivering, and the impact of continuously delivered, secure, high-quality software. This level of accountability and visibility would force vendors, and government program managers, to prioritize mission-critical efficacy.

While internal competition may seem inefficient or duplicative, the opposite is true. Value is performance divided by cost and schedule–the value of an ineffective system is zero, regardless of how cheap it is to produce. Real competition exposes this and breeds a winning culture so that over time, it’s easier to identify effective and efficient solutions and the government program managers that lead them.

This approach could benefit the Department of Government Efficiency (DOGE). While it seems duplicative at first, the internal competition would enable DOGE to use merit to identify and adopt modern software solutions while eliminating legacy technology and redundant roles. This targeted strategy could result in “efficiency” as a byproduct of effectiveness as competition fosters accountability and innovation. 

Efficacy and Efficiency: The Role of Continuous Authority to Operate (cATO)

The federal government cannot afford disruption. In an era where data security is paramount, ongoing authorization tailored for continuous software delivery — aka continuous Authority to Operate (cATO) — is a game-changer. Speed, stability, and security are interlinked; cATO enhances security and fosters innovation by enabling change without the delay and red tape of the traditional re-authorization process.

Embracing cATO is not about ticking compliance boxes, it's about revolutionizing the speed and safety of software delivery. cATO is modular—reusing components and their corresponding assessments to eliminate redundant work. Instead of re-certifying the same components repeatedly, teams leverage prior authorizations to accelerate continuous delivery and maintain compliance.

The continuous delivery of secure, high-quality software reduces costs and enables faster iteration. By turning compliance into an enabler, cATO adoption allows teams to focus on delivering software that truly matters—faster, safer, and at lower prices.

Putting it all Together: A Path to Efficiency

By better leveraging commercial technology while making smart decisions about lock-in, we can focus effort above the value line. For the government, this means on mission capability, not commodities like infrastructure, accounting software, or performance review software.

Next, by implementing cATO for commercial software platforms, we unlock a faster path to production, with both higher quality and reduced risk. In doing so, we unlock the feedback loops that will help us iterate towards efficacy and mission value.

This leads to the perfect competition platform. Not a competition on paper for a contract, but a competition in production between contracts. Bad capabilities can’t hide in prod like they can in PowerPoint. A byproduct of this is not only more effective solutions, but also more efficient ones. By doing it this way, in this order, we also create a bridge across the valley of death for innovation efforts that can prove real value and outcompete more traditional efforts. Because I started with Kessel Run, it feels appropriate to close with, “This is the way.”