Can Nations Operate Like Startups?

By: Sophia Martinez, Jude Canady, Franco Garcia

Imagine a world where DARPA calls you — not the other way around. Where you don’t need a SAM registration, 19 acronyms, and three past performance citations to show the government something new. Where “innovation” means working software instead of 400 pages of boilerplate and one slide deck with an eagle watermark.

We know. It’s a fantasy–at least for now.

What If We Flipped the Incentives?

Let’s try a thought experiment. What if the system actually rewarded:

• Speed to prototype instead of speed to proposal?
• User testing over technical volumes?
• Clarity instead of compliance?
• Impact per dollar over dollars obligated?

What a world it could be.

A Nod to Human-Centered Design

Instead of fixating on contract vehicles and lengthy paperwork, agencies could begin by defining who a solution is for and what problem it must solve. They could prioritize rapid testing in real‑world sandboxes rather than requiring monumental RFIs, five‑year roadmaps, and endless Gantt charts. In that environment, startups would sprint to build for actual users, leading to more experimentation, faster iteration, and systems that truly work for the people who rely on them.

Human-centered design, an approach pioneered by IDEO, offers powerful insights that could transform the government contracting process. Rather than centering processes solely around regulatory compliance and documentation, a human-centered approach would position contracting officers and other acquisition professionals as vital end-users whose needs, workflows, and decision points shape how systems and tools are developed. This could lead to more intuitive interfaces for risk assessment, options for safe-to-fail pilots before committing to full procurements, and mechanisms for easily integrating new experimental tools as needs evolve — mirroring the adaptability developers experience when working with APIs. By making the process more user-focused, government contracting could shift from a static operation to a more flexible and responsive system.

Integrating human-centered design principles into government procurement would also support the adoption of iterative practices such as rapid prototyping, hypothesis testing, and transparent feedback loops. These methods, proven in product and service design, would equip contracting professionals to quickly test new approaches, respond to feedback, and make data-driven adjustments throughout the acquisition lifecycle. Reforming procurement in this way would not require abandoning essential structure or oversight. Instead, it would introduce space for responsible risk-taking and continuous improvement, fostering a culture that can adapt to emerging technologies and changing mission demands. Ultimately, leveraging the insights of human-centered design could make the government contracting process more agile, resilient, and efficient, benefiting both acquisition teams and the missions they support.

A New Contracting Model, Powered by Curiosity

Under this new model, agencies don’t wait for an annual procurement cycle to explore new technology. Instead, they hold dedicated exploration budgets tied directly to pressing operator needs. They run fast, low‑cost bake‑offs between tools to see what actually works in the field. They publish the problems they need solved — without dictating solutions — and invite builders to prove, not pitch, their capabilities. Success is measured by demonstrated progress in real‑world environments, not by who can best mirror the language of a 100‑page RFP. Proposal timelines shrink from three‑month marathons to three‑week prototyping sprints. Gatekeeping through acronyms and opaque processes gives way to open calls, with clear criteria and decision paths.

We know this isn’t how federal procurement typically works today. But it could be — if only in targeted areas. Making it happen would require agency champions who value experimentation and are willing to sponsor it. It would require building guardrails so that trying something new doesn’t end in career‑ending failure. It would require vendors who lead with working solutions rather than glossy sales decks, and procurement platforms designed with the end users of innovation — not just the contracting officers — in mind.

This isn’t about tearing down the Federal Acquisition Regulation. It’s about creating deliberate side doors — smaller, faster pathways that enable government organizations to explore, test, and adopt transformative capabilities before they’re obsolete.

Changing of the Guards

The government procurement process wasn’t designed to kill innovation — it was designed to reduce risk, prevent fraud, and ensure taxpayer dollars are well spent. But in trying to mitigate every risk, it often rewards the same entrenched vendors moving at a slow, predictable pace. For years, startups saw the process as a Kafkaesque maze.

That’s starting to change. Defense‑tech startup Anduril has rapidly expanded its footprint, winning a $642 million Marine Corps counter‑UAS contract, a major Air Force prototyping deal, and a three‑year agreement with U.S. Special Operations Command. It’s even set to take over development of the Army’s $22 billion IVAS mixed‑reality program from Microsoft. Similarly, AI company Anthropic launched Claude Gov, an AI model built for federal agencies, and in mid‑2025 won a $200 million Department of Defense contract alongside other leading AI labs.

These breakthroughs show that startups are no longer occasional outliers — they’re becoming key players in federal procurement. These companies have demonstrated that smaller, faster-moving firms can deliver mission‑critical capabilities at scale. Combined with policy efforts like the Department of Defense’s “Replicator” initiative to accelerate procurement, these wins point to a broader shift: the government is actively seeking ways to bring more innovative companies into the fold, not just to modernize technology, but to stay competitive in a rapidly evolving geopolitical and AI landscape.

In the Meantime: Build Anyway

At 5I, we prototype like this world already exists. We build quickly, test with real users, and improve based on feedback. Sometimes that means we’re too fast for the system. Sometimes we find a champion who gets it. But every time we ship, we learn something useful because even if the incentives don’t line up yet, the needs still do. We believe in meeting these needs — with tools that are both cool and critically needed.