“Move Fast and Break Things” Doesn’t Work in Public Sector
Why Silicon Valley Advice Will Derail Your Digital Transformation
“Move fast and break things.” It’s the mantra that built Facebook, Uber, and a generation of startups. In Silicon Valley, speed is everything. Ship the minimum viable product, gather feedback, iterate. If something breaks, fix it in the next release. Users adapt. This advice does not transfer to public institutions. When a startup’s app crashes, users complain on Twitter. When a government portal crashes, citizens can’t access benefits, staff drown in manual workarounds, and trust erodes in ways that take years to rebuild.
The trust problem Public trust in government is already fragile. According to the OECD, only 39% of citizens across member countries report medium or high trust in their government. (https://www.oecd.org/en/publications/oecd-survey-on-drivers-of-trust-in-public-institutions-2024-results_9a20554b-en.html) In the United States, the Partnership for Public Service found that only 23% of Americans trust the federal government — down from 35% in 2022. (https://ourpublicservice.org/publications/state-of-trust-in-government-2024/) Every failed digital service reinforces the narrative that government can’t deliver. And unlike a startup, government doesn’t get to pivot to a new market when users lose faith. As one former U.S. Digital Service leader put it: “When people move fast and things break in government, it doesn’t just have real harm to people, but it has a lasting impact on trust in public institutions.” (https://federalnewsnetwork.com/technology-main/2025/07/how-healthcare-gov-botched-rollout-led-to-a-digital-services-revolution-in-government/)
The Healthcare.gov lesson The 2013 launch of Healthcare.gov remains the most visible example of what happens when government tries to move fast without the foundations in place. On day one, four million users visited the site. Six successfully registered. (https://www.henricodolfing.com/2022/12/case-study-launch-failure-healthcare-gov.html) The problems were predictable: fragmented leadership across multiple agencies, unrealistic scheduling, constant policy changes during development, and contractors awarded cost-reimbursable contracts with no accountability for outcomes. (https://www.brookings.edu/articles/a-look-back-at-technical-issues-with-healthcare-gov/) The site eventually worked. But the reputational damage was done — and it took a dedicated “tech surge” team and the eventual creation of the U.S. Digital Service to stabilize the system and prevent future disasters. Healthcare.gov wasn’t an outlier. Research shows that over the past decade, 94% of large federal IT projects were unsuccessful — more than half delayed, overbudget, or below expectations, and 41% judged complete failures. (https://d3.harvard.edu/platform-rctom/submission/the-failed-launch-of-www-healthcare-gov/)
Why “move fast” fails in institutions The startup playbook assumes conditions that don’t exist in the public sector: 1. You can’t “break” citizen services. When a permit system goes down, construction projects stall. When a benefits portal fails, vulnerable people don’t get support. The cost of failure isn’t measured in churn metrics — it’s measured in real harm. 2. You can’t iterate your way out of compliance. GDPR, accessibility requirements, procurement rules, and data protection laws aren’t optional. You can’t ship now and fix compliance later. The legal and reputational exposure is too high. 3. You don’t control the timeline. Political cycles, budget approvals, and stakeholder alignment create constraints that startups don’t face. A six-week sprint means nothing if procurement takes six months. 4. Your users can’t leave. Citizens don’t have the option to switch to a competitor. They’re stuck with whatever you build. That raises the bar for getting it right — not lowers it.
The alternative: move deliberately, build trust None of this means institutions should move slowly. Slow is its own kind of failure — delayed modernization, mounting technical debt, frustrated staff, and citizens stuck with outdated services. The answer isn’t speed or caution. It’s structure. Start with a baseline. Before committing to a plan, document what exists: systems, workflows, responsibilities, constraints. This prevents building on assumptions. Phase the work. Break large initiatives into smaller deliverables with defined outcomes. Each phase should produce something usable — and a checkpoint before committing to the next. Assign named ownership. Every deliverable needs a person responsible for its success. Not a committee. Not a steering group. A name. Document and hand over. Each milestone should include documentation sufficient for someone else to maintain and improve the work. This protects against staff turnover and vendor lock-in. Measure against baseline. Track what actually changes. If an improvement doesn’t outperform the starting point, don’t ship it. This approach isn’t slower — it’s more predictable. It reduces the risk of catastrophic failure while still delivering visible progress.
What institutions actually need Public institutions don’t need startup culture. They need partners who understand institutional constraints and can deliver within them — on time, on budget, with documentation that survives leadership changes. The firms that succeed in this space aren’t the ones promising disruption. They’re the ones who baseline before they build, phase their delivery, and measure outcomes against real-world impact. Speed matters. But in public services, trust matters more. And trust is built through consistent, reliable delivery — not through moving fast and hoping nothing breaks.
DIGIPART helps public institutions digitalize with structure: phased delivery, named ownership, documented handover, and measurable outcomes. If you’re planning a transformation — or recovering from one that stalled — we can help you find a path that balances progress with stability.
Sources:
OECD Trust Survey 2024: https://www.oecd.org/en/publications/oecd-survey-on-drivers-of-trust-in-public-institutions-2024-results_9a20554b-en.html Partnership for Public Service 2024: https://ourpublicservice.org/publications/state-of-trust-in-government-2024/ Federal News Network / USDS: https://federalnewsnetwork.com/technology-main/2025/07/how-healthcare-gov-botched-rollout-led-to-a-digital-services-revolution-in-government/ Healthcare.gov case study: https://www.henricodolfing.com/2022/12/case-study-launch-failure-healthcare-gov.html Brookings / GAO report: https://www.brookings.edu/articles/a-look-back-at-technical-issues-with-healthcare-gov/ Harvard / IT project statistics: https://d3.harvard.edu/platform-rctom/submission/the-failed-launch-of-www-healthcare-gov/
