You’ve been there. 2 AM. A blinking cursor. You’re on a government website trying to renew your passport, file an unemployment claim, or access your VA benefits. The page loads slowly. Then it logs you out. Then it asks for a verification code that never arrives. Then it says your password is wrong—but you just reset it. Then a dropdown menu has 47 options and none of them describe your situation.
You close the laptop. You curse the government. You wonder why a country that can send a rover to Mars can’t let you change your address online.
Here’s the uncomfortable truth: It’s not a funding problem. It’s not a tech problem. It’s an incentive problem.
Government modernization is measured by activity—new projects, new contracts, new stack, new dashboard. Not by whether anyone can actually use the damn thing. And when you’re measured by activity, you run in place. Faster and faster. Newer and larger. But you never get closer to the people you’re supposed to serve.
Modernization has become a treadmill: we keep running, but we never get closer to the people we’re supposed to serve.
Let me give you a concrete picture. In 2013, Healthcare.gov crashed the moment the country needed it most. In 2020, unemployment websites buckled under a wave of filings. In 2024, a veteran with a service-related injury still has to print, sign, and fax a PDF from a form built in 2008. Not because the tools don’t exist. Not because people don’t care. Because the system is optimized for something else entirely.
It’s optimized for press releases. For ribbon cuttings. For the phrase “digital transformation” said in a State of the State address. Every new initiative becomes a line item. Every line item becomes a contract. Every contract becomes a product. And every product is measured by how many features it has, not by whether the person who needs it can actually get through.
That’s the treadmill. And it’s not just expensive—it’s dangerous.
Think about the mother on a fixed income trying to renew her food stamps. The farmer navigating a flood insurance claim. The immigrant trying to check their citizenship status. These are not edge cases. These are the people the government exists for. And they are the ones being locked out by design.
You can’t fix a broken system by making it bigger.
So what does the fix look like? It’s not another round of “digital modernization.” It’s not generative AI chatbots. It’s not a new design system with 400 components. It’s the opposite.
It’s fewer systems. Simpler paths. And a hard, measurable, non-negotiable rule: if a vulnerable constituent can’t complete the core task, the project doesn’t get funded. Period.
Some places have tried this. The UK’s Government Digital Service once insisted that every service must be tested with real users, and if it failed, it didn’t launch. That discipline worked. Websites got simpler. Times dropped. Confidence rose. But even that model got nibbled away by politics and vendor inertia.
Why? Because simplicity doesn’t run on a procurement schedule. Accountability does not generate flashy headlines. And telling a department “you don’t need a new system—you need to delete 80% of what you have” is not a four-year contract.
But it’s the only way off the treadmill.
Here’s what I’m proposing, bluntly: tie every modernization dollar to a single metric—task completion by the intended user. Not page views. Not satisfaction surveys. Not “digital engagement.” Did the person who needs help get help? If yes, pay the vendor. If no, claw the money back. And if a project has been running for more than three years without a demonstrable improvement in that metric, cancel it. Not pause. Cancel.
Stop funding modernization. Start funding usability.
Now, let’s address the objection you’re already forming. “But that’s unrealistic. Government is complex. Regulations change. Services overlap.”
I’m not saying it’s easy. I’m saying it’s necessary. Complexity is a choice. Every form, every dropdown, every 2-factor authentication text that goes to a landline is a choice. Right now, we’re choosing to build for the institution. We need to choose to build for the citizen.
The technology exists to do all of this. The hard part is not code. It’s courage.
Courage to say to a vendor: “We don’t want your enterprise platform. We want a single page that works.” Courage to tell a department: “Your initiative is living proof that we’re not serious about helping people.” Courage to end the treadmill and stand still long enough to measure what actually matters.
Because if we don’t, we get exactly what we’re getting now: a system that feeds itself while starving the people it’s supposed to feed.
So the next time a politician brags about their “digital transformation” budget, ask them one question: Can a 70-year-old in a rural town use it in the next ten minutes? If the answer is no—or they start talking about “roadmaps”—you have your answer.
It’s not modernization. It’s a monument.
And it’s time to stop building monuments and start building ramps.
FAQ
Q: Isn't the private sector full of bad websites too? Why single out government?
A: Yes, private companies launch broken products all the time. But they face consequences: users leave, revenue drops, stock falls. Governments have no such market check. There's no bankruptcy for a DMV portal. That's exactly why we need a different, stronger standard—one tied to who actually gets served.
Q: What's the first concrete step to get off the treadmill?
A: Stop starting new projects. Audit the top 10 citizen-facing services and measure how many of their intended users can successfully complete a core task. If that number is below 50%, cancel the project and rebuild it around one single user journey, with funding tied to completion rate—not deliverables or feature lists.
Q: Isn't technology the real solution? Shouldn't we modernize more, not less?
A: That's the treadmill talking. More tech without fixing the incentive problem just adds complexity. Every dollar poured into 'digital transformation' so far has produced more systems, not more access. Real progress comes from subtraction: fewer forms, fewer fields, plain language, and ruthless user testing. The answer isn't more tech—it's more accountability.