We talk a lot about shiny new tech, but the most boring thing sometimes makes the biggest difference, user experience. The way governments design services determines whether someone gets help, or gives up and goes hungry for a month. I had a long conversation with Markus Poschenrieder about this exact failure mode, and it turns out it is less about code and more about choices, organization, and tone.
The problem is simple to state, brutal to live, and oddly expensive. Germany has dozens of social benefits, with application processes that feel like stumbling through a bureaucratic jungle. Markus put it well, we guide through this complete jungle, but the core issue is people simply do not know what they are entitled to, or they attempt to apply and fail. The numbers are grim, 35 to 60 percent of people entitled to benefits do not claim them, and for one major program, roughly 1.2 million applications are processed annually. Of those, about 90 percent are submitted wrong, incomplete, or unreadable. Every misfiled application is wasted time, money, and patience, and at about €260 of administrative cost per living allowance application, that waste accumulates fast.
There is a term everyone in the system knows, administrative ping pong, and it is a perfect description. Someone applies for one benefit, gets sent to another office, collects another set of documents, repeats the ritual, and after hours of work, often gets denied or still stuck. Markus told a story of a man cycling through three departments before anyone could actually help him, a pattern that turns bureaucracy into an endurance sport.
Officials spend three hours on average reviewing an application, while the law assumes 90 minutes. Those are not just numbers, they are people, and they are the reason municipalities are exhausted. The administration faces massive staffing gaps, hundreds of thousands of roles unfilled, which means digitalization is not indulgent, it is necessary. Not optional.
This is where good UX matters, because poor forms and vague guidance blow up those times, inflate denial rates, and push human effort into triage instead of service. Instead of fixing the root causes, many systems dump unreadable scans in a bucket and hope a human will rescue the mess. We can do better than that.
User-centric design here is not a buzzword. It is a checklist of practical changes that immediately reduce friction.
The immediate payoff is twofold, citizens get what they need faster, and municipalities save time and money. That is about as win win as government projects get.
AI can be useful, but it has to be careful around the legal boundary between providing information and offering legal consultation. Germany has a Legal Services Act that restricts who can give case-specific legal advice. Markus emphasized the distinction, when it comes to single-case counseling you need a lawyer, but rule-based calculators and informational guides are fine.
Use AI for the heavy lifting it excels at, parsing documents, validating forms for inconsistencies, producing plain language explanations and even accessible language conversions. Do not let AI pretend it is a lawyer, do not let it offer a definitive adjudication on a complex, ambiguous case. The sweet spot is explainable, sourced, rule-based assistance with a human in the loop for edge cases. If a chatbot clearly states it is providing general information, cites its sources, and flags ambiguous cases for human review, it is a useful helper rather than a liability.
One small, misunderstood lever is tone. Government letters often read like threats, and dozens of people are scared of opening official mail. This fear is absurd for a democratic welfare system, but it is real, and it drives avoidance. If the state wants people to engage, communications must be human, calm, and unintimidating. A digital account that notifies you in plain terms, with friendly copy and clear next steps, will beat a scary letter every time.
A better system would not make people beg for what is already their legal right. Ideally, the state would pre-identify entitlements from existing data and ask, do you want this support, rather than forcing a citizen to start from zero. That requires interoperability and political will, but it is a simpler UX problem than the one we have accepted for decades.
This is not about inventing another flashy startup. It is about focusing on the basics, centralizing information, making forms conditional and accessible, and using AI as a helper, not the final judge. Design choices are policy levers. If the goal is to make sure benefits reach the people who need them while freeing up administrative time, start with user-centric design. It costs less than the status quo, and it saves human dignity along the way.
Commenting Rules: Being critical is fine, if you are being rude, we’ll delete your stuff. Please do not put your URL in the comment text and please use your PERSONAL name or initials and not your business name, as the latter comes off like spam. Have fun and thanks for your input.
Join a growing community. Every Friday I share the most recent insights from what I have been up to, directly to your inbox.
