You know that feeling when a government form looks like it was designed to test your will to live, rather than actually help you get support? I had a long conversation with someone who has been living inside that chaos, and it confirmed something obvious, but painful, social support systems are suffering from legacy software, fractured data, and organizational inertia. The result is people left without benefits they are entitled to, administrators buried in avoidable work, and a system that spends more time denying help than delivering it.
Here are two numbers that should make you uncomfortable. 90 percent, the share of living allowance applications that arrive incomplete or incorrect. €260, the average internal cost to process a living allowance application in Berlin, when you factor officer time and overhead. Couple those with 1.2 million beneficiaries, and you stop thinking about spreadsheets and start thinking about waste. The system is not just inefficient, it is actively expensive and harmful.
This problem is a mix of messy rules and messy implementation. There are dozens of benefit types, rules with 35 different income categories, and household situations that could make even a logic professor weep. People get bounced from one office to the next, a phenomenon we heard called administrative ping pong. The state expects citizens to navigate a jungle, while each department shrugs and says, that is not my responsibility.
The obvious answer is usually obvious for a reason. Centralize the information, make it findable, and make forms behave like decision trees, not trick courses. Instead of 50 questions staring you in the face, ask one question at a time, show only what matters, validate entries in real time, and pre fill repeated fields. The aim is not to be clever, it is to stop scaring people away.
There are some simple UX fixes that cut error rates dramatically. Show contextual lists of the exact proofs someone needs, not vague instructions about income proofs. Use document scanning that produces readable files, so authorities do not spend hours dealing with blurry phone photos. When a user reaches a showstopper inside one application, offer to switch to the right application and transfer the data. These are not sci fi fixes, they are engineering and product design.
This is where the politics of procurement comes in. Many municipalities are tied into lengthy contracts with legacy software providers who become effective gatekeepers. Change happens slowly because these systems are woven into municipal work. There is also fear, if we replace the system what else breaks. So startups that offer better interfaces face two problems, convincing procurement officials to try something new, and proving they will not make the whole office implode.
There is a strange market dynamic here, if a startup makes a department less necessary, that department is unlikely to buy the product. Yet at the same time municipalities are short staffed, and the public sector needs help. One clear fact is zero churn, once a municipality signs a competent provider, they do not leave. That should be attractive to innovators and investors, if you can cross the trust barrier.
AI can help a lot, but there is a legal nuance you cannot ignore. Under the Legal Service Act, giving legal advice is regulated, you cannot act as a lawyer unless you are a lawyer or authorized. That means a public facing system must stick to information, not single case consultation. As the people I spoke to put it, we are informing, we are not consulting. That sweet spot lets AI power calculators and guided decision trees, but it stops short of telling someone definitively what to do in a legally binding sense.
Practically that looks like clear, rule based calculators and explainers, with human in the loop for edge cases. It also means benchmarking and transparency are non negotiable, show sources, show logic, and make sure the user never leaves thinking they received legal counsel from a bot.
There are three levers that change everything, and none of them are flashy. One, interoperability, APIs and shared formats so systems can talk to each other instead of reinventing the wheel. Two, user centered design that reduces non take up rates, the percentage of people entitled to benefits who never claim them. In Germany this is 35 to 60 percent, an absurd figure given the social purpose of these programs. Three, sensible procurement that encourages modular systems, not vendor lock in.
If you want to go a level higher, the ultimate ideal is proactive state action. Instead of citizens begging for benefits, the state notices life events and offers support. Some countries already do this, when a child is born payments can start automatically. That requires integrated registries and trust, it also removes the stigma of sounding like a beggar when you ask for help.
This is not a technology problem alone, but tech is the lever you can use fast. Validation, better scanning, tailored checklists, and guided flows cut processing time, reduce mental load, and keep people from falling into poverty because of paperwork. Fixing that also reduces needless administrative labor and helps municipalities steer resources to complex cases that actually need human judgment.
If you are an executive in local government, your decision does not have to be radical overnight. Pilot, measure, and scale what actually reduces manual processing and increases correct applications. If you are building solutions, do the homework, know the legal lines, and design for the humans who have to both use and process the forms. For everyone else, stop assuming complexity means it must be this way. It does not.
Change will not be painless, but the alternative is quietly expensive and morally annoying. If we want public administration to work for people, we need to stop glorifying complexity, and start building simple, honest systems that actually deliver.
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.
