> ## Content Index
> Fetch the complete content index at: https://www.siliconsnark.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Deep Dive: Germany’s Startup Scene Broke Records. The Notary Is Still Reading Page 47.
- URL: https://www.siliconsnark.com/deep-dive-germanys-startup-scene-broke-records-the-notary-is-still-reading-page-47/
- Published: 2026-08-30T02:30:08.000Z
- Updated: 2026-08-30T02:30:08.000Z
- Description: Germany’s startup scene is setting records in AI, robotics and defense while founders battle notaries, thin growth capital and procurement fog.
- Author: CircuitSmith
- Tags: Deep Dive, Startups, Germany, Venture Capital, Deep Tech, Europe

[This post from Patrick Collison](https://x.com/patrickc/status/2093717019896197186?ref=siliconsnark.com) made me think I had to do a deep dive into the state of the German startup scene.

Stripe’s co-founder had met a German founder and asked whether all the stories about building startups in Germany were exaggerated. “No, they’re understated,” came the reply. The founder then described spending an entire day listening while a notary read aloud a 90-page investment contract, a procedure he said was mandatory under German law. The notary charged €30,000\. That was company number one. Company number two, with the efficiency one develops after being professionally read to for eight hours, was incorporated elsewhere.

It is a perfect internet story because every element feels engineered in a lab to torment a Californian. There is a startup, a document, an ancient formal ritual, a five-figure invoice and a statutory section number. In Silicon Valley, a founder closes a round from an airport lounge by tapping DocuSign on 4 percent battery. In Germany, a state-licensed professional performs the shareholders’ agreement as a one-man audiobook, and the company buys the premium edition.

The irritating part is that the story is not some meme assembled from lederhosen and DIN standards. German law really can require the relevant instrument to be read aloud in the notary’s presence. Venture documents really can become trapped inside that instrument because they contain share transfers, capital measures, pre-emption rights or other promises that need notarization. Fees really are set by law and rise with transaction value. A €30,000 bill for a large financing is possible.

The equally irritating part is that “German startup scene strangled by notary” is now hopelessly incomplete. Germany recorded [3,053 new startup formations in the first half of 2026](https://startupverband.de/media/file/rs-nextgenerationstartupneugruendungenh12026?ref=siliconsnark.com), more than in all of 2024 and the strongest half-year since the data series began in 2019\. One-third had a clear AI connection. Industrial startup formations more than doubled. German startups raised roughly €5.1 billion in venture capital, while robot maker NEURA announced a round of up to $1.4 billion, defense company Helsing raised $1.8 billion and fusion startup Proxima pulled in €411 million.

Germany, in other words, is conducting two experiments at once. The first asks whether Europe’s largest industrial economy can turn world-class research, manufacturing expertise and a suddenly urgent appetite for sovereignty into a generation of globally important technology companies. The second asks how much administrative scar tissue a founder can develop before moving the holding company to Delaware.

This is not a story about a country with no startups. It is a story about a country that is unusually good at producing the ingredients of hard companies and unusually talented at placing friction between those ingredients. The German startup scene is healthier, broader and more consequential than the jokes suggest. The jokes, unfortunately, have excellent documentation.

## The Notary Will Now Read the Internet

Let us begin with the audiobook.

[Section 13 of Germany’s Beurkundungsgesetz](https://www.gesetze-im-internet.de/beurkg/%5F%5F13.html?ref=siliconsnark.com), the law governing notarization, says the notarial record must be read aloud in the presence of the notary, approved by the parties and signed by hand. This is not a quaint customer-service option for founders who miss bedtime stories. It is a formal safeguard intended to ensure that everyone understands a legally consequential transaction and that a neutral official verifies it.

That rationale is less ridiculous than Twitter makes it sound. Startup financings redistribute control, economics and sometimes the company itself. Investors can acquire vetoes, liquidation preferences and rights whose significance is difficult to explain with a celebratory GIF. Germany’s civil-law tradition would prefer evidence that the humans involved knew what they were doing.

The problem is scope. A German GmbH’s capital increases and transfers of shares require notarization, and venture agreements commonly bundle those acts with shareholder arrangements, drag-along clauses, vesting, options and obligations to transfer shares later. As [Chambers’ 2026 German venture guide](https://practiceguides.chambers.com/practice-guides/venture-capital-2026/germany/trends-and-developments?ref=siliconsnark.com) explains, the investment and shareholders’ agreements usually get notarized together because their provisions are entangled with acts that require the form. The law does not merely bless the signature page. It can conscript the whole negotiated organism.

Now the notary must read the organism.

For an ordinary company formation, the viral bill is wildly unrepresentative. Germany’s Federal Chamber of Notaries lists [typical notarial costs ranging from roughly €105 for a one-person UG using a standard protocol to around €630 for a multi-person GmbH](https://www.notar.de/themen/notarkosten/beispiele?ref=siliconsnark.com), before some extras and tax. The famous €30,000 is not the admission price for opening a currywurst app. It is what can happen when a high-value financing, a huge bespoke document and statutory value-based fees meet in the same room.

That distinction matters. It also does not rescue the system. A financing round is precisely when a startup needs law to become a quiet piece of infrastructure. Instead, Germany sometimes turns it into immersive theater.

The country has digitized parts of the performance. Since August 2022, cash formations of GmbHs and UGs can be notarized by secure video; since August 2023, the online process has covered more capital measures and amendments. The [official notary portal](https://online.notar.de/en/?ref=siliconsnark.com) is real, works in English and removes the need to appear physically for eligible procedures. This is useful progress. It is also a very German digital reform: the founder may now watch the ritual remotely.

The deeper issue is not whether the reading happens in an oak-paneled office or through a webcam. It is whether standard venture transactions need this much mandatory ceremony at all. Germany has improved the waiting room while preserving the appointment.

## Founders Are Forming Companies Anyway

If bureaucracy alone determined entrepreneurial activity, the country should be an empty Handelsregister surrounded by tax advisers. Instead, startup formation is accelerating.

The German Startup Association and startupdetector counted 3,053 new startups in the first six months of 2026, up 52 percent from the second half of 2025\. June alone produced 600\. Every federal state grew. Berlin led in absolute formations with 429, followed by Hamburg with 212 and Munich with 208\. Measured per capita over the prior 12 months, Munich remained the leading large city at 24.6 formations per 100,000 residents; Berlin and Düsseldorf followed at 21.3.

The composition is more important than the headline. Software remained the largest category, with 844 formations. AI appeared in 34 percent of all new startups, up from 27 percent in 2025\. But industrial startups were the fastest-growing category: 136 were formed, a 125 percent increase that moved the sector from twelfth to sixth place.

That is what an economy rearranging itself looks like. Germany is not merely producing another crop of meal-delivery optimizers and mobile banks with gradient cards. It is creating companies in robotics, energy systems, defense, materials, manufacturing software and scientific hardware—the sort of products that require laboratories, supply chains and customers with loading bays.

The broad entrepreneurship data is less euphoric but just as revealing. The [KfW Entrepreneurship Monitor](https://www.kfw.de/PDF/Download-Center/Konzernthemen/Research/PDF-Dokumente-Gr%C3%BCndungsmonitor/KfW-Gr%C3%BCndungsmonitor-2026.pdf?ref=siliconsnark.com) counted 690,000 people starting businesses in 2025, up from 585,000, although 70 percent were part-time founders and the category extends far beyond venture-backed startups. Sixty-four percent identified bureaucracy as a constraint. Founders spent an average of 5.1 hours a week on legal and regulatory demands; full-time founders spent 8.7 hours, and those with employees 12.3\. For the average founder, administration consumed a quarter of the working week.

Germany is therefore not short of entrepreneurial intent. It is charging an administrative subscription fee against it.

## Berlin Built the Club. Munich Built the Bill of Materials.

There is no single German startup scene. There is a collection of regional systems with different accents, industries and tolerance for wearing company hoodies to dinner.

Berlin remains the country’s startup capital in raw numbers and international mindshare. Its venture-backed tech ecosystem was worth an estimated €169 billion at the end of 2025, or 43 percent of Germany’s total, according to [Dealroom’s Berlin ecosystem review](https://dealroom.co/reports/the-berlin-tech-ecosystem-2025-in-review/?ref=siliconsnark.com). It has the deepest founder network, the most cosmopolitan hiring pool and the largest supply of investors who will take a meeting about software that does not yet technically exist.

Munich is the counter-model. It combines the Technical University of Munich, Ludwig Maximilian University, the Max Planck network, BMW, Siemens, Allianz, Munich Re, aerospace, chips and an industrial base that can both manufacture a robot and insure the factory it bumps into. [Bavaria had 4,406 active startups and scaleups](https://startupverband.de/presse/pressemitteilungen/bayern-monitor-2026-muenchen-fuehrt-bei-startup-dichte-und-finanzierung-17-03-2026/?ref=siliconsnark.com) in early 2026, with 2,445 in the Munich region. Munich startups raised €2.7 billion in 2025, edging Berlin’s €2.4 billion; more than €1 billion of Munich’s total went into defense technology.

The contrast is sometimes overstated into caricature. Berlin is not only consumer apps, and Munich is not a board meeting wearing a loden jacket. Berlin has serious climate, biotech and defense companies. Munich created Celonis, one of Europe’s most important software firms. But the ecosystems do have different center fields. Berlin excels at international talent, fintech, consumer products and company creation. Munich excels when the pitch deck contains a cutaway diagram.

Hamburg adds logistics, commerce, media and a rapidly growing formation rate. Cologne and Düsseldorf bring commerce, media, enterprise customers and the dense corporate landscape of North Rhine-Westphalia. Karlsruhe has cybersecurity and deep university ties. Stuttgart has automotive and industrial engineering. Dresden carries semiconductor expertise. Freiburg somehow produced a frontier AI company between the vineyards.

This polycentric structure is a strength because German industry is polycentric. It is also a networking tax. A founder in San Francisco can collide with capital, customers, researchers and alumni over lunch. A German founder may need a BahnCard.

## Berlin’s First Product Was Berlin

Germany’s modern startup story began with a city that was cheap, strange and only partially finished.

After reunification, Berlin offered low rents, empty buildings, universities, artists and a government eager to remake the capital. International young people could afford to arrive before they had a job. That population produced nightlife, design, technical talent and the social density from which companies emerge. Berlin’s first successful startup product was Berlin: a global brand for people willing to trade salary for possibility and central heating of uncertain conviction.

The 2000s brought the Samwer brothers and Rocket Internet, a company-building machine often summarized as “American idea, German operations manual.” Rocket took models proven elsewhere, launched local versions at speed and expanded them across markets Silicon Valley had treated as a map labeled Here Be Tax Codes. The strategy attracted derision for its derivative products and admiration for its execution. Both were deserved.

Out of that era came Zalando, Delivery Hero and HelloFresh, plus an enormous alumni network of operators who learned growth, logistics, international expansion and the spiritual cost of a weekly dashboard. Rocket did not make Berlin synonymous with original research. It did something more foundational: it trained people to build companies at scale.

The next generation broadened the model. N26 pursued banking, GetYourGuide travel, Contentful infrastructure, Enpal solar, and a long tail of enterprise, climate and health companies. Venture capital followed. So did the predictable effects of success: higher rents, higher salaries, a professional services layer and founders who no longer needed to explain that Berlin was in Germany.

The city now has a mature ecosystem’s best and worst features. Repeat founders can raise on reputation. Employees have seen hypergrowth up close. Investors understand the local market. And every possible variation of “AI for the Mittelstand” already has a pre-seed deck.

## Germany Discovered B2B Software Has Fewer Returns

Germany’s great software successes tend to make sense once you remember what the country already knows: complex organizations, industrial processes, regulation and the monetizable agony of enterprise workflows.

SAP, founded in 1972 by five former IBM engineers, established the national archetype. German software did not have to be charming. It had to know where the inventory was.

[Celonis updated that instinct for the cloud era](https://www.celonis.com/company/about-us?ref=siliconsnark.com). Founded in 2011 by three students from the Technical University of Munich, it built process-mining software that reconstructs how work actually moves through a company rather than how management’s flowchart claims it moves. Celonis became a unicorn in 2018 and a decacorn after a $1 billion round in 2021\. It now says it has roughly 3,000 employees and 1,400 customers. The product is almost cosmically German: software that watches an organization and points out that the process is not being followed.

Personio turned human-resources administration for small and midsize companies into a multibillion-dollar business. DeepL built machine translation of such quality that it became one of Europe’s rare consumer-facing AI names before every company added “AI” to the lobby. Munich’s commercetools helped define headless commerce. Berlin’s Trade Republic brought mobile investing to millions and, in December 2025, [completed a secondary transaction at a €12.5 billion valuation](https://assets.traderepublic.com/assets/files/251217%5FSecondary%5FPressRelease%5FIT%5FEN3.pdf?ref=siliconsnark.com) after reporting more than 10 million customers and three profitable years.

These are not consolation prizes for failing to create Google. They reflect a different demand environment. Three-quarters of German startups sell primarily to businesses. Europe’s economy contains an endless supply of complicated firms with legacy systems, regional rules, old equipment and budgets for making the complication slightly less complicated. As we found when [German database company Exasol turned its platform into a sovereign AI panic room](https://www.siliconsnark.com/exasol-turned-its-database-into-a-sovereign-ai-panic-room/), geopolitical anxiety has made local control another enterprise feature.

The weakness is that German B2B excellence can become an excuse for timid ambition. A startup can spend years customizing pilots for prestigious domestic customers and wake up as a profitable consultancy with a product tucked inside. The country knows how to make a customer happy. A global platform sometimes requires making the customer adapt.

## The 2021 Money Fountain Had a German Setting

The cheap-money era inflated German venture capital just as it inflated everything else with a recurring-revenue chart.

German startups raised about €18.8 billion in 2021, according to KfW’s venture dashboard. Valuations jumped, late-stage rounds multiplied and founders could plausibly believe Europe’s capital gap was closing. Then interest rates returned from sabbatical. The market fell sharply, weak companies met their burn rates, and “path to profitability” re-entered the language as if it had merely stepped out for coffee.

By 2025, [annual investment had stabilized around €7.2 billion](https://www.kfw.de/About-KfW/Newsroom/Latest-News/Pressemitteilungen-Details%5F876672.html?ref=siliconsnark.com), essentially flat with 2024 and 2023\. The number of rounds declined from 1,540 in 2024 to 1,444\. This was not a collapse. It was a smaller and more selective system operating far below the carnival peak.

Artificial intelligence and security technology began to absorb more of the available money. Defense and resilience themes accounted for 17 percent of 2025 venture volume. That would have been politically improbable before Russia’s full-scale invasion of Ukraine and commercially improbable before Europe noticed that cloud infrastructure, chips, drones and software supply chains can all become strategic objects between lunch and the next European Council meeting.

The shift resembles the wider European sovereignty boom we have covered in [Microsoft’s attempt to put Mistral behind Europe’s firewall](https://www.siliconsnark.com/microsoft-is-spending-billions-to-put-mistral-ai-behind-europes-firewall/) and [the EU’s €10 billion AI gigafactory tender](https://www.siliconsnark.com/europe-opened-a-eu10-billion-ai-gigafactory-tender-good-luck-with-electricity/). Capital is no longer chasing only growth. It is chasing strategic reassurance, preferably with a procurement budget attached.

## The New German Startup Is Wearing Safety Glasses

For a decade, European founders were told to build software because software scales without factories, inventory or a conversation with a municipal planning office. Germany listened just long enough to create a strong software scene. Then history began ordering hardware.

Energy insecurity, defense spending, AI compute demand, labor shortages, industrial decarbonization and supply-chain risk all favor companies that work on atoms. Those are expensive businesses. They are also businesses for which Germany has useful unfair advantages: applied research institutes, engineering talent, chemical and automotive supply chains, machine-tool expertise, specialist manufacturers and customers who understand why a bearing matters.

The H1 2026 formation data captures the change. Industrial startups were still only the sixth-largest category, but their 125 percent growth was faster than any other sector. Venture data tells the same story at much larger denominations. EY counted €1.6 billion for hardware startups in the first half of 2026, ahead of software and analytics at €1.2 billion. Defense technology received €579 million.

Software people like to say hardware is hard. Germany hears that as a product requirement.

The opportunity is bigger than robots with photogenic torsos. It includes power electronics, grid equipment, sensors, batteries, novel materials, industrial heat, autonomous systems, space hardware, synthetic biology and the software layers controlling physical assets. Berlin startup [NextGO Epi’s attempt to commercialize gallium-oxide power electronics](https://www.siliconsnark.com/nextgo-epi-raised-eu2-million-to-make-power-electronics-wear-gallium-oxide/) is a good miniature of the model: years of public research, a technical advantage invisible to ordinary consumers and a market created by the limits of existing materials.

These companies also expose every weakness in European finance. A SaaS startup can revise its pricing page. A materials startup cannot negotiate with a half-built pilot plant. Deep tech needs larger rounds, longer timelines, specialized diligence and investors who remain calm when revenue arrives after the equipment.

## Helsing, NEURA and the Billion-Dollar Workshop

The round announcements of 2026 looked less like German venture capital and more like an industrial policy fever dream.

Munich-based Helsing raised [$1.8 billion in a Series E at an $18 billion valuation](https://helsing.ai/de/newsroom/helsing-sammelt-1-8-mrd-us-dollar-in-der-series-e-finanzierungsrunde-ein?ref=siliconsnark.com) in July. The company builds AI-enabled defense systems and says it remains majority European-owned. Whatever one thinks about defense technology—and Europeans spent years preferring not to think about it—Helsing demonstrates that a European company can attract global-scale capital when product demand becomes strategically unavoidable.

NEURA Robotics, based in Metzingen, announced [a Series C of up to $1.4 billion](https://neura-robotics.com/record-series-c/?ref=siliconsnark.com) in June. The wording matters: “up to” is not the same as cash already wired, and the company’s reported billion-dollar order book is its own claim. Even with the footnotes, the round reflects intense investor appetite for so-called physical AI—robots that bring machine learning out of the chat window and into places where they can drop a torque wrench.

We have already examined why [factories, startups and billionaires suddenly want humanoid robots](https://www.siliconsnark.com/humanoid-robots-explained-why-factories-startups-and-tech-billionaires-suddenly-want-a-mechanical-workforce/). Germany has one of the strongest answers: because its industrial employers face demographic pressure, high labor costs and production environments designed around human bodies. If humanoids ever become commercially useful, Germany is not merely a market. It is the exam.

Proxima Fusion, spun out of the Max Planck Institute for Plasma Physics, [raised €411 million at a €2.4 billion valuation](https://www.proximafusion.com/press-news/proxima-fusion-raises-eu411-million-to-build-europes-commercial-fusion-champion?ref=siliconsnark.com) in July, bringing its announced public and private funding to €650 million. Its stellarator approach may take years to prove. Fusion is what happens when deep tech looks at normal hardware risk and asks for something more ambitious.

Black Forest Labs offers the software-side counterexample. Founded in Freiburg in 2024 by researchers behind Stable Diffusion, the company [raised $300 million at a $3.25 billion post-money valuation](https://www.globenewswire.com/news-release/2025/12/01/3197140/0/en/UPDATE-Black-Forest-Labs-Announces-Series-B-Investment-to-Accelerate-Frontier-Visual-Intelligence.html?ref=siliconsnark.com) in late 2025 for frontier visual AI. It showed that German research talent can still create a fast-moving model company without a turbine, a foundry or a procurement officer.

Four giant financings do not constitute a broad capital market. They do, however, destroy the lazy version of the German story. The country is not merely watching American firms build the future while perfecting the fax cover sheet. It is producing companies at the center of AI, defense, robotics and energy. Its problem is making that outcome repeatable below the headline tier.

## Research Is Germany’s Hidden Co-Founder

Germany’s startup advantage begins years before a pitch deck, usually in a building where the coffee machine has a grant number.

The country [spent €129.7 billion on research and development](https://publica.fraunhofer.de/entities/publication/1165b1e6-d54a-443c-a851-0cf9430aa85c?ref=siliconsnark.com) in 2023, about 3.1 percent of GDP. Its institutional landscape is unusually dense: Max Planck institutes pursue fundamental science; Fraunhofer institutes focus on applied research; Helmholtz centers operate large-scale facilities; the Leibniz Association spans the gap; technical universities produce engineers and spinouts.

The [2025 European Spin-off Report](https://www.mpg.de/25797163/european-spin-off-report-2025?ref=siliconsnark.com) ranked the Max Planck Society second among European research institutions by spinout creation, Helmholtz fifth and Fraunhofer tenth. Max Planck spinouts had produced four unicorns and more than $67 billion in aggregate value. [Fraunhofer reported 29 new spinouts](https://www.fraunhoferventure.de/de/news/presse-2026/bilanz-2025.html?ref=siliconsnark.com) in 2025, up from 21 the year before.

This is the machinery behind companies like Proxima and a wider crop of photonics, quantum, biotech, climate and advanced-material ventures. It also explains why the German ecosystem looks increasingly suited to a period in which technology policy has rediscovered power grids, factories and defense procurement.

But research transfer remains the national hinge. Scientific excellence does not automatically produce a clean intellectual-property license, an entrepreneurial principal investigator or an investor willing to finance four years of engineering. Universities and institutes differ in their spinout terms, speed and appetite for risk. A brilliant researcher can find that the hardest system to model is the technology-transfer office.

The capital market has begun adapting. Specialist funds increasingly target companies too scientific for a conventional software screen, a category we discussed when [Floating Point started raising $120 million for startups too complicated for a sound bite](https://www.siliconsnark.com/floating-point-is-raising-120-million-to-fund-startups-too-complicated-for-a-sound-bite/). That is the correct direction. Europe does not need its labs to imitate consumer internet startups. It needs an investment system capable of understanding what the labs already contain.

## The Mittelstand Is an Unfair Advantage With a Six-Month Pilot

Germany’s famous Mittelstand gives startups access to something most ecosystems spend fortunes trying to manufacture: demanding customers with global operations, deep domain expertise and real industrial problems.

A robotics company can test with an automaker. A carbon-accounting platform can sell into chemicals. A logistics startup can find businesses whose supply chains cross six borders before breakfast. Hidden champions—often family-owned companies leading obscure global niches—offer both design partners and acquisition targets.

In theory, this is a flywheel. Startups bring speed and new technology. Established companies bring data, distribution, equipment and money. In practice, the flywheel has been placed behind vendor onboarding.

A [2025 Startup Association survey](https://startupverband.de/presse/pressemitteilungen/report-kooperationen-zwischen-startups-und-etablierten-unternehmen-%25E2%2580%2593-grosse-chancen%2C-aber-es-hakt-an-vielen-stellen-21-08-2025/?ref=siliconsnark.com) found that 90 percent of corporates considered startups important to their future, but only 11 percent of startups perceived a high willingness to cooperate. Fifty-nine percent cited slow processes and 49 percent cited corporate risk aversion. The share of startups working with established companies fell from 72 percent in 2020 to 56 percent in 2025.

This is Germany’s industrial paradox in miniature. The customers most capable of validating a new technology are often the least institutionally capable of buying it quickly. A corporate innovation office will invite a founder to three workshops, photograph the Post-its and then discover that procurement requires three years of audited accounts from the company incorporated nine months ago.

Startups need the Mittelstand. The Mittelstand increasingly needs startups. Both parties agree, and a steering committee has been formed to explore the synergy.

## The Funding Ladder Is Missing Its Top Rungs

German venture capital in 2026 looks strong from the helicopter and narrow from the ground.

[KfW’s Q2 2026 dashboard](https://www.kfw.de/PDF/Download-Center/Konzernthemen/Research/PDF-Dokumente-Dashboard/KfW-VC-Dashboard-Q2-2026.pdf?ref=siliconsnark.com) recorded €5.1 billion invested in the first half, including €3.4 billion in the second quarter. EY, using a different methodology, put the total at €5.3 billion, up 14 percent year over year. Seven rounds above €100 million transformed the quarter. Scaleups absorbed €2.5 billion, or 72 percent of Q2 capital. Seed companies received €187 million.

The number of deals worth at least €1 million fell to 184 for the half-year. Q2’s 86 deals were the weakest second-quarter count in the series. Median round size rose from €2.2 million in 2023 to €5 million in the first half of 2026.

Money did not flood the ecosystem. It selected a few boats and lifted them onto the dock.

AI alone took €2.2 billion in Q2, 63 percent of all invested capital, across 35 rounds. One enormous robotics transaction can make a country look magnificently funded while hundreds of ordinary seed and Series A companies experience a colder market. This is not fraud by statistics; it is why totals need company.

Investor origin reveals both Germany’s strength and dependence. German investors supplied 33 percent of Q2 capital; US investors supplied 29 percent. Foreign capital is welcome and necessary. The vulnerability appears later, when the largest follow-on rounds, likely acquirers and liquid public markets sit abroad. Germany can finance invention locally and still export ownership as the price of scale.

The government has responded with an expanding alphabet of public-private capital. KfW Capital had committed €2.9 billion to 153 venture funds by the end of 2025\. The €1 billion Growth Fund Germany was designed with nearly 70 percent private capital. The WIN initiative targets €12 billion in commitments by 2030\. The broader Deutschlandfonds uses public money and guarantees in an effort to mobilize far more private investment.

These programs are directionally sensible. European pension funds and insurers invest too little in venture and growth equity, and public capital can absorb some risk. But a funding ecosystem cannot consist of American mega-funds at the top, German state acronyms at the bottom and a laminated diagram explaining how they connect.

## The Exit Sign Mostly Points West

Venture capital is not merely money entering startups. It is a system for money eventually leaving them. On that side of the ledger, Germany still exports too much of the climax.

A [KfW analysis of 986 exits by venture-backed German startups since 2005](https://www.kfw.de/PDF/Download-Center/Konzernthemen/Research/PDF-Dokumente-Fokus-Volkswirtschaft/Fokus-2025/Fokus-Nr.-521-November-2025-Start-ups.pdf?ref=siliconsnark.com) found that roughly 92 percent were acquisitions, 5 percent buyouts and only 4 percent initial public offerings, with rounding in the categories. Fifty-seven percent of acquired startups went to foreign buyers. American companies alone accounted for 24 percent of acquisitions; German buyers accounted for 43 percent.

The IPO numbers are small but strategically loud. Forty-four percent of German startup IPOs took place abroad. Nasdaq hosted 32 percent, while German exchanges hosted 54 percent. IPOs represented only about 4 percent of exits, but the companies going public accounted for 51 percent of employment across exited startups. The rare company capable of becoming a large independent employer is also the company most tempted by a deeper foreign market.

An acquisition is not a failure. Founders and employees receive liquidity. Technologies find distribution. Capital returns to funds and can finance another generation. Foreign buyers are often the best owner. The problem is cumulative: when a country repeatedly sells its strongest companies before they become global platforms, it loses headquarters, senior jobs, strategic control, acquisition capacity and the alumni who would otherwise seed the next ecosystem.

The United States is not powerful merely because it starts companies. It has a compounding machine in which public markets, giant technology firms, venture funds and employee wealth continuously recycle the winnings. Germany has constructed much of the first half and outsourced too much of the second.

Trade Republic’s €1.2 billion secondary sale in 2025 is an interesting bridge. It gave shareholders liquidity at a €12.5 billion valuation without forcing the company into a sale or premature listing. Secondary markets can keep companies independent longer. They cannot substitute forever for a European public market willing to price growth companies, attract research coverage and accept that quarterly earnings might occasionally contain investment.

## Employee Equity Improved, Which Is German for “Bring Counsel”

Startups compete for talent partly with ownership. Germany spent years making that sentence sound like a tax-policy threat.

The historical problem was “dry income”: employees could owe tax on shares before receiving cash to pay it. Section 19a of the Income Tax Act and subsequent reforms improved the position by allowing taxation to be deferred until a later event, such as a sale, the end of employment or a statutory deadline. The rules now cover more companies and offer better protections when employment ends.

That was meaningful progress. It did not turn German equity compensation into a one-page American option grant. GmbH shares remain formal objects. True ownership can involve notarization, valuation, payroll questions and legal administration. Many startups use virtual share programs, which mimic the economics of equity without making the employee an actual shareholder. Virtual shares are easier to administer and easier to explain right up to the sentence where one admits they are contractual promises from the company rather than the thing itself.

This matters because salaries in Munich, Berlin or Hamburg compete with London, Zurich, New York, remote American employers and established German corporations. A startup cannot always win on cash. It must make the upside legible, portable and credible to an engineer deciding whether “employee number 43” is a career move or an unusually elaborate way to finance the founder’s risk.

The German government recognizes the issue; every startup strategy eventually discovers stock options. The remaining job is not another declaration that participation matters. It is making real equity cheap enough to issue, simple enough to understand and predictably taxed when value becomes liquid. Founders should not need a tax-law companion podcast.

## International Founders Get a Visa and a Scavenger Hunt

Germany’s demographic arithmetic makes foreign talent essential, not decorative.

The 2025 Migrant Founders Monitor found that [14 percent of German startup founders were born abroad](https://startupverband.de/presse/pressemitteilungen/migrant-founders-monitor-2025-14-prozent-der-gruender%2Ainnen-im-ausland-geboren%2C-23-prozent-bei-unicorns--doch-deutschland-muss-attraktiver-werden-12-05-2025/?ref=siliconsnark.com). Among founders of German unicorns, the share rose to 23 percent. That is significant and still far behind the United States, where the equivalent figure was 44 percent. Only 46 percent of foreign-born founders rated access to local networks positively, compared with 57 percent of founders overall.

Berlin’s international culture helps. English is the operating language of many teams, landlords have heard of a foreign passport, and a founder can build a social network without first mastering the dative. Munich is more uneven: only 46 percent of its startups used English as their working language in a 2026 Bavaria study, compared with 67 percent in Berlin.

Germany [offers a residence route for self-employment](https://www.make-it-in-germany.com/en/working-in-germany/setting-up-business/visa?ref=siliconsnark.com) when the business serves an economic interest or regional need, is expected to have positive economic effects and has secured financing. A successful founder can potentially obtain permanent residence after three years. On paper, this is competitive. In practice, experiences vary by city, caseworker, appointment supply and the applicant’s ability to produce the exact document that was not listed on the website.

The country has modernized skilled migration rules and introduced more digital processes. Yet founders still report friction around visas, local registration, banking, tax numbers, health insurance and family relocation. None of these is individually fatal. Together they form the German administrative boss battle: many small opponents, each holding a stamp.

The inclusion problem is not limited to nationality. Women accounted for only [19 percent of German startup founders in the latest Startup Association analysis](https://startupverband.de/presse/pressemitteilungen/darum-gibt-es-in-deutschland-zu-wenig-gruenderinnen-02-04-2025/?ref=siliconsnark.com), a slight decline. The gap reflects who receives capital, who accumulates technical and commercial networks, who absorbs unpaid care work and which founder archetype investors recognize quickly enough to fund. An ecosystem cannot complain about a shortage of founders while repeatedly selecting from the same narrow room.

## The State Wants Innovation, Provided It Has Three References

Germany is trying to become a major market for defense, climate, health and digital infrastructure. In each case, government is not just a regulator. It is a buyer. This creates a simple test of startup policy: will the state purchase from startups?

The answer has historically been a carefully formatted “perhaps.”

The [2026 GovTech Startup Monitor](https://startupverband.de/presse/pressemitteilungen/govtech-startup-monitor-2026-29-06-2026/?ref=siliconsnark.com), based on 225 founders and 33 public administrators, found that only 11 percent of startups serving government considered it a more attractive customer than the private sector. Seventy-nine percent struggled to find the correct contact. Eighty-five percent said decisions took too long. Seventy-two percent found documentation too complex. Seventy-seven percent believed procurement favored established providers.

This is not simply startups moaning that government refuses to buy a chatbot after a cappuccino. Public procurement must prevent corruption, protect taxpayer money, maintain continuity and comply with European rules. The state cannot beta-test a benefits system as casually as a consumer app tests a new button.

But risk management can become risk displacement. Choosing an incumbent with a mediocre product feels administratively safe because the failure is familiar. Choosing a young company with a better product feels risky because someone must own the decision. The outcome is a public sector that demands innovation in strategy documents and rewards longevity in tenders.

Germany has started changing the rules. [Procurement reforms effective in July 2026](https://www.bundeswirtschaftsministerium.de/Redaktion/DE/Pressemitteilungen/2026/06/20260610-bmwe-vereinfacht-oeffentliche-beschaffung.html?ref=siliconsnark.com) simplify procedures, and federal rules now allow direct awards to young startups up to €100,000 in certain circumstances. Negotiated awards can also be directed to a single startup within defined age limits. These are practical changes, not slogans. Their value will depend on whether purchasing officers are encouraged to use the flexibility or merely receive another circular describing it.

For defense and deep tech, this matters more than another conference. Early government contracts validate products, generate revenue and help European companies survive until larger programs arrive. A sovereign technology strategy without sovereign first customers is a motivational poster.

## Europe Is 27 Markets in a Trench Coat

Many problems blamed on Germany are actually European problems with a German return address.

A startup formed in Delaware begins inside a single country of roughly 340 million people, one dominant language, deep capital markets and broadly compatible commercial rules. A startup formed in Germany begins inside the European Union’s single market, a monumental achievement containing 450 million people, 24 official languages, national tax systems, varying employment rules, fragmented public procurement and enough local compliance to keep every country’s advisory class comfortably indoors.

The single market is real for goods and weaker for services, finance and company administration. Expansion into France, Italy or Poland can mean new contracts, regulators, payroll systems and consumer rules. Fintech feels this acutely, which is why [cross-border money apps tend to accumulate compliance footnotes](https://www.siliconsnark.com/bloxley-turns-cross-border-money-into-one-app-and-several-compliance-footnotes/). The European Central Bank’s attempt to build [a digital euro that taps like Visa and bills like sovereignty](https://www.siliconsnark.com/the-ecb-wants-a-digital-euro-that-taps-like-visa-and-bills-like-sovereignty/) is partly an effort to create shared infrastructure where fragmented commercial systems remain.

Fragmentation also explains Europe’s obsession with sovereignty. The continent has the researchers and customers to support important AI companies, but less compute, less risk capital and no unified procurement machine. Even Mistral’s [€830 million bank financing for American chips](https://www.siliconsnark.com/mistral-borrowed-830-million-from-seven-banks-to-buy-american-chips-and-call-it-european-independence/) neatly captures the dependency. European independence currently accepts Nvidia.

Germany cannot solve the continental market alone. It can decide whether to compound the fragmentation. National gold-plating, language requirements, slow registrations and local formalities turn the theoretical single market into 27 domestic markets wearing a blue trench coat with stars.

## EU Inc. Is Delaware With a Legislative Calendar

In March 2026, the European Commission finally proposed the thing founders had been demanding for years: an optional, digital-first company form usable across the Union.

[The EU Inc. proposal](https://cyprus.representation.ec.europa.eu/news/commission-presents-proposal-eu-inc-unlocking-full-potential-single-market-europes-entrepreneurs-2026-03-18%5Fen?ref=siliconsnark.com) promises incorporation within 48 hours for no more than €100, no minimum capital, a single digital interface, flexible share classes, online corporate actions and a Union-wide employee stock-option framework that taxes gains at sale. It would reduce mandatory intermediaries for share transfers and allow founders to use one recognizable structure across borders.

It is, conceptually, Delaware with umlauts removed.

It is also a proposal. Parliament and the Council must agree. National systems must implement it. The Commission wants a political agreement by the end of 2026 and a functioning “one Europe, one market” environment by 2028\. Europe is capable of creating a company in 48 hours once it completes a two-year legislative process explaining how.

The German Startup Association welcomed the plan while calling it a minimum viable product. Its [April position paper](https://startupverband.de/fileadmin/startupverband/mediaarchiv/Politik/260414%5FPosition%5FPaper%5FGerman%5FStartup%5FAssociation%5FEU%5FInc.pdf?ref=siliconsnark.com) warned that EU Inc. remains partly dependent on national law, lacks a complete specialist dispute system and could be sabotaged by the same domestic bottlenecks it is supposed to avoid. A company may exist after 48 hours and still wait months for a German tax number. National authorities could preserve notarial controls through implementation. Standard early-stage financing documents remain absent.

Those are not pedantic objections. A corporate form is an operating system. If national plug-ins can reintroduce the old friction, the new icon on the desktop does little.

EU Inc. may become one of the most important European startup reforms in a generation. It would not eliminate local tax, labor or sector law. It could give investors, employees and founders a shared legal chassis, making cross-border companies easier to finance and administer. Germany should be its most enthusiastic implementer. A country whose founders are escaping the notarial audiobook has every reason to sponsor streaming.

## Berlin Has 150 Measures. Of Course It Does.

Germany’s federal cabinet adopted a new Startup and Scaleup Strategy on July 22, 2026\. It contains roughly [150 measures across eight fields](https://www.bundeswirtschaftsministerium.de/Redaktion/DE/Pressemitteilungen/2026/07/20260722-start-up-scale-up-strategie-beschluss.html?ref=siliconsnark.com), informed by more than 300 submissions and ten workshops.

Naturally, the strategy for reducing complexity has 150 measures.

That joke is cheap because the substance is serious. The strategy addresses financing, company growth, research transfer, bureaucracy, talent, AI, defense technology, public procurement and European market integration. It builds on the WIN initiative, Deutschlandfonds, improved equity rules and procurement reform. The government promises annual implementation reports, which is exactly what a broad cross-ministry program needs if it is to become something more than a handsome PDF.

The policy diagnosis is now broadly correct. Germany needs faster formation, more growth capital, better spinouts, simpler options, greater public purchasing and deeper European markets. Founder surveys, investors and ministries have converged on the same list. There is little analytical mystery left.

Implementation is the entire product.

Germany’s political system disperses authority across federal ministries, states, municipalities, regulators, courts and professional bodies. This protects against reckless central power and makes coordinated reform exhausting. Digital company formation touches notaries, registries, tax offices, identity systems and state-level administration. Procurement reform can change federal law while local buyers preserve cautious habits. Immigration law can liberalize while appointments remain scarce.

Every institution can agree on the destination and retain a procedural veto over the bus.

## Germany Does Not Need to Become California

The most boring advice in European technology is that Europe should copy Silicon Valley. It cannot, and where it can, it often should not.

California’s technology ecosystem grew from American defense spending, elite universities, immigration, a unified continental market, permissive bankruptcy, stock options, public markets, giant platforms and several decades of self-reinforcing wealth. Germany cannot reproduce that history by adding beanbags to a Munich insurance office.

Nor should it discard its own strengths. German vocational training, worker protections, research institutes, engineering culture, manufacturing base and long-term industrial relationships are not bugs to be refactored. They are why the country has a plausible position in robotics, energy, mobility, quantum, materials and industrial AI. The goal is not a Europe where everyone moves fast and breaks the high-voltage switchgear.

Germany’s opportunity is to build a different compounding machine: public research feeding spinouts; patient capital financing industrial scale; Mittelstand customers providing first deployments; European procurement creating a home market; and founders retaining enough ownership and independence to create the next generation.

That machine will produce fewer overnight consumer sensations and more companies whose first meaningful milestone is a certified pilot line. It may look slow to a software investor. The output can still be globally decisive.

The rise of open-weight AI makes the strategy more plausible. As our [deep dive into open-weight models](https://www.siliconsnark.com/deep-dive-open-weight-ai-from-checkpoints-to-china/) showed, companies can increasingly build proprietary systems on accessible model foundations rather than training a frontier model from scratch. Germany can combine those models with industrial data, robotics, scientific instruments and domain expertise. It does not need to win the general chatbot leaderboard to build the intelligence layer of a factory.

## What Germany Actually Needs to Fix

First, Germany should reserve mandatory notarization for the acts that genuinely need an impartial public witness and remove the gravitational field that drags entire venture agreements into the ceremony. Standard financings, employee equity and share transfers should be digital, fast and priced for administration rather than dramatic value. A founder may still hire excellent lawyers. The state need not require a live reading of their output.

Second, company formation must mean operational formation. A 48-hour registration is not success if the tax number, bank checks, payroll registration and sector permits arrive on separate geological schedules. Founders need one interface, once-only data submission, reliable English support and published service deadlines. The relevant metric is days until a company can employ, invoice and receive investment—not days until a registry generates a PDF.

Third, Germany and Europe need more domestic growth capital and better exits. Pension funds and insurers should be able and encouraged to allocate prudent slices to venture and growth funds. Secondary markets need depth. European exchanges need a more attractive path for high-growth listings. Public programs should catalyze private underwriting, not become permanent substitutes for it.

Fourth, the state and large corporations must become first customers. Procurement officers need permission, training and incentives to buy from young companies, with staged contracts that manage risk through milestones rather than excluding risk through eligibility. Corporates should measure pilots by how many enter production, not how many enter PowerPoint.

Fifth, research organizations should standardize fair spinout terms, publish timelines and reward technology transfer as a core output. Germany already pays for extraordinary science. Losing years in IP negotiation is a bizarre way to protect the investment. Specialist investors can handle scientific risk; they cannot finance indefinite institutional ambiguity.

Sixth, talent policy must be designed around the whole arrival. Visa, spouse employment, city registration, housing, tax and English-language administration are one founder experience even if government distributes them across seven offices. Employee ownership should be real, comprehensible and taxed on liquidity.

Finally, Germany has to become more comfortable with entrepreneurial failure. A country built around engineering reliability naturally dislikes businesses whose base rate includes death. But startups are a portfolio method for discovering what works. If every failed company stains a founder while every cautious incumbent receives another contract, the ecosystem will optimize for defensible smallness.

None of these changes requires Germany to abandon the rule of law, worker protection or fiscal seriousness. They require the rules to distinguish between preventing abuse and preventing motion.

## The Verdict: Germany Is Making Success Wait Outside

Patrick Collison’s post works because it condenses the German startup problem into one room. The founder, investors and lawyers have agreed on a transaction. The company wants capital. Everyone is present. Then the system insists that progress become audible.

But outside that room, the German startup scene is not waiting quietly. It is forming companies at a record pace. Munich has become one of Europe’s strongest deep-tech centers. Berlin remains a major international startup hub. Industrial formations are surging. Researchers are leaving elite institutes to build fusion, materials and AI companies. Global investors are writing billion-dollar checks for German defense systems and robots.

The ecosystem’s weakness is not creation. It is conversion.

Germany converts too little research into companies, too few pilots into orders, too few startups into scaleups, too few scaleups into independent public companies and too little global success into domestic compounding capital. At each transition, the country adds friction that looks reasonable in isolation and absurd in aggregate.

That is why the second company in Collison’s anecdote matters more than the €30,000 bill. The notary earned a statutory fee. Germany lost a future set of shareholders, employees, tax relationships and strategic choices before company number two had built anything at all.

The good news is that this is fixable. Germany does not lack founders, science, customers or technical seriousness. It does not need to summon an entrepreneurial culture from a government workshop. The culture is already producing 3,053 startups in six months while spending a quarter of its week on administration.

Germany’s startup scene is not dead, timid or irrelevant. It is one of Europe’s most important technology engines, attached to a legal and financial transmission designed when horsepower was suspicious.

The founders are ready. The robots are ready. The fusion coils are warming up.

The notary has asked everyone to turn back to page 47.