Fabian Westerheide is a founding partner of the AI-focused venture capital investor AI.FUND and has been privately investing in AI companies through Asgard Capital since 2014. Westerheide provides strategic advice to public and private institutions in the area of AI and invites you to the Rise of AI conference in Berlin every year.
Anyone who talks about AI financing these days quickly ends up with numbers that feel like science fiction: 200 million, 650 million, one billion. Rounds that seem to arise out of the blue and give the impression that the logic of the venture market has fundamentally changed.
This is exactly what is dangerous for founders. Such financing not only generates headlines, but also false benchmarks. Suddenly a solid seed round seems small. A cleanly structured B2B sales department slowly. A company with real customers is almost boring. Often exactly the opposite is true.
Read too
I have been observing and investing in AI since 2014. I know mega rounds. And above all, I know the pattern behind it: these windows open and they close again. Usually after two to three years. Then the mood changes. Then “Growth at any cost” becomes “Show me the numbers” again.
That’s why founders shouldn’t romanticize the current phase. You should understand them. Because the AI mega rounds are not just the new normal. They are a window of FOMO in a market that operates heavily on winner-takes-most logic. This combination can create real champions. But it also produces a lot of expensive illusions.
Super rounds are often less business logic than reputation logic
There are financing rounds that are actually reputation investments. Not just for founders, but also for VCs. Anyone who is a fund in the cap table of one of the big AI names is not just buying a share in a company. He’s buying a ticket to the powerhouse of the next tech cycle: access to deal flow, access to co-investors, access to talent, access to partnerships.
That can be rational. But it is not proof of product-market fit. First of all, it is evidence of network, narrative and positioning.
Read too
The error starts where founders misread these mechanics. High round equals high quality. Sometimes that’s true. Often it’s not true. Sometimes it’s simply a signaling business. A portfolio statement. A market entry sign: We’re there.
This is important for founders because they must not confuse these signals with their own strategic benchmark. Just because another team raises capital based on pedigree, network and market imagination doesn’t mean the same logic applies to their own company.
For whom such rounds are really rational
However, there is a small group for which such super rounds are actually understandable: exceptional teams that have already played in the top league. Teams from the environment of the large AI labs. People with real research depth, implementation skills and access to markets. Founders who not only understand models, but also distribution, platform logic and the speed with which markets can be occupied.
If there’s someone sitting there who has a proven ability to build a platform game, then the logic isn’t: How much revenue will this company make next quarter? But rather: Can this team take a central position in a market that is currently being reorganized?
Such bets are extremely risky, but not irrational. However, they are rare. And they are not the benchmark for 99 percent of the market. This is exactly where the distortion arises: everyone looks at the top of the pyramid and thinks it is the pyramid.
For everyone else: Sales and traction beat hype
For most founders, a simpler and much more brutal rule applies: You don’t get a shortcut for free. The best defense against the cycle is not a pitch deck. It’s a business.
After ten years in the AI market, I can say this very clearly: Real companies live longer. In the end they survive. I’ve lost money with hype companies that didn’t provide numbers. Nice story, strong slides, big vision, but no reliable unit economics, no repurchase rates, no real pull from the market.
And I made money with startups that eventually became companies: with customers, contracts, recurring revenue, clear budgets on the customer side and a product that was not “nice to have” but “must have”.
Hype can carry you all the way to seed. Reputation can open doors for you. But by the time the markets turn, no one will ask how good your narrative sounds. Then cohorts, churn, sales cycle, gross margin, net revenue retention count. Then the demo no longer counts. Then the question that counts is: Will AI become a business?
Read too
What German founders should now learn from this
For founders from Germany, the most important change of perspective is: build a company, not just a model. Many teams confuse technical excellence with market excellence. This is understandable because AI is fascinating. But the market doesn’t pay for fascination. He pays for impact.
- That means First: Go-to-market from day one. Not: We build first and sell later. Rather: We sell while we build. Ten paying design partners are worth more than a hundred demo compliments.
- Prove ROI quickly: Especially in the B2B context, if an AI product does not have an impact in a short time, it disappears from the priority list. It must reduce costs, increase quality, reduce risks or save time. Otherwise it remains an experiment.
- Own distribution, not just technology: The sustainable AI companies not only have a model, but a workflow. You are close to the process, close to the data, close to the integration and close to the change in the company. Not “we also have a model”, but rather: we become part of the standard process.
- Understand capital efficiency as a negotiating weapon: If you are optional, you don’t beg. Anyone who can show that the company is growing even without a perfect next round will gain negotiating power. And most importantly: He survives the cycle.
This is particularly relevant for German founders. We will not win every round of capital in international comparison. But we can build companies that are closer to real industrial, regulatory and operational problems. Especially in B2B, industry, medium-sized businesses, automation, compliance, engineering and productivity, there are fields in which substance is more important than volume.
What does this mean for German VCs?
For investors, too, the crucial question is not: How do I get into the hype? But: How do I build a portfolio that survives the change in mood?
This leads to three rules:
- Follow-on beats FOMO. Anyone who invests in Seed must know beforehand how the story can continue in Series B and C. Not as a wish, but as a plan. Otherwise you end up with portfolios that starve in the middle because the music stops before the company gets to real numbers.
- Reputation signals are an accelerator, but not a replacement for substance. Cap table logos, well-known names, top funds in the background. All of this can help. It can accelerate recruiting, open partnerships and strengthen PR. But it does not replace value creation. Late investors and buyers end up paying not for logos, but for reliable cash flows, moats and distribution.
- Specialization beats convenience. In a market that is massively concentrated, generalists lose more quickly. Whoever wins has a clear thesis: certain verticals, certain workflows, certain infrastructure layers. And anyone who invests in Germany should honestly ask themselves: In which game do we have structural advantages and where do we just play because it feels good?
Read too
The super rounds are a window, not a climate
The mega rounds are real. But they are a window, not an air conditioner. When that window closes, two types of companies will be left: those with a team that can truly win the platform league. And those that have sales, traction and real customer loyalty.
Everything in between was often just a bet on mood. For Germany, this does not mean that we have to fall back. It just means we should stop confusing the top of the pyramid with the entire market.
And founders shouldn’t try to imitate American mega-rounds if they don’t have the skills to do so. You should build the things that work in every market phase: companies that solve real customer problems, draw budgets, replace processes and not just survive hype, but shape a market.