AI in Sales: Why SMEs are Losing Out

Marktanalyse · 25. April 2026 · Rebecca Kupka

Billions are flowing into AI Sales, but SMEs are footing the bill. A ruthless market analysis of why AI in sales needs a different approach.

Do you know that moment? You're sitting in one of those typical German executive boardrooms – lots of beech veneer, the obligatory foosball table in the hallway as an alibi for modernity – and the sales director proudly presents the new wonder weapon. A software, "AI-powered" of course, that supposedly puts sales on autopilot. I experienced it again just last week. Eyes lit up, PowerPoint slides promised double-digit growth rates. And all I could think was: they're not just putting the cart before the horse, they're harnessing it from the wrong side of the globe.

The bold thesis I'm presenting to you today is brutal but necessary: the entire hype around AI in sales, this flood of billions upon billions of dollars currently being pumped into AI sales startups, isn't just bypassing German SMEs – it will ultimately cost them dearly. We are on the sidelines, cheering on a game whose rules we don't understand and in which we've already lost before kick-off.

The Great Illusion: Why We Completely Misunderstand the Game

Honestly: most managing directors and sales managers in our latitudes still believe in a fairy tale. The fairy tale of purchased efficiency. You read in the Handelsblatt about the wonders of artificial intelligence, hear about a 1.5 percent productivity increase predicted by BlackRock, and think: "Great, I'll buy a piece of that pie." You look for a provider – usually one with a fancy office in San Francisco or Tel Aviv – book the demo, sign the SaaS contract, and wait for the promised revenue miracle.

This isn't just naive. It's extremely dangerous. Because what we're seeing here isn't a normal market cycle. This isn't a fair competition of the best ideas or the cleverest algorithms. What we're experiencing is the cementing of a new world order, driven by a handful of American hyperscalers who are building a closed system with unimaginable sums of money. And us? We dutifully provide the data and the monthly subscription fees. We are degraded from creators to users – or worse: to digital subjects.

The fallacy lies in the assumption that it's about technology. But it's about capital. It's about power. It's about who controls the digital infrastructure of the 21st century. And while in German conference rooms, people are still debating the intricacies of GDPR when implementing a new CRM plugin – a topic I by no means want to downplay, but it's not the whole picture – facts are being created on the other side of the Atlantic. Facts in the form of 13-digit dollar amounts.

The Uncomfortable Truth: A Look at the Numbers Nobody Wants to See

So let's look at this soberly. Forget the marketing brochures for a minute. Sequoia Capital, one of Silicon Valley's godfathers, is launching a $7 billion fund – explicitly for AI investments. Andreessen Horowitz (a16z), the other big name, is doubling down on AI startups. We're not talking about peanuts here. In 2025 – that was last year, not in the distant future – over 50 percent of all global venture capital went into AI companies. Half! For the first time, a single sector has dominated the entire VC landscape to such an extent.

But that's just the tip of the iceberg. The truly insidious game is played on a deeper level and is called "Circular Financing." A circular financing deal. Sounds harmless, but it's key to understanding the new power architecture. The principle is as simple as it is brilliant: Microsoft "invests" $13 billion in OpenAI. A large part of this money flows back to Microsoft immediately because OpenAI needs the computing power of Microsoft's Azure cloud to train its models. Oracle commits to providing data centers worth $300 billion for OpenAI's "Stargate" project, securing a customer for decades. OpenAI, in turn, receives shares in cloud provider CoreWeave worth $350 million to secure its computing power. Do you notice something? The money never leaves the ecosystem.

It's a way for a company to secure its own success and growth by financing it on its own terms. Companies do this because it fits their narrative and keeps them at the forefront of the market, where cash and speed are paramount.

— Omar Rajjoub, VP at HudsonPoint Capital

Rajjoub hits the nail on the head. This isn't an investment in the classic sense, where capital meets an innovative idea. This is strategic power consolidation. The tech giants – NVIDIA, Microsoft, Amazon, Google – control the scarce resources: the chips and the cloud infrastructure. They force the most promising startups into their system and finance their growth with money that ultimately returns to them. A perfect cycle from which there is hardly any escape. And every medium-sized mechanical engineering company that buys an "AI-powered" tool from one of these startups feeds this system.

But Productivity! The Strongest Counter-Argument – and Why It Doesn't Hold Up

Now I can already hear the objection from the controlling department: "Mr. Müller, that's all well and good, but if AI ultimately brings us 1.5 percent more productivity, as BlackRock says, then it pays off! That's $1.1 trillion in additional economic output!" Yes, that sounds fantastic. A dream. A margin that has not been dreamed of in the manufacturing industry since the golden years of the economic miracle.

The question no one asks is: who exactly pockets these $1.1 trillion? Do you think they end up one-to-one in the accounts of medium-sized users? I allow myself a certain skepticism there. BlackRock's analysts say it themselves, if you read between the lines: the tech companies will secure their share through "market expansion and IP monetization." In plain English: they will skim off the profits. The business model has shifted. Previously, you bought a software license. A one-time investment, then it was yours. Today? Subscriptions. Monthly, annually increasing fees for access to the black box. You are not investing in an asset; you are renting a service. And the landlord – in this case, a monopolist with virtually infinite resources – can increase the rent at any time. There's no getting around that.

The risk is completely shifted to the user. The hyperscalers make their investments in infrastructure with debt capital that they raise through bonds on the capital market. The SME, on the other hand, has to stem the rising subscription costs from its operating cash flow. It finances the party but is not allowed to celebrate. That's the deal. And many don't even notice it.

What I See in Practice: AI in Sales Between East Westphalia and Silicon Valley

I travel a lot. I'm as at home in the factory halls of Swabia as I am at tech conferences in Las Vegas. And the contrast couldn't be greater. On one side, there's this mechanical engineering company in East Westphalia, a hidden champion, a global market leader in its niche. The sales manager, a capable man, was convinced to implement an American AI sales platform. Cost: 60,000 euros per year. The result after six months? Frustration. The AI spits out leads that have nothing to do with the highly specialized product. The "automated" emails sound like a bad translation. Why? Because the AI was trained with B2C data from the US market and doesn't understand the concept of a German purchasing committee for a 2-million-euro plant. The thing is: the data basis is wrong.

And on the other side? The USA. There, it's not about implementing such tools. It's about creating the foundations. About controlling computing power. And in Europe? What is our answer to Sequoia's $7 billion funds? Well, former Ryanair manager Peter Bellew is launching a fund for AI in aviation. Ticket sizes: between 21,000 and 100,000 euros. Don't get me wrong, every initiative is good. But it shows the discrepancy in dimensions. We're playing with water pistols in the sandbox, while aircraft carriers are fighting next door.

The Great Divide: A Regional Comparison That Hurts

The data here is clear and depressing. While deals in the triple-digit billions are becoming normal in the US to secure basic infrastructure, we are holding conferences on proper governance. I recently attended PLEXUS's "AI Compact" in Liechtenstein virtually. A clever mind, Marco Wunderlich from GET Capital, said the quote of the day there: "Technology is not the problem - governance is." That perfectly sums up the German – indeed, the European – soul. We want to define all the rules for a game that has long been running and whose score is already 10:0 for the other team. We are world champions in raising concerns.

The Amplifa ICP Playbook: The Foundation for Real Sales Success Before you invest a single cent in AI software: Do you even know who your ideal customer is? This playbook shows you how to sharpen your Ideal Customer Profile (ICP) – the absolute basis for any intelligent sales strategy.

Questions from the Executive Suite: What SMEs Need to Know Now

Has the AI train already left the station for German SMEs?

No, absolutely not. But you have to stop trying to jump on the American express train. It's too fast, too expensive, and it's going in a direction that isn't yours. The right path for SMEs is the regional train: slower, but it stops exactly at the stations that are relevant to you. This means: focus on niche problems that the large general-purpose AIs cannot solve. Your strength is your deep domain knowledge, not the ability to process terabytes of data.

How do I recognize a truly good AI sales solution?

Quite simply: a good solution asks you more questions at the beginning than it gives answers. It wants to understand your processes, your customers, your specific challenges. A bad solution shows you a shiny dashboard and promises "more leads." Be extremely skeptical of providers who claim to have a "one-size-fits-all" solution for AI in sales. For complex B2B sales in German mechanical engineering, there isn't one off the shelf. Look for providers who respect your data sovereignty and give you control over the algorithms, instead of swallowing them in a global black box.

Should we now invest in AI development ourselves?

For 99% of SMEs, the answer is: God forbid, no. You cannot and will not compete with OpenAI, Google, or Anthropic. That would be like trying to build your own car factory just because you need a new company car. Instead, invest your money and time in what you can control: your data. Clean up your CRM data. Standardize your processes. Build a clean, structured data foundation. This is the currency of the future. And it's more valuable than any "intelligent" algorithm you could buy.

Amplifa Outbound Engine: The Alternative to the Shotgun Approach Once the ICP is established, it's time for targeted outreach. Instead of sending thousands of irrelevant emails, this method helps you reach the right decision-makers at the right companies with a relevant message.

What Needs to Happen Now: A Four-Step Appeal

Sitting around and waiting for better times is not an option. Nor is complaining. A strategic reorientation in the minds of decision-makers is now vital. In my experience, there are four steps that are now crucial for survival:

1. Stop chasing the hype: Stop running after every digital trend. Take the time to understand the actual mechanisms behind buzzwords like "AI Sales." Don't just read the success story, but also the fine print. A healthy cynicism is worth its weight in gold here.
2. Focus on unfair knowledge: You know more about your customers and your market than any AI in the world. This "unfair" knowledge is your greatest asset. Instead of trying to adapt your processes to generic software, look for ways technology can scale your specific knowledge. That's the crucial difference.
3. Data sovereignty as the highest directive: Treat your sales and customer data as the most important raw material of your company – because they are. Every decision for new software must be made under the premise of data sovereignty. Where is the data? Who has access? Can I extract it at any time? Anyone who ignores this commits strategic suicide on installments.
4. Smart partnerships instead of blind purchasing: Don't look for suppliers, but for partners. These can be smaller, specialized European startups that understand your business model. Or research institutes. Or even competitors with whom you can build a neutral data platform. Enter into cooperations to form a counter-power to the hyperscalers, instead of individually submitting to their dictates.

Amplifa ABM Framework: Precision Instead of Shotgun Account-Based Marketing is the antithesis of the mass-market approach of large AI platforms. Concentrate your resources on the few, highly profitable target customers. This framework shows you the way.

The Americans have the capital. The Chinese have the data. And us in the DACH region? We have the engineering skill, the quality, and a loyal customer base, grown over decades, in the most complex B2B markets in the world.

The crucial question for the coming years is therefore not whether we use AI in sales. But how we use it to expand our unique strengths, instead of becoming an interchangeable cog in the machinery of global tech giants. Because at the end of the day, we all want the same thing: not just to build the best products, but to sell them damn well too. Our way. Discuss with me on LinkedIn. I'm curious about your – unvarnished – opinion.