SAP Runs Half the World's Big Businesses From a German Town of 15,000. That's Not the Weakness It Looks Like.

Walldorf is a town of roughly 15,000 people in southwest Germany, the kind of place you would drive through without noticing. It is also the headquarters of a company that ended 2024 with €34.18 billion in revenue and a cloud order book — work already signed but not yet billed — of €63.29 billion.¹ No Silicon Valley campus. No glass tower in a financial capital. The software that runs the payroll, the supply chains, and the closing of the books for a huge slice of the world's largest companies is engineered, to a degree no other tech giant can claim, from a quiet German exurb. The instinct is to treat that as a quirk SAP overcame. It is closer to the reason SAP exists at all.

The official story is that SAP is a German software company that happened to go global. The truer story runs the other way: SAP went global precisely because it was German — because it grew up embedded in the factories and finance departments of Germany's industrial Mittelstand, writing code that did the unglamorous work of actually running a business, and writing it so deep into the customer that leaving became almost unthinkable.

Five engineers who walked out of IBM with a different idea

SAP was founded on April 1, 1972 by five former IBM employees — Dietmar Hopp, Hasso Plattner, Claus Wellenreuther, Klaus Tschira, and Hans-Werner Hector — under the name SystemAnalyse Programmentwicklung.³ Note what that name is not. It is not a marketing phrase. The now-famous backronym — Systeme, Anwendungen und Produkte in der Datenverarbeitung — came later, in 1976, when the sales subsidiary was created.⁷ The acronym, in other words, predates the name it supposedly stands for. That detail is the company in miniature: the engineering came first, and the story was bolted on afterward to explain it.

The founders had no venture capital and no plan to disrupt anything. They had a conviction that companies should process their data in real time, on standard software, instead of paying programmers to hand-build one-off systems forever. So they did the thing that turns out to matter most: they built the product alongside the customer, on the customer's premises, against the customer's actual processes. The first systems were written for German firms doing German industrial accounting. That is not a limitation. That is how you learn exactly how a business runs — and exactly how to make it dependent on you.

Why a customer can't simply leave

Here is the mechanism, worked down to the bottom. Enterprise software like SAP's doesn't sit beside the business; it becomes the nervous system of the business. The chart of accounts, the procurement workflow, the way an order turns into a shipment turns into a recognized euro of revenue — all of it gets encoded inside the system, customized over years, and woven into the habits of thousands of employees. Ripping it out means re-mapping every one of those processes, retraining everyone, and risking the company's ability to close its own books. The switching cost is not the license fee. The switching cost is the terror of turning the nervous system off.

That depth is why SAP's revenue is so strikingly predictable. By the end of 2024, recurring revenue had reached 83% of the total⁸ — a business that mostly knows in January what it will earn in December. And the €63 billion cloud backlog is the same fact seen from the front: it is years of committed future spending from customers who have already decided that whatever happens, they're staying.¹⁸ The Walldorf engineering culture — patient, close to the customer, allergic to the flashy rewrite — is what produced software worth being that loyal to.

And the moat is being defended in real time. SAP poured €6,514 million into R&D in 2024 — more than 19% of total revenue⁸ — most of it aimed at one transition: moving the locked-in base from software it owns and installs to software it rents from the cloud. The numbers show it working. Cloud revenue hit €17.14 billion in 2024, up 25% in a year, while old-style software-license revenue fell 21% to €1.40 billion.¹ One line is being deliberately starved to feed the other.

But isn't SAP no longer even number one?

Here is the honest objection, and it lands. The comfortable claim that SAP is the world's #1 ERP vendor stopped being true in 2024. By Apps Run The World's count, Oracle pulled ahead in ERP-specific revenue — about $8.7 billion to SAP's $8.6 billion.⁶ If the whole thesis rests on dominance, the thesis just sprang a leak.

Except dominance was never the point, and the same dataset shows why. SAP served 141,399 ERP customers in 2024, and the entire top-ten roster of ERP vendors held only 26.5% of the global market between them.⁶ This is not a market one company owns; it is a vast, fragmented field where the prize isn't share — it's stickiness. A few hundred million dollars of revenue ranking is a rounding error against a €63 billion backlog. What actually matters is whether the locked-in base follows SAP into the cloud faster than a rival can pry it loose. So far the cloud line is compounding at 25% a year while the legacy line bleeds out on schedule.¹ The real risk isn't Oracle taking a revenue crown for a year. It's a botched cloud migration loosening the one thing that made the address in Walldorf irrelevant: the customer who couldn't leave.

SAP — long since converted from SAP AG to the pan-European SAP SE⁵ — is sometimes treated as a relic that survived by being too entrenched to dislodge. That gets the causation backwards. It got entrenched on purpose, one customer's nervous system at a time, from a town most of its own customers couldn't find on a map. The provincial address was never the handicap the Valley assumed. It was the discipline. The danger now isn't that Walldorf is too far from the action. It's that the action — the move to the cloud — is the one moment a 52-year-old moat can be drained from the inside, by SAP's own hand, in front of everyone.