On April 1, 1972, five engineers walked out of IBM’s Mannheim office in West Germany and founded a company they called Systems, Applications, and Products in Data Processing.
Their names were Dietmar Hopp, Hasso Plattner, Klaus Tschira, Hans-Werner Hector, and Claus Wellenreuther. They had been working together at IBM on a project they thought was important. IBM had cancelled the project. The five engineers thought IBM was wrong. So they quit and started a company to build what IBM had refused to build.
Their first customer was ICI. Their first product was called R/1, released in 1973. R/2 followed in 1979. R/3 in 1992. Then a long sequence of subsequent products — mySAP, S/4HANA, and dozens of others — extended the platform over the following decades.
Today, SAP is the world’s dominant enterprise software company. Approximately 87% of global commerce transactions pass through SAP systems at some point. It is one of the most successful business software companies in history.
And here is the detail almost no one appreciates. The project the five engineers started in 1972 is, in a specific technical sense, still unfinished.
What they were trying to do
The project that IBM cancelled and that the five engineers thought was worth quitting for was real-time integrated data processing for enterprise operations.
The vision, in its original 1972 form: an enterprise’s financial system, its inventory system, its production system, its sales system, and its human resources system should all operate on the same underlying data, updated in real time, with any transaction in any part of the business immediately visible in every other part.
This vision was, in 1972, technically impossible at the scale required. The computing infrastructure did not exist. The database technology did not exist. The networking capacity did not exist.
Fifty-four years later, most of the technical pieces do exist. And yet, in most enterprises deploying SAP in 2026, the original vision has still not been fully realized. Most enterprises still have data silos. Still have batch reconciliation processes. Still have finance closing periods that require days of work.
The unfinished project is not a technical failure. It is a business reality. Enterprises change more slowly than their technology permits.
What this means for AI in 2026
Every senior leader I work with in 2026 is being sold the idea that AI transformation is a three-to-five year project.
The SAP story is the cleanest available counter-example. The transformation the five engineers started in 1972 has taken fifty-four years and is still not complete. It has produced enormous value along the way. But the underlying ambition remains a multi-generational commitment, not a project timeline.
The senior leaders who understand AI transformation as a multi-generational commitment will make dramatically different decisions than the senior leaders who understand it as a project.
Multi-generational commitments require ongoing investment. Project mindsets look for completion.
Multi-generational commitments accept partial progress. Project mindsets require deliverables.
Multi-generational commitments compound. Projects end.
The Hopp lesson
Dietmar Hopp, one of the five founders, has said in interviews that the project they started in 1972 was not a business decision. It was a conviction about what enterprise operations should look like — and about what work was worth doing.
The conviction survived IBM’s cancellation. It survived multiple technology paradigm shifts. It survived the transition from mainframe to client-server to cloud. It survived the arrival of dozens of competing frameworks and vendors.
What made the conviction durable was that it was rooted in a fundamental observation about what enterprises needed, not in a specific technology or a specific market opportunity.
Three practical questions
One: is your organization’s AI transformation a project or a multi-generational commitment? The distinction shows up in specific decisions. Budget structure. Reporting cadence. Success metrics. Leadership tenure expectations. Projects have all of these on short cycles. Multi-generational commitments do not.
Two: what is the underlying conviction that would sustain your AI transformation across five different technology paradigms? If the answer is “we want to be more efficient,” you have a project. If the answer is a specific claim about how work should be done and what the enterprise should be capable of, you have a conviction.
Three: what is your organization’s version of the unfinished project? The specific ambition that will not be fully realized in a decade but that is worth pursuing anyway because the compounding value along the way justifies the ongoing commitment.
The closing thought
The five engineers who founded SAP in 1972 did not know their project would take fifty-four years and would still be incomplete. They knew that the project was worth starting and that IBM was wrong to cancel it.
The senior leaders who make similar bets on AI transformation in 2026 — bets that require multi-generational commitment, that will not fully complete within any single leader’s tenure, that will compound value across decades — are the ones whose organizations will be dominant in 2050.
The senior leaders who require project completion within their tenure will produce project deliverables. Their organizations will not compound. Their successors will not inherit multi-generational capability. The bet will not have been made.
The world has changed. The leaders who notice will be the ones the next decade is built around.
