Oracle and SAP share the same incentives. Only one is built to protect its product.
The views below are my own, based on publicly available information. Oracle and SAP would each tell this story differently, and I welcome that conversation.
Compare Oracle and SAP long enough and you stop seeing two software companies. You start seeing two answers to a single question: who is a company actually for? The products, the support deadlines, the capital strategies are all downstream of that question. And the two firms answer it differently because their countries make them.
The same incentives, two different boardrooms
Start with what they share. Both are public companies built to make money, and both are very good at it. The profit motive is identical; neither is greedier than the other. What differs is who sits in the room when the decisions get made, and that is set by law, not character.
Why US corporate law rewards harvesting
And it is worth saying plainly, because it is easy to get wrong: this is not about bad people. The engineers and support staff closest to JD Edwards often care about it as deeply as any SAP team cares about theirs, and they spend their days trying to add value. But they have no say in its future. They can polish the deck chairs and serve their lovely guests drinks all they want, but they are still standing on the Titanic, and they cannot change its course. That course, JD Edwards as a cash cow to be milked, maintained but never given a roadmap, is set above them, by people with no structural obligation to it. The good people are there. They just have no say.
Why German codetermination protects the product
Germany runs the same engine with a different boardroom. No German law requires a company to stay focused on its core product, and SAP is free to diversify. But SAP is governed under German codetermination, or Mitbestimmung, which puts employees in roughly half the seats on the supervisory board (the Aufsichtsrat), the body that hires and fires top management. The management board's own legal duty is to the interest of the enterprise as a whole, balancing shareholders, employees, and the public good, not shareholders alone. The people whose craft, careers, and livelihoods are bound up in the product being good hold real, binding power over its future.
That is why SAP looks more product-focused, and it is worth being precise about why. It is not that SAP is the more virtuous company. It is that when the people who build the product hold board power, nobody votes to hollow out the engineering organization to fund a financial pivot. The integrity is real in its effects, but it is manufactured by structure, not chosen by character. So SAP does the rational thing for its system: it sunsets ECC into S/4HANA, a real successor, and keeps pouring R&D and AI into the product its customers actually run.
And this is not a recent experiment. Board-level codetermination has been German law since 1922. Hitler abolished it in 1933, and after the war the Allies and German unions rebuilt it deliberately, to keep concentrated industrial power from ever again answering to a regime. Germany extended it to all large firms in 1976. The board that quietly protects your SAP investment today is the direct descendant of a century-old decision about who industry is allowed to serve.
What it means for your roadmap: JD Edwards versus S/4HANA
Watch where that leaves you, the customer. Be honest first: neither path is cheap or quick. ECC to S/4HANA is a major, multi-year program, and plenty of shops effectively reimplement to get there. But the difference is not the effort; it is the destination. S/4HANA is the designated successor to ECC, the same vendor's flagship evolved, with a conversion path that can carry much of your configuration and data forward. JD Edwards has no such successor. Oracle slides support past 2037 with no real deadline, and it offers a path to Fusion, its flagship, with migration tools and partners to match.
But Fusion is a different product, not an evolution of JD Edwards: there is no in-place upgrade, only a re-implementation onto a system built finance-first, thinner for the manufacturing, construction, and project-based industries JDE was built to serve. And whatever migrates, it is not your JD Edwards at the end of it: the customizations and integrations you shaped to the business over the years do not make the trip; they are rebuilt from scratch or abandoned. You may reimplement either way. The difference is what waits at the end: one vendor is investing in the destination it is sending you to; the other is sending you onto a product built for someone else, while the one you run gets no future at all.
The limits of the comparison
Be fair about the limits. SAP is no saint: a profit-seeking, publicly traded German company that runs layoffs and builds its own cloud lock-in. And codetermination has a price, since German firms are often slower and less willing to make a radical bet. For most software that would be a weakness. But for ERP, a system you run for twenty or thirty years, continuity is exactly the integrity that matters.
And be fair to the American model, because it can produce fierce product loyalty too, usually through a single founder with both the conviction and the control to resist the system, the way Elon Musk has kept Tesla on its mission when the incentives would have rewarded a pivot. JD Edwards once had exactly that founder. Ed McVaney, the Ed in J.D. Edwards, built it in Denver and ran it as a product its mid-market customers loved, the way you defend something with your name on it. But founder loyalty is personal, not structural; it lives and dies with the founder. PeopleSoft acquired JD Edwards, Oracle acquired PeopleSoft, and the moment the product passed from the man who built it to an owner optimizing for the next platform, the harvesting began.
Larry Ellison is every bit as powerful a founder as McVaney was, yet Oracle harvests JD Edwards without hesitation, because his conviction points at the next profit engine, not the products that built the last one. JDE's protector was not defeated; he was bought out, and nothing structural took his place. Codetermination does not need a hero, which is exactly why it keeps protecting the product long after any individual believer is gone. Tradeoffs, not heroes and villains.
So greed was never the real story. Both firms want your money with equal appetite. The difference is whose interests are seated at the table when the decisions get made, and everything downstream, the roadmap, the support horizon, the migration bill, flows from that.
That is the part worth sitting with. The most consequential decision about your next decade was not made in a sales meeting or a roadmap deck. It was made years earlier, in law: German codetermination on one side, American shareholder primacy on the other, written by people who will never know your name. A support date that keeps sliding is not a reprieve, it is a business model. A vendor's loyalty to its own product is not virtue; it is the boardroom that virtue gets built into.
What you actually control
You cannot legislate a board. You could vote with your feet, migrate off JD Edwards to a vendor like SAP, and inherit governance that actually protects its product. But that is a multi-year rip-and-replace to solve a governance problem, and it trades away the fit your business already runs on. A real option for some; a free one for no one. For the many who will run JD Edwards for years yet, the lever is not the vendor at all. You have just turned the variable that matters most from a surprise into a known quantity. Oracle's incentive to harvest JD Edwards does not mean the software breaks, and it does not mean support stops. EnterpriseOne 9.2 gets patches, security fixes, and regulatory updates well into the next decade; the maintenance is real, and plenty of companies will run it profitably for years.
What you will not get is a future for the product itself. JDE is not the flagship, there is no successor that preserves your investment, the sales motion is built to move you onto Fusion, and the partner, ISV, and talent ecosystem around it has been allowed to thin. Be clear-eyed: no wave of vendor R&D is coming, and operating the system brilliantly will not change that. Product and operations are separate dimensions, and no amount of the second buys you the first.
Where the two meet is your budget, and that is the whole game. A dead-end product does its real damage through the run-rate. Let support costs drift up year after year and you spend everything just to stand still, with no money, no capacity, and no runway left to deal with the future you were just warned about. That is how a shop gets stampeded into a migration on the vendor's timeline, underfunded and on the back foot.
So the play is not to out-engineer Oracle. It is to compress the run-rate hard and deliberately, until JD Edwards costs a fraction of what it does now, and to hold the difference as optionality. Then the real decision, migrate now, migrate later, or run lean for years, is one you fund and time on your terms, not one the vendor forces. A vendor with no structural reason to protect your product is not a reason to panic. It is a reason to get the cost down while the choice is still yours, with someone operating it whose interests are aligned with keeping it that way.
Read each vendor as the output of its ownership, not its goodwill. Then stop waiting on the boardroom you cannot change, and go build the operating model you can.
