One pattern keeps coming up in my recent tech updates. People start discussing re-platforming without understanding what that would cost. More importantly, they haven’t always asked whether the IBM i platform is really causing their pain.
In many cases, IBM i isn’t the issue at all. The push to re-platform often comes from old interfaces, which can be modernized instead of replaced. A Windows environment can have a horrible interface too. More often, it’s an application problem. Companies also have tons of data on IBM i, and it makes more sense to update and modernize what they already have than to start over.
Fortra’s 2026 IBM i Marketplace Survey points in the same direction. Of the IBM i professionals who responded, 73 percent run applications written in-house, and 62 percent named application modernization as a top concern. Those numbers tell me the applications, and the interfaces in front of users, deserve a closer look before anyone replaces the platform underneath.
A few weeks ago I spoke with a client who showed me how fast this thinking can run away with a decision. They had replaced their internal IT with a new team that was coming in to re-platform. When I asked why, the answer was that nothing runs modern on the IBM i. In fact, there are tons of cutting-edge ERP systems on the platform, some more modern than the Windows interfaces people hold up as the alternative.
Stability is also a big strength of IBM i on the back end. I showed this client one of my own systems, running for 521 days without ever being shut off or turned down, and asked them to show me a Windows system that runs a business like that. A little education got this client to look at the problem differently and evaluate their plans instead of assuming that just because something runs on IBM i, it’s old. Getting rid of IBM i won’t solve all problems and could create new ones.
For executives, the decision comes down to cost. Before moving to a newer platform, step back and spend a little time understanding the true problem. I’ve been digging into the total cost of ownership since sitting down with IBM i Chief Engineer Kris Whitney in Rochester, and the hidden expenses really stand out.
A re-platform isn’t just trading IBM hardware for x86. You also have to migrate business data sitting on the system for 20, 30, or even 40 years. That significant business burden is amplified if something goes wrong. A GAO review of one federal agency’s effort to replace a financial system more than 20 years old shows how that can play out. The project fell behind schedule, its estimated cost rising by $13.4 million, to $71.9 million. About $4 million of this price adjustment came from reverse-engineering the old system’s interface requirements, building an integration layer, and developing test plans.
Modernizing the right way also means remembering what a business needs, no matter which platform it runs on. Automation, monitoring, security, and high availability are required either way, and going to Windows doesn’t mean you get all of those pieces on the other side. You still need solutions to cover them. Security in particular deserves serious attention, and a hasty switch might leave organizations vulnerable.
No doctor would suggest a heart transplant if the patient’s heart is healthy. Wanting the most up-to-date system is understandable, but it’s important never to confuse novelty with reliability, or to assume a system that’s older is also less robust. Make sure the IBM i is really the problem before you replace it. Old interfaces can often be modernized, while the true cost of re-platforming goes well beyond hardware. Solve the business problem in front of you, rather than creating a different set of complications with a new system.