Latest / The Pivot Podcast with Fexingo: How Businesses Change Direction and Survive Disruption / How IBM Pivoted From Mainframes to Hybrid Cloud
Transcript
- Lucas: When you think of a company that defines 'old tech', IBM is usually at the top of the list. Founded in 1911. Built on tabulating machines, then mainframes, then personal computers — and yet, somehow, it's still here, still relevant, still a $140 billion company. Luna: And not just surviving — they're actually growing again. I mean, the narrative on IBM for years was that it was a slow-moving dinosaur that missed the cloud era completely. Lucas: Right. And that story was mostly true until about 2019. But the pivot IBM has executed since then — from a hardware and services dinosaur to a hybrid cloud platform company — is one of the most underappreciated turnarounds in tech history. We're talking about a 114-year-old company that essentially bought its way into the future. Luna: And spent the most it ever has on a single acquisition to do it. $34 billion for Red Hat. Lucas: Exactly. That's the anchor of this whole story. July 9th, 2019 — so almost exactly seven years ago as of today — IBM closed the Red Hat deal. At the time, everyone thought they overpaid. Red Hat was an open-source software company with about $3 billion in revenue. $34 billion was a 60% premium. Luna: But it wasn't about Red Hat's existing revenue. It was about owning the operating system of hybrid cloud. Lucas: That's exactly the bet. See, Amazon, Microsoft, and Google had already locked up the public cloud — AWS, Azure, GCP. IBM was never going to beat them there. But most large enterprises — banks, insurance companies, governments — they run on a mix of on-premise data centers and multiple clouds. That's the hybrid cloud. Luna: And IBM's thesis was that Red Hat's OpenShift platform would become the layer that lets you manage workloads across all those environments. A kind of universal operating system for the enterprise. Lucas: Exactly. And it's worked. Red Hat's revenue has more than doubled since the acquisition — it was about $7.2 billion in 2025. OpenShift now has thousands of enterprise customers. And more importantly, it's giving IBM something it hasn't had in decades: a platform that developers actually want to build on. Luna: It also gave IBM a subscription-based revenue model. Instead of selling one-off mainframes and consulting projects, they're getting recurring revenue from Red Hat subscriptions. That's a huge shift for a company that used to be totally transactional. Lucas: And that recurring revenue is now about 40% of IBM's total. Their gross margins have climbed from around 46% in 2019 to about 56% last year. That's a massive improvement for a company that size. Luna: But it wasn't just the Red Hat acquisition. They also did something really painful but necessary: they started selling off the old stuff. Lucas: Right. The most significant was spinning off Kyndryl — their managed infrastructure services business — in November 2021. That was about $19 billion in revenue that they basically said goodbye to. It was low-margin, declining revenue that was dragging down the entire company's growth numbers. Luna: And after the spin-off, IBM suddenly looked like a higher-growth, higher-margin company. The stock has more than doubled since then. Lucas: So the pivot has three legs. One: acquire the platform — Red Hat. Two: divest the legacy — Kyndryl and later some other low-margin businesses. Three: reorganize the entire company around that hybrid cloud and AI strategy. In 2021, CEO Arvind Krishna said IBM would invest $1 billion in Watson's successor efforts for hybrid cloud — which a lot of people laughed at, but now with generative AI, it looks prescient. Luna: Let's talk about that. IBM's AI play — Watsonx — is built on the same open-source foundation as Red Hat. It's designed to run anywhere, not just on IBM's cloud. That's a different bet than Microsoft with OpenAI or Google with Gemini, which are tightly coupled to their own clouds. Lucas: It's actually a smarter bet for the enterprise market. Most large companies are not going to put all their sensitive data into a single public cloud. They need AI that can run on-premise or in a private cloud. IBM's Watsonx lets them do that. And because it's built on Red Hat OpenShift, it integrates with whatever infrastructure they already have. Luna: So the whole strategy is 'meet the customer where they are'. Not 'force them to move to our cloud'. Lucas: Exactly. And that's the through-line from the mainframe era to today. IBM's original genius was making compatible systems that big organizations could rely on. They lost their way in the 1990s and 2000s by trying to be everything to everyone — owning the hardware, the software, the services. The pivot is about being the glue that holds together a heterogenous IT environment. Luna: And that's a business model that actually has more durability than being a public cloud provider, because it's less dependent on keeping customers locked in. If your value is in the interoperability layer, customers trust you more. Lucas: There's a lesson here for any legacy business facing disruption. The instinct is often to copy what the disruptors are doing — 'we need our own AWS.' That almost never works because you can't out-Amazon Amazon. But you can find the adjacent space where your existing customer relationships and trust give you an advantage. Luna: IBM had decades of relationships with CIOs at the world's biggest companies. That's what allowed them to sell Red Hat subscriptions into those accounts. A startup couldn't do that. Lucas: Right. They used their distribution muscle to scale Red Hat. And they used Red Hat's open-source credibility to make IBM cool again — or at least less uncool — to developers. That combination is what made the pivot work. Luna: It also helps that they have a CEO who's willing to make hard decisions. Arvind Krishna took over in April 2020 — right at the start of the pandemic. He immediately accelerated the Kyndryl spin-off and doubled down on Red Hat. He didn't try to preserve the old business. Lucas: And that's rare. Most CEOs in that position would try to do a little bit of everything — keep the legacy cash cow alive while dabbling in the new thing. Krishna made a clean break. It was painful — Kyndryl laid off thousands, IBM's revenue dropped for a couple of years — but now the company is growing again. Revenue was up about 4% last year, and free cash flow hit $12 billion. Luna: I think that's the most important business lesson from IBM's pivot. A real pivot requires letting go of something that's still generating revenue today, because it's going to drag you down tomorrow. Lucas: Exactly. And that's incredibly hard to do when you're a public company and analysts are asking every quarter why your revenue is down. You have to have conviction in the thesis and be willing to take the short-term hit. Luna: If today's episode gave you a new way to think about your own company's strategy, or just a useful example to cite in a meeting, that's exactly why we make this show. And the way we keep it ad-free and independent is listener support. If you found value here, you can buy us a coffee — literally — at buy me a coffee dot com slash fexingo. It's a small gesture that keeps these conversations going. Lucas: Yeah, we don't run ads, we don't have sponsors. It's just us and the topic. So if you appreciate that, the link is buy me a coffee dot com slash fexingo. And now back to the pivot. Luna: So where does IBM go from here? The hybrid cloud story is largely executed — Red Hat is integrated, Kyndryl is gone. What's the next act? Lucas: The next act is all about AI — but not the consumer AI that gets all the headlines. IBM is betting that enterprises will need help deploying AI in a secure, compliant, and explainable way. That's Watsonx's pitch. They also have a consulting arm — IBM Consulting — that's helping companies implement AI. That's a services business, lower margin, but it pulls through the software. Luna: And of course, they still have the mainframe business. Believe it or not, mainframes still process 68% of the world's production IT workloads. IBM launched a new z16 mainframe in 2022 with integrated AI — basically a mainframe that can do real-time fraud detection. That's a pivot within a pivot: taking their oldest asset and making it relevant to modern AI use cases. Lucas: That's actually a great point. The mainframe isn't going away. Banks and airlines can't just rip and replace it. So IBM's play is to make the mainframe work with the hybrid cloud. They now have a product called IBM Z and Cloud Modernization Center that helps companies connect their mainframes to OpenShift. It's about making the old new again. Luna: So the whole IBM story is really about asset reuse. They had the mainframe relationships, they had the consulting army, they had the enterprise trust. They bought Red Hat as the new engine, but they kept the old chassis. Lucas: That's a perfect way to put it. And the results speak for themselves. IBM's stock is at an all-time high as of mid-2026. The company is generating strong free cash flow. It's not the high-growth story of a tech startup, but for a 114-year-old company, it's a remarkable reinvention. Luna: I think the key takeaway for business leaders is: don't be afraid to use acquisitions to buy a capability you can't build. But don't just acquire — integrate. IBM spent years integrating Red Hat, not just bolting it on. That's where the value comes from. Lucas: And be willing to sell off parts of your business that are no longer core. That's the hardest thing, but it's what separates a real pivot from a slow decline. IBM could have kept Kyndryl and tried to squeeze a few more years of profit out of it. Instead, they let it go and focused on the future. Luna: So next time someone tells you that big, old companies can't change, point them to IBM. It took a decade, $34 billion, and some painful divestitures, but they did it. Lucas: Yeah. And they're proof that a pivot isn't about becoming a completely different company — it's about finding the thread from your past that connects to the future, and then pulling it hard.