The 30% That HP Cannot Ignore and Microsoft Has Already Stopped Pretending Doesn't Exist
When the chief financial officer of one of the largest PC manufacturers on the planet mentions, in a conversation with investors, that 30% of its installed base is still running Windows 10, she is not communicating a technical anecdote. She is describing the anatomy of a replacement cycle that has not yet ended and that, precisely for that reason, continues to generate revenue for the entire supply chain.
Karen L. Parkhill, Executive Vice President and Chief Financial Officer of HP Inc., stated that figure in the context of explaining the company's growth in EMEA and APJ during the last quarter. Her argument was optimistic: the migration to Windows 11 is working as a sales catalyst in regions that have historically lagged behind North America. But the data point that continued to resonate in specialized markets was the other one, the one that looked backward: three out of every ten HP devices currently in circulation have still not made the transition.
What is relevant about that figure is not its size but its origin. For years, statistics on Windows version adoption came from sources with well-known limitations. Statcounter estimates based on web traffic, with geographic and device-related biases. Steam's monthly survey captures exclusively users of that platform, who do not represent any corporate or institutional profile whatsoever. None of those sources has access to what HP does have: the actual data from its own installed base, support contracts, and active hardware records. The fact that a CFO states this in front of investors — under the obligations of truthfulness that this implies — turns that number into an entirely different category of data.
What the Data Reveals About How the Market Actually Moves
The standard narrative of recent months painted a relatively orderly transition. Windows 11 surpassed 70% desktop market share according to Statcounter by mid-2026. Manufacturers reported sustained demand for new PCs. Microsoft had declared the end of support for Windows 10 in October 2025. Everything pointed to a migration already well underway.
But the HP figure forces us to adjust that narrative. If the second-largest PC manufacturer by global unit volume — holding approximately 21% of the shipped units market — reports that a third of its installed base has not migrated, the absolute number of active Windows 10 devices in the world remains enormous. We are not talking about a marginal segment or resistant users at the fringes. We are talking about a substantial portion of the corporate and domestic computing infrastructure that continues to operate on a system without active support from its manufacturer.
That has direct and measurable consequences. The first: Microsoft responded to that reality before HP confirmed it publicly. The extension of the Extended Security Updates program — known as ESU — until October 12, 2027, was not a gesture of technical generosity. It was a decision made in response to signals coming from the market indicating that the mass migration was going to take longer than the official calendar admitted. Someone in Redmond had information similar to what HP has just confirmed publicly.
The second consequence is less visible but equally relevant: the ESU converts Windows 10 into a late-payment product. Organizations that failed to complete their transition within the original deadline now have the option to purchase additional time. That time has a price. And the size of the Windows 10 base — now validated with internally sourced data from an OEM — justifies exactly that model: there is sufficient demand for the program to be financially viable for Microsoft and sufficiently attractive for organizations to adopt it rather than rushing into a costly migration.
Why Large Organizations Did Not Migrate When They Should Have
The most interesting question is not how many devices are still on Windows 10. It is why they are still there after official support ended.
The answer has nothing to do with negligence or lack of awareness. It has to do with the actual structure of replacement cycles in medium and large organizations. A company with 5,000 devices does not migrate because Microsoft changes a date. It migrates when three conditions converge: that the hardware budget is available, that the internal systems dependent on the operating system are compatible with the new version, and that the IT team has the operational capacity to execute the process without disrupting operations. Those three conditions rarely align simultaneously with the calendars dictated by a software vendor.
The result is predictable and recurrent. It happened with Windows XP, which remained the majority operating system long after Microsoft announced its end of life. It happened with Windows 7. And it is happening now with Windows 10, albeit on a smaller scale thanks to the fact that Windows 11 has a faster adoption cycle than its predecessors.
What HP revealed is that that lag is not a distortion of the market: it is the market. Organizations make infrastructure decisions based on their own fiscal calendars, their dependencies on legacy software, and their internal execution capacity. Manufacturer deadlines are pressure points, not operational mandates. The gap between the official calendar and the actual adoption behavior is the friction that no product announcement alone can eliminate.
For HP, that friction is a documented opportunity. Every percentage point of the 30% that has not yet migrated represents devices that, at some point over the next 12 to 24 months, will need replacement or upgrade. HP's CFO was not reporting a problem: she was describing a sales pipeline with a known expiration date.
The Window That Closes in 2027 and What Each Stakeholder Must Resolve Before Then
October 12, 2027, is not merely the end of the Windows 10 ESU program. It is the moment when the last safety net disappears for organizations that have not completed their transition. After that date, devices running Windows 10 will continue to function, but without security patches, exposed to vulnerabilities that will accumulate without any official remediation. For any organization with regulatory compliance obligations, that simply ceases to be a viable option.
This means that the next wave of hardware and operating system renewal has a harder deadline than the previous one. And unlike the original October 2025 deadline, the 2027 cutoff arrives after organizations have already consumed one extension. The market's tolerance for a third postponement is significantly lower.
For HP and other manufacturers, that horizon defines a window of predictable and bounded demand. The growth that Parkhill described in EMEA and APJ is not the peak of the cycle: it is the beginning of its most intense phase. The organizations that are migrating now are those that first met the necessary conditions. Those still on Windows 10 are the ones with the most complex friction points, which means they will require more time, more support, and, in all likelihood, higher unit-value equipment due to the hardware requirements that Windows 11 demands.
For Microsoft, the period leading up to 2027 is a double monetization opportunity: revenue from ESU paid by organizations buying additional time and revenue from Windows 11 licenses as devices are renewed. Those two streams are not contradictory; they are sequential. The ESU finances the transition, and the transition generates the sale of a new operating system.
What the HP figure makes visible, beyond the specific statistic, is that major technological transitions do not follow the deadlines dictated by vendors. They follow the convergence of budget, compatibility, and operational capacity within each individual organization. Microsoft has known this since Windows XP. HP has it measured in its installed base. And the corporate PC market will continue to revolve around that real-world rhythm for at least the next two years, regardless of when the official calendar says it should have already been over.
