Do you like roller coaster rides? Do you crave those steep drops where the car plunges more than a thousand feet at an insane angle, subjecting your body to a g-force of 3.5? Do you squirm with anticipation as the car clack-clack-clacks slowly up to the top and then hurtles down, leaving your stomach behind as you speed to the bottom?
Buckle up as we go for a decade-long ride with Microsoft's Surface division, which has delivered a performance worthy of a Gerstlauer Euro-Fighter 1000. After an inauspicious beginning in 2012, the division began growing steadily. Back in 2018, I wrote a recap of that progress, noting that the division had brought in total revenue of nearly $5 billion.
Many companies would have given up at that point, but not one run by Steve Ballmer, who famously described Microsoft's approach as "long-term, tenacious, and partner-centric."
As he once told an arena full of partners, "We don't go home. We just keep coming and coming and coming. Tenacious, tenacious, tenacious." (As Ashlee Vance noted when transcribing those remarks for the New York Times, "The man likes to talk in threes.")
In fact, while PC shipments overall have been flat or down for the past four years, the Surface business has been growing at a compound annual rate of better than 22 percent a year.
The Surface division hit its high point at the end of FY22, bringing in nearly $7 billion of revenue for Microsoft. And then… Well, see for yourself.
That's a pretty spectacular collapse in FY2023. (Note that the number for FY2024 in the chart above is a projection.)
Teasing those numbers out of Microsoft's financial reports was a bit of a challenge. The company stopped reporting actual Surface revenue numbers in FY22, instead choosing to disclose changes from the previous year's revenue in percentage terms. To make comparisons even murkier, the company in 2023 changed the "Surface revenue" line item to "Devices revenue growth," folding in revenue it receives from sales of keyboards, mice, and other peripherals.
Even with all that massaging, the picture for FY2023 (which ended on July 30, 2023) was grim. In its annual report, Microsoft noted "Devices revenue decreased $1.8 billion or 24% as elevated channel inventory levels continued to drive additional weakness beyond declining PC demand."
Or, to put it another way: The PC market is going through a major correction right now, and Surface is doing even worse than its rivals.
The carnage isn't over. In the Microsoft FY23 fourth quarter earnings conference call, Satya Nadella and his team projected that Devices revenue for the quarter ending September 30, 2023 would continue plunging:
In Devices, revenue should decline in the mid-30s due to the overall PC market and adjustments we made in our portfolio with an increased focus on our higher margin premium products.
It's hard to imagine that things will get much better before the end of 2023. The most successful Surface products are the Surface Pro and Surface Laptop, both of which are more than a year old. The Surface Laptop Studio 2, a high-priced machine made for a niche market, was in the spotlight at this year's fall Surface event in New York City and isn't likely to be a runaway hit.
In the chart above, I've projected what the Surface revenue number might look like if the company rebounds to a loss of only 24%, repeating this year's performance. Basically, they're back to what the business was in 2016 and 2017, which is… not good?
At Microsoft, a division needs to be able to bring in $10 billion of revenue per year to be considered a "needle mover." Surface once looked like it was on its way to building that kind of steady, growing business. It doesn't anymore.
Maybe this explains the sudden departure of Windows & Devices boss Panos Panay, just days ahead of the company's fall Surface showcase event. His portfolio had expanded dramatically in recent years to cover not just Surface devices but also Windows 11. A report in Business Insider, quoting "anonymous insiders," says Panay was "unhappy with recent changes in the Windows + Devices division," including "significant cuts to simplify the Surface business ... and focus more on Microsoft's hits rather than the more experimental devices the company funded in flush times."
The Surface Pro and Surface Laptop lines are undoubtedly the "hits" referred to in that report. Notably, both products are relatively mature designs whose physical appearance has barely changed in recent years. The addition of Thunderbolt 4 support in the Surface Pro 8 and Surface Laptop 5 was welcome, but otherwise, most upgrades in these product lines have been simple spec bumps, not the sort of thing to win over new customers.
A pair of high-profile flops probably didn't help anyone making the case for Surface in front of Microsoft management. The Surface Duo, a dual-screen phone powered by Android, was uncomfortably expensive and just plain odd. I bought one shortly after its release in 2019 and returned it six weeks later. (For details, see "Five reasons why I returned my Surface Duo.")
And then there was the dual-screen Surface Neo, which was announced at the same 2019 event as the Surface Duo and became one of the first casualties of the pandemic the next year. It never shipped, and neither did the operating system it was supposed to run, Windows 10X. If you've forgotten about that project it's understandable. Hey, a lot happened in 2020! But the bottom line is it was supposed to be a "cloud-powered" competitor to ChromeOS.
Windows 10X is debuting as the native operating system for a new PC form factor, with multiple screens. But its core innovation is the ability to run traditional Windows desktop apps in secure containers that are isolated from the core of the operating system. It's easy to imagine this technology migrating to more traditional form factors before long.
A dual-screen PC is an expensive gimmick and unlikely to become a mass market hit. That's why the project was killed and why it's unlikely to ever come back. But imagine if you can accomplish many of the same goals of the Windows 10X project on traditional hardware, like (say) an inexpensive laptop. At that point you have a credible competitor to Google's Chromebook, and possibly a way to once again make Windows relevant in schools.
That is apparently what Microsoft has been working on for the past three years with its CorePC project, which might ship next year (in whole or in part) under the Windows 12 moniker. If you need an explainer, Zac Bowden at Windows Central can fill you in on this "modular and customizable variant of Windows."
My sources tell me CorePC will allow Microsoft to finally deliver a version of Windows that truly competes with Chromebooks in OS footprint, performance, and capabilities. A version of Windows that only runs Edge, web apps, Android apps (via Project Latte) and Office apps, designed for low-end education PCs is already in early testing internally, and is roughly 60-75% smaller than Windows 11 SE.
Meanwhile, at the "high-margin, premium device" end of the PC spectrum, Microsoft desperately needs a way to compete with Apple's M2-powered MacBook line (and Macs based on more powerful M3 processors should be on the shelves in Apple Stores next year). Unfortunately, getting performance and battery life that are competitive with Apple's flagship products is going to require Arm-based processors, and right now the entire PC industry is waiting for Qualcomm to ship its next-generation Nuvia SoCs. This is what I wrote on the subject a few months ago:
As pretty much every rational observer has noted, Windows on Arm is far behind Apple, at least using Qualcomm's current SoC designs. But they could play catch-up, and maybe even leapfrog their Cupertino competition, with a successful launch of the Nuvia-based Oryon architecture, especially if they can build some custom Windows features into it. (The SQ3-based Surface Pro 9 has a couple of nifty AI-based features not found in its Intel sibling, including eye-tracking and background noise reduction for video calls.)
That's probably why Microsoft didn't have anything to say about upcoming iterations of Surface Pro or Surface Laptop at its recent hardware event. Another spec bump won't move the needle at all. But an Arm-powered device that gets double the battery life of x86-based hardware, has a neural processing unit for advanced AI tasks, and can run x86 apps in secure containers? That might be worth waiting for.
Fighting off determined competitors on the low end and the high end at the same time is hard enough, and the challenge is worse when you need to do it in the face of budget cuts and the loss of a key leader. Get ready for another ride on the Surface roller coaster.