Thinking About the Surface Release Cadence (Premium)

The rate at which Microsoft improves its various products is interesting, and all over the map. And when I look at Surface, specifically, I see the firm doing the exact opposite of what it is doing with Windows. Which is to say that it is releasing product revisions too slowly and too haphazardly.

I think this should be fixed. But more important, I think it can easily be fixed.

Let’s forget, for a moment, that Surface has one foot stuck in the past, thanks to Microsoft’s decision to stick with the USB-based Surface Connect architecture. Let’s forget, too, that Surface still suffers from being almost a boutique PC business, in that Microsoft simply doesn’t offer the same range of product customization choices that are far more typical at bigger PC makers like HP, Lenovo, Dell, and others.

(There’s an argument to be made that Microsoft needs for Surface to achieve at least the same type of model customization that Apple does. But again, I’m trying to stay on target here.)

The reason I want to ignore some of those real-world concerns is that I feel they will solve themselves over time. That Microsoft will modernize and expand its capabilities, and that we simply don’t have to worry about either of those things. They will happen.

No, the bigger issue—and this is one that, frankly, should be an obvious issue to anyone at Microsoft, given the firm’s enterprise bent—is that Surface product revisions are not predictable. They don’t happen on a consistent cadence. They seem, frankly, to be somewhat haphazard. Non-agile, if you will.

Consider the 18-month gap that separated the release of Surface Pro 4 from the release of its successor, Surface Pro (2017). And the 24-month gap that separated the releases of Surface Book and Surface Book 2. Surface Pro 4 and the original Surface Book were announced and released on the same days, and they suffered from the same endemic reliability issues for at least the first nine months or so of their respective life cycles.

Why—why on earth—did Microsoft not issue revisions of these products based on Intel’s Kaby Lake architecture as soon as it became available? Other PC makers do this all the time, and most do so without physically changing a PC’s form factor or design in any meaningful way, if at all.

This type of update should be built into a product’s life cycle. And while we might debate the difficulty and benefits of various schedules, I think that tying yourself to Intel’s release cycles makes plenty of sense for a PC maker. That is, if you can release a Surface device based on Skylake in 2015, you can pretty easily update it to Kaby Lake in 2016. There are huge benefits to doing so. And in the case of Surface Pro 4 and Surface Book, specifically, excellent business reasons to make that change as well.

But here we are in late 2017. Intel has released its 8th-generation Core processors, and has dramatically increased preformance by moving to quad-core processors across the board. Yes, Microsoft is on-board with its new Surface Book 2, but this product is arguably exactly the type of mid-stream update I’m talking about: The same body with some internal differences. Plus, a USB-C port.

But what about Surface Laptop and Surface Pro? These products would be dramatically better, right now, if Microsoft simply updated them with quad-core 8th generation Intel processors across the board. No external changes are necessary, though it wouldn’t be hard to mail in some new Type Cover and, for Surface Laptop, body point colors, to spice things up for the new model year.

But Paul, you argue. Microsoft just released those products.

Yes, I know. The thing is, Microsoft knew that the 8th generation Intel processors were coming when it released those products, and it likewise knew that it was months behind, in the case of Surface Pro, in releasing an update. And it made the choice it made.

And there is precedent for being agile. Apple in late 2016 released its most widely-panned MacBook Pro models ever, so it revved those devices in mid-2017 to address the complaints. It was just six months later.

Today, Microsoft is bringing dual-core weapons to a quad-core fight. It is artificially limiting its ability to compete. That doesn’t make sense.

CES is right around the corner, so Microsoft does have a venue to announce something it’s never once done with Surface before. Which is to basically keep the existing product in market but replace the guts with a newer Intel chipset. In other words, these new products wouldn’t be Surface Laptop 2 and Surface Pro (2018), they’d just be Surface Laptop and Surface Pro.

Here’s another way to look at this plan. On the most recent episode of Windows Weekly, Mary Jo, Leo, and I were discussing the difficulty of introducing major new features in every version of Windows 10, given the maturity of the product and the twice-yearly release schedule. So Mary Jo hit on something that I think is a great idea: Maybe Microsoft should return to the “R2” release schedule for Windows, where every other release is a major version and the interim releases are minor, or R2 (for “release 2”) versions.

That’s essentially what I’m suggesting for each Surface product model, though there may be further revisions (R2, R3, whatever), based on whatever Intel is doing with their chipsets and/or whatever major revisions Microsoft might want to make with its hardware otherwise.

As Brad pointed out when Microsoft announced Surface Book 2, the company is naming things wrong. Surface Book 2 is an arbitrary name, as is Surface Pro (2017), which should simply has been called Surface Pro 5 given what it is. But there’s consistency and predictability to be had. And Microsoft’s customer based pretty much demands it.

I’m surprised that they aren’t embracing this model. It’s the right way forward for the future.