Monday morning quarterbacking: Surface RT Debacle Edition

Hal Berenson regularly shares some of the greatest insights into Microsoft that I see appear around the web. His latest post, Fixation on Margins: The Surface RT Debacle Edition, does a great job conveying both the ambition of Surface RT and its failure to achieve its lofty goals. That inspired me to share some thoughts I’ve been having for a while (and occasionally tweeting about), regarding where Microsoft went wrong with Surface. This isn’t an “I told you so,” (because I didn’t, and it wasn’t my job to). It’s not even a “what I would have done,” though maybe a “what I’d like to think I would have done.” 🙂

In Hal’s post, I particularly like his invocation of Maslow’s hierarchy of needs in describing the value of Office on a tablet. I agree 100% that Office is a highly valued, worthwhile differentiator for tablets. That is, I am certain that a very sizable percentage of iPad and Android tablet users would consider Office a desirable addition to their tablet experience. What they won’t say is that Office is so worth having that they would sacrifice other things they love about those tablets. In particular, the apps they cherish most.

I don’t think any of this is news to the folks behind the Surface and Windows RT efforts. In fact, achieving that first level of “Maslow’s tablet hierarchy” (or maybe that should be Berenson’s tablet hierarchy?) was of the utmost priority for Microsoft. This was entirely the purview of the Windows team (and by way of partnership, DevDiv). Just some of the efforts they/we undertook to achieve that are obvious:

• One store for all Windows 8 (and RT) devices, which enabled all the developer marketing around reaching “hundreds of millions of users” within the first year basically by default.
• The better-than-them store deal, where developers retain 80% of their sales rather than 70%, after exceeding the $25,000 threshold.
• A huge investment in tools for Windows 8 apps, along with the multi-pronged platform approach designed to appeal to three groups:
  • Web developers, with the first-class JS/HTML5 platform, new JS library, and kick ass tools.
  • XAML / .NET developers, with the WinRT XAML platform and familiar tools.
  • Native developers, by finally (finally) providing them with a modern UI framework (XAML), investing in C++ 11, WRL library, CX extensions, and long overdue tooling updates for native C++ developers. Oh, and continuing to invest in DirectX.
• Restricting many of the new platform investments to Metro applications (though I’m sure other factors around time/resources contributed to this).
• Booting to the start screen, setting “Metro” apps as the default file association handlers on all PC/device form factors, and generally treating the desktop as the “classic” environment, much in the way that Windows 95 treated DOS apps.

That’s not to say that this was the only motivation for some of those decisions (particularly the latter ones). I can attest to the fact that those in charge of those later UX decisions were “true believers” who felt each of those decisions was not just supported, but often mandated, by the Metro-inspired design principles they’d adopted at the outset. You can call it over-confidence, or even arrogance, or an over-correction at the sight of Apple’s success with what seemed to be “the Jobs/Ive way or the highway.” But from my perspective I saw no reason to blame those sort of decisions on out-of-touch executives or inherent organizational flaws, and certainly not on incompetence by the front-line engineers or designers.

What I might attribute to organizational or leadership failings is the lack of cross-group coordination, particularly in the marketing and basic product development areas. A quick note on my perspective for those who don’t know me: I was a senior IC developer on the Windows 8 User Experience team, who left the company earlier this year for a variety of reasons, but mostly to scratch an entrepreneurial itch that my 8 years at Microsoft had left unscratched. So you could say that as an engineer it’s easier for me to criticize these “farther away” disciplines. You might be right.

The prime example for me, though, is the fact that you’ve probably heard of Windows RT. From my perspective, that name was chosen because it didn’t matter. People don’t buy tablets based on operating systems, the thinking went. They buy tablets. That is, you don’t buy an iOS device, you buy an iPad. You don’t buy an Android tablet, you buy a Galaxy or a Nexus.

You can see how this could have extended to the Surface. I wasn’t in the marketing or product marketing part of the organization, but my sneaking suspicion is that on the Windows side, this is how the naming of Windows RT came about. Surface would be the brand (or in the case of Samsung it’d be something like Ativ, Yoga for Lenovo, Vivo for Asus, etc). It made sense to me, anyway. I never expected to have to explain Windows RT to anyone. I was, however, quite prepared to say “Surface runs new Windows 8 apps and comes with the new version of Office, but isn’t backward compatible with older versions of Windows.” I figured normal humans would grasp that concept. The last couple decades have taught people the “backward compatible” concept. If they didn’t get that, you could just make Ballmer cringe and say “it’s a tablet, not a full computer.” Wrong or right, I bet a lot of people would understand what you meant. So yeah, simple message: Windows 8 is backward compatible, but Surface is not. It’s Microsoft’s iPad. It’s super well made, priced competitively, and runs tablet apps. But it also runs Office, has a kickstand, and you can get these cool keyboard covers. Awesome, right?

Sure, it still had the apps thing to overcome. But everyone has to overcome that. Apple did. Yes, Apple had phone apps on the iPad at launch, but those were next to useless and honestly such a poor experience (particularly for non-games) that I’m still shocked Apple actually did it. Anyway, there were a lot of reasons Microsoft and the Windows team thought this would be fine. The aforementioned message to developers was expected to be super compelling. The fact that tablet users spend the vast majority of their time in the web browser was another (and I still think Metro IE on Win8 is the best tablet browsing experience bar none). The plethora of Microsoft-made apps was also expected to play a big part in satiating early adopters until the ecosystem began to thrive on its own.

And so, the story of Surface made sense to me. Then the announcement came, and I learned of the Surface Pro. Cool! Right? But wait, now the message got complicated. “Surface is the cool tablet that runs new stuff and Office, but isn’t backward compatible. Surface Pro runs old stuff, but it doesn’t come with Office. It’s more expensive, but it’s not as good at tablet basics like being small, always-on, and having a long-lasting battery.” Phew. That’s a mouthful, and probably takes a few repetitions to sink in.

No no, it isn’t that bad, I thought. The Pro is going to be more of a niche thing, and not even there at the start. It’s designed for people who were wiling to pay more and compromise their tablet for something fast, higher resolution, with an active digitizer pen. So maybe I wouldn’t have to tell most people about it. Even if they saw the Pro at the store, they’d see the price tag and the added heft and those outside the target niche would move along. My earlier “Surface is a tablet with Office but isn’t backward compatible” explanation would still suffice for the general case. Or so I hoped.

You know, odd as it may sound, I initially thought that compatibility with “classic” Windows apps on the Pro was an extra bonus. A footnote, more than a selling point. Unless you’re a developer, maybe. Or part of an even smaller niche who thought it’d be the perfect device if it just ran Photoshop. In retrospect, I can’t help but wonder if this “feature” may have won over a tree at the expense of the forest… But I’ll get to that in a moment 🙂

Compatibility differences aside, the other big complicating factor the Pro introduced was the “Oh, the expensive one doesn’t include Office but the cheap one does” conundrum. I’ve heard several Best Buy associates try valiantly to explain this, and every time I can’t help but feel sorry for them. It reminds me of trying to explain the Windows Vista SKU model. It’s one of those things that can only make sense from the perspective of a product development person assuming that it’s someone else’s job to make sense of it (i.e., “marketing will explain it to consumers”). Sadly, most of those explanations ended with a bewildered customer walking away from the Surface displays, presumably to find something that made their brain hurt less.

So with all that in mind, and the substantial benefit of hindsight, here are ways I think the messaging could have been kept simple:

• By not releasing the Surface Pro at all, at least the first year when establishing the brand and expectations about it.
• Make it not available at retail but instead just a development device for Surface. “Surface Developer Edition” or something. Would’ve sold out Build faster I bet 🙂
• Just plain not use Surface in the name of the Pro, embracing the “laptop first” model it better fits in. Two clear messages instead of one muddled one.
• Or… brace yourself for this one. Are you sitting down? Good.
  Ship the Pro with an Intel version of Windows RT. No classic desktop apps here. Blasphemy? I don’t think so. Maybe make an exception for VS (or have “developer edition” variant), because Surface + fast+ Office + VS = awesome dev device. But otherwise it’s the niche power-tablet. Still a tablet. Still a Surface with the same, simple message. Just made for people who are okay sacrificing $$, battery, and weight, to have tomorrow’s performance today. If people will buy giant 8-core phones, I could believe there’s a solid niche willing to make this sacrifice.

You sell/give some of those “Intel RT”-running Pro devices to developers, with only Office and dev tools on the desktop, and I bet some of them build apps for it just because they want to use them. I’d bet money you’d at least have a Bit Torrent client by now.

While I’m Monday morning quarterbacking the whole thing, I’ll share another idea I had a little while back: Don’t include Office with the Surface. Include it with the cover.

The initial Surface RT price may have seemed ambitious, but the price of the Touch and Type covers are what really hurt. Maybe if the Surface price had been something you could call cheap, then the cover price could be “where they get ya.” In my experience, people are okay with things like that. They just feel smart for recognizing it. But with the Surface RT, you were already paying at least what Apple charged just for the tablet. If you’re paying someone the same thing Apple charges, they’ve already “got” you.

But let’s say they offered the Surface RT at the iPad competitive price, and convinced you that it was a good tablet and that your favorite apps, if not there, would be coming soon. You’re interested, and weighing the pretty tiles and kickstand against the iPad you saw at the Apple store. Then the salesperson says, “Best of all, for just $120 you can have Office and this awesome keyboard cover, so you might not even need your laptop anymore.” Well I don’t know about you, but that almost sounds like a deal!

The other big leadership failing that happened somewhere is almost too obvious to bother mentioning. The lack of proper “Metro” versions of Office is inexcusable. Hell, they could’ve jammed the Office Web Apps into Win8 apps (remember that awesome JS/HTML platform designed specifically so that people could do things like this? Apparently someone important didn’t). Even if they intended to replace them a year or two later, you can’t convince me that they couldn’t have had something compelling there. Then you could’ve had RT sans desktop. Or at least without it pinned by default, left in as “debug/admin mode” the way command prompts have been for ages now.

All of those things said, when I look at the future of Windows and Surface, I don’t see doom and gloom. I don’t think any of the above “things I would’ve done” will happen, or even need to happen. For the most part, we’re talking about ships which have long sailed. Things is, I see this whole “Surface Debacle” as a stumble. Could that lead to them falling flat on their face? Maybe. But if there’s a tech company that knows how to power through a stumble and come out laughing, its name starts with an M and ends with “icrosoft.”