Microsoft has reported a $900 million write-down on Surface RT inventory in its latest financial results. Was Surface RT a big mistake?
A loss of that size is a massive blunder, but the concept behind Surface RT is good and Microsoft should persevere. Here’s why.
Surface RT is experimental in two ways:
- It was the first Microsoft-branded PC (or tablet if you prefer).
- It was among the first Windows RT devices. Running on the ARM processor, Windows RT is locked down so that you can only install new-style Windows 8 apps, not desktop apps. However, the desktop is still there, and Microsoft bundles Office, a desktop application suite.
Microsoft had (and has) good reason to do both of these things.
Historically, DOS and Windows prospered because it was open to any hardware manufacturer to build machines running Microsoft’s operating system, creating a virtuous circle in which competition drove down prices, and abundance created widespread application support.
This ecosystem is now dysfunctional. The experience of using Windows was damaged by OEM vendors churning out indifferent hardware bundled with intrusive trial applications. It is still happening, and when I have to set up a new Windows laptop it takes hours to remove unwanted software.
Unfortunately this cycle is hard to break, because OEM vendors have to compete on price, and consumers are seemingly poor at discriminating based on overall quality; too often they look for the best specification they can get for their money.
Further, Windows remains a well understood and popular target for malware. One of the reasons is that despite huge efforts from Microsoft with User Account Control (the technology behind “do you really want to do this” prompts in Windows Vista onwards), most users outside the enterprise still tend to run with full administrative rights for their local machine.
Apple exploited these weaknesses with Mac hardware that is much more expensive (and profitable), but which delivers a less frustrating user experience.
Apple has been steadily increasing its market share at the high end, but an even bigger threat to Windows comes from below. Locked-down tablets, specifically the Apple iPad and later Android tablets, also fixed the user experience but at a relatively low price. Operating systems designed for touch control means that keyboard and mouse are no longer necessary, making them more elegant portable devices, and a wireless keyboard can easily be brought into use when needed.
Microsoft understood these trends, although late in the day. With Surface it began to manufacture its own hardware, an initiative which alongside the bricks-and-mortar Microsoft Stores (supplying trialware-free Windows PCs) aims to counter the corrosive race to the bottom among OEM vendors.
Windows 8 also introduces a new application model which is touch-friendly, secure, and offers easy app deployment via the app store.
In Windows RT the experiment is taken further, by locking down the operating system so that only these new-style apps can be installed.
Surface RT brings both these things together, solving many of the problems of Windows in a single package.
Why Surface RT failed
Surface RT is well made, though performance is disappointing; it seems that Nvidia’s Tegra 3 chipset is not quite sufficient to run Windows and Office briskly, though it is usable, and graphics performance not bad.
There were several problems though.
- The price was high, especially when combined with the clever keyboard cover.
- It may solve the problems of Windows, but for many users it also lacks the benefits of Windows. They cannot run their applications, and all too often their printers will not print and other devices lack drivers.
- Surface RT launched when the Windows 8 app store was new. The new app ecosystem also has its problems (all these things are inter-related) and in consequence few compelling apps were available.
- Microsoft’s built-in apps were poor to indifferent, and Office was bundled without Outlook.
I was in New York for the launch of Surface RT. There were “Click In” ads everywhere and it was obvious that Microsoft had convinced itself that it could sell the device in large numbers immediately. That was a fantasy. I suppose that if consumers had taken Windows 8 to heart quickly (as opposed to resisting the changes it imposes) and if the app ecosystem had flourished quickly then it could have taken off but neither was likely.
Surface RT positives
Despite all the above, Surface RT is not a bad device. Personally I was immediately drawn to its slim size, long battery life, and high build quality. The keyboard cover design is superb, though not everyone gets on with the “touch” cover. I purchased one of the launch machines and still use it regularly for cranking out Word documents on the road.
Reviews on Amazon’s UK site are largely positive:
Surface RT is also improving as the software evolves. Windows 8.1, now in preview, adds Outlook and makes the device significantly more useful for Exchange users. Performance also gets a slight lift. The built-in apps are improving and app availability in general is much better than it was at launch, though still tiny compared to iPad or Android.
I have also been trying Surface Pro since receiving one at Microsoft’s Build conference last month. The Pro device has great performance and runs everything, but it is too bulky and heavy to be a satisfying tablet, and battery life is poor. I think of it more as a laptop, whereas Surface RT is a true tablet with a battery that gives pretty much a full day’s use when out and about.
Microsoft’s biggest mistake with Surface RT was not the concept, nor the quality of the device. Rather, they manufactured far too many thanks to unrealistic expectations of the size of the initial market. The sane approach would have been a limited release with the aim of improving and refining it.
I hope Microsoft perseveres both with Windows RT and with Surface RT. Give it better performance with something like Nvidia, Tegra 4, Windows 8.1, and improved app support, and it is near-perfect.
The future of Windows
Desktop Windows will remain forever, but its decline is inevitable. Even if it fails, we should recognise that Microsoft is trying to fix long-standing and deep-rooted problems with Windows through its Windows 8, Surface and Windows RT initiatives, and there is some sanity in the solutions it has devised. Despite a billion dollars thrown away on excess Surface RT inventory, it should follow through rather than abandon its strategy.
The Surface RT’s biggest problem by far is an inferior app store selection on a locked-down platform. This is more important than anything else, and could be overcome only if Microsoft were willing to undercut cheap Android tablets as pure web browsing platforms, but that’s unlikely. As long as Apple keeps making tablets with equivalent hardware, better usability, and a much bigger ecosystem, a pure Windows Store tablet doesn’t stand a chance. Its only niche would be enterprise interoperability — forget the consumer market. It would be the BlackBerry of tablets, basically.
Given how many people worldwide I’ve seen complaining about how long it took to get Surface available in their markets, a limited release would have led to more frustration. Actually marketing and explaining RT a lot more, beyond the trendy design stylings of Click In and lots of TV ads for picture password would have helped; as would the physical stores where people buy PCs being less awful at the process… RT needs to improve but it needs to be presented properly for the strengths it has now.
Following copied from Twitter (Regarding this blog article by Tim):
Steven Sinofsky @stevesi
“@edbott: The best analysis I’ve seen so far of the Windows RT/Surface RT: http://www.itwriting.com/blog/7530-windows-rt-and-surface-rt-why-microsoft-should-persevere.html …” // some interesting thoughts
Tim, hard to argue with what you’ve written, even the conclusion.
I’m trying to make sense of the writedown amount ($900m). If that reflects only the $499 to $349 price cut, then it would seem that there must be about 6 million unsold Surface RTs. Is that a big number to sell at $349? Hard to say, since we don’t know how many have been sold to date or how many a tablet maker other that Apple can reasonably be expected to sell in a quarter or two.
Two scenarios come to mind:
(1) There are fewer than 6m unsold Surface RTs. That would suggest MS anticipates further price cut(s) to unload the inventory.
(2) There are 6m unsold Surface RTs but MS is unable to sell them at $349. That would mean another big writedown in some future quarter.
This writedown also undercuts the argument that pricing them lower to start would have been a good idea. Remember, the last couple quarters’ results reflect Surface RT sales at $499 and inventory valued at $499 – adjust the quarters accordingly for a $349 hypothesis and Entertainment and Devices Division will be even more in the red.
“Blunder” and “fantasy” are right. It’s hard not to be reminded of the fate of HP and RIM’s tablet excursions.
The simple fact is that very few People want those pads. And I sure understand. The Interface is obtrusive, ugly, stressing neon-colours – and – you cant find out how they work (even called Windows).
It is not at all a surprice that MS is having it largest failure ever in this. The CEO and president in charge of designing this – should have been lesser psychopatic – and letting the users having the Choice to behold a familiar (and more elegant designed) GUI.
Phil, revenues from Surface RT are credited to the Windows and Business divisions — *NOT* to the E&D Division.
I agree that the default assumption of 6 million unsold Surface RTs does not make sense. Inventory is carried on the books at cost — not at retail price. Thus, a $150 price reduction should not have resulted in a $150 writeoff.
Some additional possibilities:
(3) Microsoft has 2.25 million unsold Surface RTs. They simply declared the entire stock to be obsolete inventory, and wrote off the full $400 cost of each Surface RT. (Cost includes $300 manufacturing, plus $100 for OS, Office RT, and keyboard.)
(4) Microsoft has 9 million unsold Surface RTs. The cost of each Surface RT is $400, and this was written down to $300 on average. (Assume 50/50 split between Microsoft Store and other retailers, and $250 wholesale cost.)
Bottom line — we have absolutely no idea how the accounting was done. You could make an argument for just about any number in the single-digit millions.
@Tom – very good. Unfortunately, we don’t know cost; we only know price, price cut and writedown. However, I think you’ve established a good range of possible inventory numbers.
As a regular user of the Surface RT I call BS on the Tegra 3 performance call! (But I do believe MS paid a helluva lot of money for those chips affecting the price, more below).
Surface RT + Windows RT is a rushed product in all respects:
1. Windows port to ARM incomplete (too much baggage from x86). Needs better battery life/more stability.
2. WinRT (development stack) is incomplete. Just look at 8.1 Preview and you’ll know why!
3. The Tablet experience is incomplete because of the above two (for example lack of synchrony between SkyDrive Store and ‘Desktop’), horribly incomplete Music/Video apps (don’t blame the app ecosystem), if MS wants a media tablet a hobbled Media/Music player out of the box is a disaster, ditto Mail App.
4. Pricing: Been yelling from the top of the roof ever since it got released, @ $499 it needed the Type Keyboard with it. I am assuming MS’ lack of tablet experience probably prevented a better deal at the supply chain level resulting in higher production cost. If they were being plain greed, well… $900MM is the course correction they needed!!! Current $359 should be accompanied with a $49 Type Keyboard. Junk the touch keyboard!
The only thing it had going for itself was Office and the ‘familiarity of Windows’ (which is a very big deal to a lot of people actually).
Good analysis of the causes of the Surface RT write down, but to understand the true internal causes you need to look at the current state of Microsoft’s culture and motivation which has been corrupted by the twin evils of the team collaboration sapping stack ranking review system and the crazy focus on manipulatable scorecards metrics. Microsoft used to be full of staff who believed in what the company was trying to do. Those people have been replaced with others who are happy to collect the pay check for a couple of years and get Microsoft on the CV. The zeal to win has been sucked out of the company.
Very interesting analysis Tim. Do you think Microsoft is still regarded as the safe choice by early adopters? They’ve had several hardware missteps, and every time any giants like HP, Blackberry, Barnes and Noble fail in becoming the “third estate” of tablet computing, they increase the risk aversion of people anything that is less established for all the other players.
Microsoft has never been interested in anything that is worth less a billion dollar business. If they don’t persevere, people won’t stick around in the future. Is Microsoft willing to go Bing’s length with the Surface? If it does, and stays in the game long enough, there’s a chance they will still be players in the future.
People moan about Apple but going through the MS/Symantec authorisation just to become a WM8 app developer is completely ludicrous. In fact its so bad we have given up. They don’t understand the whole ‘startup’ thing at all. Its way simpler with Apple app store and Google play.
The key is to make your platform accessable to devs. Make it hard and people won’t join.
Gary