On my flight today, I rewatched the documentary Objectified. I’ve seen it a few times before, but it has been several years. While I don’t jibe with 100% of the sentiment of the documentary, it made me think a bit about design, as I was headed to Dallas. In particular, it made me consider Apple, Microsoft, and Google, and their dramatically different approaches to design – which are in fact a reflection of the end goal of each of the companies.

One of my favorite moments in the piece is Jony Ive’s section, early on. I’ve mentioned this one before. If you haven’t read that earlier blog post, you might want to before you read on.

Let’s pause for a moment and consider Apple, Microsoft, and Google. What does each make?

• Apple – Makes hardware.
• Microsoft – Makes software.
• Google – Makes information from data.

Where does each one make the brunt of its money?

• Apple – Consumer hardware and content.
• Microsoft – Enterprise software licensing.
• Google – Advertising.

What does each one want more of from the user?

• Apple – Buy more of their devices and more content.
• Microsoft – Use their software, everywhere.
• Google – Share more of your information.

You can also argue that Apple makes software, Microsoft makes hardware, and Google makes both. Some of you will surely do so. But at the end of the day, software is a hobby for Apple to sell more hardware and content (witness the price of their OS and productivity apps), hardware is a hobby for Microsoft to try and sell more software and content, and hardware and software are both hobbies for Google to try and get you more firmly entrenched into their data ecosystem.

Some people were apparently quite sad that Apple didn’t introduce a ~12” so-called “iPad Pro” at their recent October event. People expecting such a device were hoping for a removable keyboard, perhaps like Microsoft’s Surface (ARM) and Surface Pro (Intel) devices. Hopes were there that such a device would be the best of both worlds… a large professional-grade tablet (because those are selling well), and a laptop of sorts, and it would feature side-by side application windows, as have been available on Windows nearly forever, and many Android devices for some time. In many senses, it would be Apple’s own version of the Surface Pro 3 with Windows 8.1 on it. Reporters have insisted, and keep insisting that Apple’s future will be based upon making a Surface clone of sorts. I’m not so sure.

I have a request for you. Either to yourself, in the comments below, or on Twitter, consider the following. When was the last time (since the era of Steve Jobs return) that you saw Apple hardware lean away, in order to let the software compromise it? Certainly, the hardware may defer to the software, as Ive says earlier about the screen and touch on the iPhone; but the role of the hardware is omnipresent – even if you don’t notice it.

I’ve often wondered what Microsoft’s tablets would look like today if Microsoft didn’t own Office as well as Windows; if they weren’t so interested in preserving the role of both at the same time. Could the device have been a pure tablet that deferred to touch, and didn’t try so hard to be a laptop? Could it have done better in such a scenario?

Much has been said about the “lapability” of the Surface family of devices. I really couldn’t disagree more.

More than one person I know has used either a cardboard platform or other… <ahem> surface as a flattop for their Surface to rest upon while sitting on their lap. I’ve seen innumerable reporters contort themselves while sitting in chairs at conferences to balance the device between the ultra-thin keyboards and the kickstand. A colleague recently stopped using his Surface Pro 2 because he was tired of the posture required to use the device while it is on your lap. It may be an acceptable tablet, especially in Surface Pro 3 guise – but I don’t agree that it’s a very good “laptop”.

The younger people that follow me on Twitter or read this blog may not get all of these examples, but hopefully will get several. Consider all of the following devices (that actually existed).

• TV/VCR combination
• TV/DVD combination
• Stand mixers with pasta-making attachments
• Smart televisions
• Swiss Army Knife

Each of these devices has something in common. Absent a better name to apply to it, I will call that property toasterfridgality. Sure. Toasterfridge was a slam that Tim Cook came up with to describe Microsoft’s Surface devices. But regardless of the semi-derogatory term, the point is, I believe, valid.

Each of the devices above compromises the integrity with which it performs one or more roles in order to try and perform two or more roles. The same is true of Microsoft’s Surface and Surface Pro line.

For Microsoft, it was imperative that the Surface and Surface Pro devices, while tablets first and foremost (witness the fact that they are sold sans keyboard), be able to run Office and the rest of Win32 that couldn’t be ported in time for Windows 8 – even if it meant a sacrifice of software usability in order to do so. Microsoft’s fixation on selling the devices not as tablets but as laptop replacements (even though they come with no keyboard) leads to a real incongruity. There’s the device Microsoft made, the device consumers want, and the way Microsoft is trying to sell it. Even taking price out of the equation, is there any wonder that Surface sales struggled until Surface Pro 3?

Lenovo more harmoniously balances their toasterfridgality. Their design always seems to focus first on the device being a laptop – then how to incorporate touch. (And on some models, “tabletude”.) Take for example, the Lenovo ThinkPad Yoga  or Lenovo ThinkPad Helix. These devices are laptops, with a comprehensive hinge that enables them to have some role as a tablet while not completely sacrificing… well… lapability. In short, the focus is on the hinge, not on the keyboard.

To view the other end of the toasterfridge spectrum, check out the Asus Padfone X, device that tries to be your tablet by glomming on a smartphone. I’m a pretty strong believer that the idea of “cartridge” style computing isn’t the future, as I’ve also said before. Building devices that integrate with each other to transmogrify into a new role sounds good. But it’s horrible. It results in a device that performs two or more roles, but isn’t particularly good at either one. It’s a DVD/VCR combo all over again. Phone breaks, and now you don’t have either device anymore. If there was such a model that converted your phone into a desktop, one can only imagine how awesome it would be reporting to work on Monday, having lost your “work brain” by dropping your phone into the river.

I invite you to reconsider the task I asked of you earlier, to tell me where Apple’s hardware defers to the software. Admittedly, One can make the case that Apple is constantly deferring the software to the hardware; just try and find an actual fan of iTunes or the Podcasts app, or witness Apple’s recent software quality issues (a problem not unique to Apple). But software itself isn’t their highest priority; it’s the marriage of that software and the hardware (sometimes compromising them both a bit). Look at the iPhone 6 Plus and the iPad Air 2. Look how Apple moved – or completely removed – switchgear on them to align with both use cases (big phones are held differently) and evolving priorities (switches break, and the role of the side-switch in iOS devices is now completely made redundant by software).

Sidebar: Many people, including me, have complained that iOS devices start at 16GB of storage now. This is ridiculous. With the bloat of iOS, requirements for upgrading, and any sort of content acquisition by their users, these devices will be junk before the end of CY2016. Apple, of course, has made cohesive design, not upgradability, paramount in their iOS devices. This has earned them plenty of low scores for reparability and consumer serviceability/upgradeability in reviews. I think it is irresponsible of Apple, given that they have no upgradeability story, to sell these devices with 16GB. The minimum on any new iOS device should be 32GB. Upgradability or the ability to add peripherals is often touted by those dissing Apple as limitations of the platform. It’s true. They are limitations. But these limitations and a tight, cohesive hardware design, are what let these devices have value 4 years after you buy them. I recently got $100 credit from AT&T for my daughter’s iPhone 4 (from June, 2010). A device that I had used for two years, she had used for two more, and it still worked. It was just gasping for air under the weight of iOS 6, let alone iOS 7 (and the iPhone 4 can’t run 8). There is a reason why these devices aren’t upgradeable. Adding upgradeability means building the device with serviceability in mind, and compromising the integrity of the whole device just to make it expandable. I have no issue with Apple making devices user non-serviceable for their lifespan, as I believe it tends to result in devices that actually last longer rather than falling apart when screws unwind and battery or memory doors stop staying seated.

I’ve had several friends mention a few recent tablets and the fact that USB ports on the devices are very prone to failure. This isn’t new to me. In 2002, when I was working to make Windows boot from USB, I had a Motion Computing M1200 tablet. Due to constant insertion and removal of UFDs for testing and creation, both of the USB ports on the tablet had come unseated off of the motherboard and were useless. Motion wanted over $700 to repair a year old (admittedly somewhat abused) tablet. With <ahem> persuasion from an executive at Microsoft, Motion agreed to repair it for me for free. But this forever highlighted to me that more ports aren’t necessarily always something to be looked at in a positive light. The more things you add, the more complex the design becomes, and the more likely it becomes that one of these overwrought features added to please a product manager who has a list of competitive boxes to check will lead to a disappointed customer, product support issues and costs, or both. USB was never originally designed to have plugs inserted and removed willy-nilly (as Lightning and the now dead Apple 30-pin connector were), and I don’t think most boards are manufactured to have devices inserted and removed as often (and perhaps as haphazardly) as they are on modern PC tablets.

Every day, we use things made of components. These aren’t experiences, and they aren’t really even designed (at least not with any kind of cohesive aesthetic). Consider the last time you used a Windows-based ATM or point-of-sale/point-of-service device. It may not seem fair that I’m  glomming Windows into this, but Windows XP Embedded helped democratize embedded devices, and allowed for cheap devices to handle cash, digital currency, rent DVDs on demand, and make a heretofore unimaginable self-service soda fountain.

But there’s a distinct feel of toaster fridge every time I used one of these devices. You feel the sharp edges where the subcomponents it is made of come together (but don’t align). Where the designer compromised the design of the whole in order to accommodate the needs of the subcomponents.

The least favorite device I use with any regularity is the Windows-based ATM at my credit union. It has all of the following components:

• A display screen (which at least supports touch)
• An input slot for your ATM/credit/debit card
• A numeric keypad
• An input slot for one or more checks or cash
• An output slot for cash
• An output slot for receipts.

As you use this device, there are a handful of pain points that will start to drive you crazy if you actually consider the way you use it. When I say left or right, I mean in relation to the display.

• The input slot for your card is on the right side.
• The input slot for checks is on the left side.
• The receipt printer is on the right side.
• The output slots for cash are both below.

Arguably, there is no need for a keypad given that there is a touchscreen; but users with low visibility would probably disagree with that. Besides that, my credit union has not completely replaced the role of the keypad with the touchscreen. Entering PINs, for example, still requires the keypad.

So to deposit a check, you first put in your card (right), enter your pin (below), specify your transaction type (on-screen), deposit a stack of checks (no envelope, which is nice) on the left. Wait, get your receipt (top right), and get your card (next down on the right). My favorite part is that the ATM starts beeping at you to retrieve your card before it has released it.

This may all seem like a pedantic rant. But my primary point is that every day, we use devices that prioritize the business needs, requirements, or limitations of their creator or assembler, rather than their end user.

Some say that good design begins with the idea of creating experiences rather than products. I am inclined to agree with this ideology, one that I’ve also evangelized before. But to me, the most important role in designing a product is to pick the thing that your product will do best, and do that thing. If it can easily adapt to take on another role without compromising the first role? Then do that too. If adding the new features means compromising the product? Then it is probably time to make an additional product. I must admit – people who clamor for an Apple iPad Pro that would be a bit of (big) tablet and (small) notebook confuse me a bit. I have a 2013 iPad Retina Mini and a 2013 Retina MacBook Pro. Each device serves a specific purpose and does it exceptionally well.

I write for a living. I can never envision doing that just on an iPad, let alone my Mini (or even without the much larger Acer display that my rMBP connects to). In the same vein, I can’t really visualize myself laying down, turning on some music, and reading an eBook on my Mac. Yes. I had to pay twice to get these two different experiences. But if the alternative is getting a device that compromises both experiences just to save a bit of money? I don’t get that.