Why Nielsen Is Wrong About Windows 8

Look, I’m a user experience designer for Nokia. I’ve designed my fair share of desktop and phone apps to know a usability issue when I run slap-bang into it. I’ve been following the noize about Windows 8 pretty closely and I have to say, I honestly don’t get it.

Sure, Windows 8 has a learning curve. Sure, it’s frustrating to use when you pick it up for the first time. Will you make mistakes? Absolutely. But, as an industry, how can we possibly progress the operating system paradigm unless we administer a little kick to move out of the local maxima users are in? The fast-becoming-traditional model of the static many multicoloured front doors that need to be walk around to and entered to get anything done on your phone, desktop or tablet is old. Microsoft’s ‘modern UI’ (the new name for the ‘metro’ design language) brings a whole series of innovations to the staid operating system space. More so than golden-boy Apple has in any of it’s recent OS releases. A ¬†headline today stated that OS X 10.9 will include Siri and Maps. Oh, please. More iOS features ham-fisted onto the desktop? That’s not innovation. If people feel self-conscious enough talking to their phones, they’re sure as hell not going to talk to their desktop Macs.

So, an ailing Microsoft realise they’re on the precipice and try to head it off by going in an interesting and new direction. Gone is the Start menu. Core interactions require gestures that aren’t as easy to perform with a keyboard and mouse. Apple-style skeuomorphism is shunned in favour of a clean, clear, crisp chrome-less design language. Content Is King!

Of course users are going to have some issues getting used to it. Initial reactions will be bad. But didn’t we all chide Microsoft with the release of Windows 7 for clinging desperately to their OS legacy and lacking the bravery to try something new?

Windows 8 is not difficult to use. It takes some mastery (perhaps a few days of fumbling). But then you find yourself getting stuff done with surprising speed and efficiency. Instantly my iPad looks old and frustrating.

Jackob Nielsen (very influential grandaddy of empirical user experience design and measurement) has weighed in on the Win8-bashing act, too. Read his views here:


Here’s why he’s wrong:

Double Desktop = Cognitive Overhead and Added Memory Load

Nielsen is criticising Windows 8’s two worlds: the modern UI Start screen (with live tiles, etc.) and the traditional Desktop. Sure, I agree, this does create cognitive load. While we’re in this transitionary period between modern UI apps and old-skool Desktop Windows apps this is an unfortunate situation. However, most of your common tasks can be performed as completely (and, I’m arguing, more efficiently) using modern UI apps. Mail, Internet, Facebook, Skype, Twitter, Photos, etc… the things you spend most of your life swinging between can be done in the modern UI. And when you need an old Desktop app, there’s a tile on the Start screen that launches it directly: you’re taken to the Desktop with your app open and a first-class citizen alongside modern UI apps. You can navigate through the application history (swiping from the left) and see your desktop app in there. A Desktop can be put into snap view. Your modern UI apps can be put into snap view alongside your fill-view Desktop app. It’s the best of both worlds: a chrome-less, information rich space when you want it, and a deep-dive intense experience for your more focussed tasks.

How Microsoft square this particular circle in the future is to be decided. Traditional Desktop apps will be around for a long time, and there are certain apps that I can’t image working as well in the paired-back modern UI without significant re-working. But this is how progress works.

Lack of Multiple Windows = Memory Overload for Complex Tasks

It’s been proven time and time again that human’s can’t multitask. It’s a fallacy driven by the ability of our brains to work so quickly. That said, we enjoy working on multiple things at once, we think we’re good at it, and we probably all find it quite an effective way to get stuff done. Once you get beyond a certain number of discrete tasks however, your total productivity will actually decrease and you’ll get less stuff done for a given time period. How many times have you lost the browser window you’ve been working with? Or forgotten which browser window has particular tab holding the web page you need right now? Apple’s Expose helps manage this chaos to a degree, but wouldn’t it be more optimal to focus your attention on just what you need to do right now? That’s why an iPad can be quite an effective tool to get stuff done – you’re locked into one app at a time and have to focus. Windows 8’s full-screen apps have this efficacy. In a critical point that Nielsen seems to completely miss, Windows 8 does allow multi-tasking – albeit in a controlled way – through snap views. You can have two apps running¬†simultaneously, the main app taking up two-thirds of the screen, and the secondary snapped app taking up a third. Implicitly there’s a priority given to the apps too – which you can change at will depending on the given task.

Far from requiring you to remember more and therefore do less, your task space is deliberately constrained and focussed and you’re given just the right amount of content to remember, with snap view further facilitating by providing a necessary reference point.

Flat Style Reduces Discoverability

Nielsen complains that Windows’ modern UI is an extreme too far –¬†skeuomorphism may feel old and cheap to most visual designers, but users love the comfort it offers. In contrast, the modern UI in Windows 8 and Windows Phone 8 shuns reflections, gloss and any pretence to real-world metaphors, focussing on strong typography and sparse carefully-designed iconography. At first glance, this can feel alienating, inhuman and characterless. Especially if you’re coming from a world as lickable as iOS. Nielsen mentions that this flat style lacks suitable affordance: users are used to bezels and chrome indicating affordance. This is an old-skool traditionalists opinion. With the chrome stripped away, the user is left with information. Some can be actioned, some not. With less visual noise, cognitive load is reduced. Microsoft haven’t completely removed all affordances. Chevrons, iconography and colour are used consistently and sparingly to afford just the right amount of suggestion that an object can be actionable.

With such a legacy of heavy-handed real-world metaphors in every digital interaction until this point, the user is bound to have issues during first-time use. But when they’ve had time to relax and let go, the modern UI is a breath of fresh air.

Low Information Density

In this point Nielsen claims that the modern UI’s approach to reducing information density is a bad thing. This one is difficult to argue, as it’s so subjective. However, it is a well-established practise of effective visual design to give information space to breathe. Negative and positive space is critical to guide the semantic processing of information. Modern UI shapes information into long horizontally-scrolling pages which are elegantly laid out, ergonomically efficient to navigate and promote visual content alongside textual content. Providing what you’re showing the user is consistently useful, relevant and easy to consume, why does it have to be crammed in ‘above the fold’?

Overly Live Tiles Backfire

Live Tiles in Windows 8 and Windows Phone 8 are not just shortcuts to open the app, they’re also channels through which the application can surface content from itself to the user without requiring them to come in and have a look. Very glancable, very responsive, very personal. Everyone’s Start screen is different – facilitated not just by the multitude of different app configurations and layouts users can achieve, but through the content itself. Even if you and I have the same apps in the same configuration on the Start screen, ours will still differ as the content pushed to those Live Tiles is unique to us individually. It is a wonderful combination of specially-atuned shortcuts and dashboard style micro-delivery of relevant content.

Nielsen’s point here is that early application developers have gone overboard with their live tile designs such that a typical user’s Start screen becomes a blur of noise and irrelevance. Granted, some apps do enjoy surfacing less than relevant content. Where this is the case, users have the option to either shrink or completely disable updates on a per Live Tile basis. However, I would argue that Microsoft have done such a good job in the algorithm that animates Live Tiles that you never see too many tiles updating at once or too quickly. Animation is applied judiciously and as a thin layer to announce the arrival of something new or relevant. Live Tiles must follow strict guidelines laid down by Microsoft on the position, amount and frequency of content updating. There are a small set of different types of Live Tile styles an app can deploy. So, while a few Live Tiles on screen may be animating at any given time, they do so in a controlled, consistent way following a set of pre-defined templates. The experience is far from chaotic and noisy – it’s relevant, measured and deliberate. Plus, the user is in complete control at all times.

Charms Are Hidden Generic Commands

Firstly, Nielsen appears to mistake Microsoft’s term ‘Charm’. Swiping from the right on a touch-operated Windows 8 device brings up the ‘Charm Bar’ containing a set of standard Charms: Start, Settings, Share, etc… Nielsen seems to be talking specifically about the Share charm in this item.

The Share charm is a generic command that allows you to plumb two apps together without them needing to know exactly what each does. For example, a Twitter app may say that it can use text and images, while a photos app may say that it can create text and images. Windows 8 will create the back- and front-end plumbing to allow the user to be in the photos app and Tweet a photo directly to whatever app can handle it. Apps create ‘contracts’ detailing what types of content they can create and consume, Windows 8 does the rest.

Nielsen criticises the share Charm as being too hidden, and that their action is inconsistent (he cites examples of users tapping the search Charm and being told “This application cannot be searched”). On the point of the Charms being hidden, swiping from the right is an alien gesture, but it quickly becomes second nature as you discover and use the Charms within. Ergonomically, a swipe from the right and tap in the vertical centre of the screen (to get back to the Start screen) is easier than taking one hand off the tablet and pressing the physical Windows key (at the bottom-centre). The Charm bar rapidly becomes a frequently-used function. Placing the share Charm in this commonly-used location exposes it to the user every time they open the bar. Far from hidden. On the point of inconsistency, the ‘This application cannot be searched’ case is so rare, I’ve never seen it. All the applications I have tried to search within have done something useful and relevant. I have neither experienced this myself, nor have I seen any user in a test experience this inconsistency.

Let’s take one moment to take stock of Nielsen’s processes. He cites examples of user behaviour in user tests he has conducted. From my own experience, the validity of findings from user tests where we bring people in off the street who have no prior knowledge of the area we’re testing should be questioned. In the case of Windows 8, there is a learning curve that must be traversed before you can accurately measure the usability of subsequent functions. Efforts to explain the unique elements of an interface like Windows 8 in the five minutes before a user test session begins are futile and artificial. Users need to be sat on their couch at home, in an unpressured environment to learn these new interaction gestures in their own time to form an effective mental model of how their computer works. Do you think Nielsen would have found such opposition to these new interactions if the users had been given the tablet to take home for a few days first and use in a more natural environment?

If what Nielsen is doing here is challenging the need for a learning curve at all, then I would argue that we’re back to the problem of local maxima. How can we find more efficient and productive paradigms for operating systems and software if we cling so rigidly to outmoded, outdated, clearly subpar approaches. Consistency and predictability in user interfaces is an absolute pre-requisite to good interaction design. Dogmatically clinging to a constraining legacy does the user no good in the long run.

Error-Prone Gestures

Here, Nielsen complains that the gestures are difficult to remember and to physically perform. I don’t see this argument as any different to the criticisms about hidden functions: they’re a product of users not given enough time to familiarise themselves with how the system works. I have first-hand experience of users frustrated with the tablet experience when they first pick it up – in fact, I can’t think of anyone who has picked up the tablet and used it without making a mistake first time. But does that mean it’s broken? Would you expect to be able to pick up a swiss army knife and be able to use all its functions first time? Of course not. A little mastery and practise is absolutely essential to get the best out of any tool. Once you learn and appreciate how the tool works, you can exploit it’s capabilities to the fullest extent.

I’m no Microsoft apologist. I use a Mac and used to be an Apple fanboi. But I have spent a lot of time with Windows 8, both personally and with real users – watching their experiences over time. Nielsen’s approach seems hopelessly outdated and behind the times, stuck in an entrenched world over fifty years old. The post-PC age is upon us. More people are interacting with their computers through touch than ever before, and current trends suggest touch-based computing will outpace exclusively mouse-and-keyboard-based computing sooner rather than later as devices converge. Microsoft has attempted to lead the charge into this new space with some unique ideas. Maybe we all need to stop panicking so much, take a deep breath, relax, and just see where it takes us. I urge you to try Windows 8 yourself (on a desktop PC or tablet). You’ll be frustrated at first, of course. But calm down and just give it a go. Not only could you be at the vanguard of the new wave of touch-first UIs, you never know, you might actually quite like it.


  1. likeshare · October 7, 2014

    Thank you for sharing your info. I really appreciate your efforts
    and I will be waiting for your next post thanks once again.

    • Lucy · October 7, 2014

      Thank you for reading! It’s funny: Nearly two years since I wrote that article and a lot has changed. Windows 10 shows Microsoft back-peddling on some of the key early decisions it made but, fundamentally, they’re trying to tackle the keyboard/mouse + touch hybrid computer paradigm head-on now. ‘Continuum’ mode (where the PC will automatically switch to the most optimal mode based on your configuration) makes perfect sense. The history strip down the left of the screen is being replaced by an expose-style visual multi window switcher and the Start menu is back. They’re listening to feedback. My only worry is that they’ll cow-tow too far and miss the unique opportunity to own the hybrid space in the near future…

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s