Jakob Nielsen is not your Windows 8 guru.    Here's Why.

November 20, 2012

I can't believe I'm about to defend Microsoft Design outloud like this. It's not something I would normally do, however when it comes to the Jakob Nielsen Windows 8 review I just can't stand to let it slide. Personally I think that entire company is still stuck in the past and has consistently failed to navigate change with a degree of accurate prediction since they declared Flash a fail (Oct 2000) (which translates to in principle to JavaScript based websites a fail).

Furthermore I think they rely on the idea that the end users are all collective virgin users who have never had to navigate or use bad UI in todays software environments. The fact that we as a human race can navigate even dumb solutions such as Sharepoint, Lotus Notes, SAP and a whole host of other really badly design UI indicates that we aren't as dumb as useit.com would have us believe. Furthermore there is a huge generational change underway whereby the concept of "experienced windows users" would be fair to say my 8yr old son fits that category.

The clue is in the audience sampled as if you get that wrong the rest of the responses are just opinions based around a skewed bias (bad baseline to draw from on their part). Here is my notes from an internal email I sent around when I was asked "what do you think of the article" from my co-workers.

My remarks:

• Novice and Power Users. "Invited 12 experienced Windows users" is a weak / broad sweeping remark to make that XYZ demographic doesn't like N-Product. Keep in mind I'm a tough critic of Windows 8's design, but even I can concede it's still usable whether the incentive is to use though is entirely different matter (Cognitive Dissonance measures Behavior vs. Incentive). I would have taken him more serious if he had of used a variety of audience(s) for this (OSX users, Seniors, GenY, IT Professionals, Sales force etc) .. everyone's experienced In Windows is my point.

Cognitive Overhead.

• Prospective Memory — I think he's building up to "learn where to go" and associating it as a bad thing. The concept of a desktop works in favor of prospective memory, meaning "I'll put x here so I can come back to it later" works in the same fashion as the start overlay. Its not ideal, but to declare this a cognitive overload is an over-reach given over time (behavior) users will settle on a rhythm that suites them. If I press START and start typing my context will adjust to the text I'm typing and so on.
• Dual Environments — The two environments in which he speaks of are WinRT and WinRT Pro, now the clue is in the word "Pro" firstly and it has to do with legacy support than actual user experience (context is annoying when you leave it out huh?). Tablet users won't interact with the said duality he's nominated so it kind of is a weak point to rest on and those that opt for the Surface Pro edition are doing so more as a finger in both pies approach to the problem at hand. If I pitched the problem that needed to be solved in that I need the user(s) to have both Windows Now and Windows vNext it shifts the results differently as if I said I need the users to solely focus on vNext only … Again, It feels more about airbrushing the facts without context (Ironic given the guy's a usability "guru" and how context is important in ux as content).
• Added Memory — I see this a lot and I wonder if UX Practitioners suffer from this concept that we all suffer from sudden memory loss at any given point. I understand interruption etc plays into this but in reality we don't multi task and phones today for example don't have this issue – if anything given the complexity between switching from apps via navigation routines (ie iPhone double hitting the rectangle and using a slider style switch). I am baffled as to what moment of brilliance the author assumes he/she is uncovering here – I'm kind of lost between whether I dislike his point or the actual website itself's design.

Multi Window.

• Responsive & Adaptive Design — I think the author again (they really should sit down and study some basic design principles to articulate the points) probably wanted to say that the design of the solution isn't responsive and/or adaptive depending on screen real estate. The said applications again don't make full use of the screen(s) they are being deployed or used upon. I concede that this could be an issue for usage of LOB solutions but at the same time I also reject it. Having window support in today's UI world is an absolute engineering challenge at the best of times and furthermore buy having to adhere and cater to this we in turn limit our future potential by sticking to the ye olde side by side window usage. As it now begs the question, why are two applications side by side if they are related? If we have a forcing function which puts emphasis on a single screen visualization would this not cut down on fragmented software delivery? What if the snap screen concept could be more broader in its execution where you allow users to have more than one window at a time but the designs themselves can be responsive to the state in which they are housed? This works better imho than just given floating cascade windows with dynamic border resize + maximize + minimize. It fixes and creates an interesting solution to much bigger problem.Again, the author is kind of saying "it's changed, I don't like it". I didn't like the day I gave up a tactical keyboard for a touch screen, but I got over it and can type just as fast now. Humans evolve.

Discoverability.

• Flat styles — I can't tell you how many times I've been driving along the highway and seen the turn arrow being flat and thought to myself "I wish that had a sense of depth, as that would give me contrast to make an informative decision". The whole idea that we need depth in order to associate action is a kind of "drawing from a long bow". If you're a virgin user and never seen something for the first time, yes, you have that moment of initial "wtf" but you explore, because now it's a puzzle and you have an incentive to figure it out. Take into account marketing and real-world surroundings it's fair to assume that the learnability of a solid icon is considered both touchable and untouchable. You will discover this fairly easily but the learnability is probably shallow but discoverability isn't – Key differentiation there. I don't agree with Metro's content over chrome metaphor and in the visual he provided it's an easy fight to pick (grouping is all wrong) but the failures here are easily misleading given he left out the constancy of the design (in that it's not isolated to one area, it's throughout and again, surprisingly we all seem to navigate over time without issue – behavior vs incentive again).
• Symbology — Probably the only thing I would agree here is that there is way too much of a strong reliance on symbology to convey the context of what the said solution does. There's no personality attached to apps and functions, meaning I think there still needs to be a balance between core operating and in-app functions and said Applications (one thing iPhone does well as the apps entry icons are able to retain a differentiation whereas Win8 it doesn't). I don't think the author articulated this very well but I sense that's the direction they were heading

Information Density.

I won't bother remarking too much on these areas, suffice to say it's like I grabbed Angry Birds app, declared iPhone a fail due to lack of 3D support. Probably helps to separate third party applications from the actual said operating system. You can grade an OS based on its actual abilities or inbuilt functions, not by what the ecosystem does with them as that's a slippery slope.

Desktop computers and horizontal control hasn't been a failure. I don't subscribe to the "well on websites it failed" it actually hasn't, its more to do with screen size, frequency of use and does the UI tease the user to carry out the action. It's not a complete failure it's more to do with context and case by case. Now the current win8 mode relies on the horizontal scroll bar or mouse wheel to navigate between the screen and yes I think the missing element here is for the mouse to do the flicking between left/right (kinetic scrolling etc).

Live Tiles.

Agreed. Probably the one area of this article he nailed well. Yeah, the live Tiles for me is like a room full of screaming kids all asking for ice cream and one asking to go to the toilet. Pray you get the later right early.

Charms.

• Progressive Disclosure has always been a double edged sword. On one hand you free up user from distraction by giving them a chunk of information to process act upon whilst on the other hand you're easily forgotten and totally rely on muscle memory / learnability to be your UX crutch. I don't think the author framed this correctly in this case by asserting that the users will "forget" the charm icons etc. I think it's got poor amount of UX friction associated to it but the idea that Novice/Power users will be absent minded users here is really again an over reach. I find the whole persona attachment in this authors writing to be disconnected and fluctuates between a virgin user and a veteran of 15 years+ user? (settle on them and grouping here clearly needs to indicate the level of friction associated to each point).Had the user stated "I sampled a user with only 6month usage of a computer" then yes, Charms would be hazardous to one's health. The reality is that's a generational issue firstly (ie they are deprecating) and secondly there is such a wash of bad UI in software today that the users in general are what I'd call "defensive" in that they have been trained over and over that UI today isn't always a case of "everything is in front of you where you need it". Furthermore if you take a step back in time and look at the green-screen terminals and how data entry operators would fly through the various fields etc one can see that a human and pattern recognition have remarkable abilities.

Gestures.

I've not used Win8 Gestures to comment. I want to take the author at his/her word but so far I'm inclined to favor Microsoft here. That being said, Microsoft and Touch have never really been that good together (even Surface Table had issues here). Suffice to say they really need to tidy up NUI in general here and its still the wild west, so in reality anything that all brands put on the table is open to this set of arguments.

Windows 8 Weak on Tablets, Terrible on PC's.

Yeah this is where the true bias shows through and why my UX spidey senses tingled. It's in this part you see the opinion shine through which can distill down to that they wanted Win8 to be tablet only UI and desktop to continue the Win7 as-is approach. It shows lack of foresight for how the mobility and desktop market's are starting to eat away into the tablet focused approach. How well we handle the ergonomics of going between a laptop to a tablet is still undecided but that's the direction ones heading. Microsoft are trying to get out ahead of this early and if that means along the way they will fumble some of the UX by giving a duality in both old and new then so be it. In my view if you are given the problem of retaining the old while moving the user base over to the new in an aggressive manner then Microsoft may actually have a winning idea (yes I just praised Microsoft). I would however say that there Metro design style is going to come back and bite them the most and from what I can tell the Author has been cherry picking the negatives in order to build up to a point of how unusable it is. No balanced proposition here other than I don't like Windows 8 and here's why (hence the whole paragraph of "I don't hate Microsoft but.." which translates to "I'm not racist, but.." …there is no "but" ..as everything you just said before it gets lost in cognitive overload (grin).

How the author then goes onto praise Ribbon Menu after spending a paragraph or two downsizing the charm bar "out of sight out of mind" makes me confused

Lastly by asserting that Win7 needs to be replaced with Windows 8 is probably the final conclusion that Microsoft marketing still sucks at its job (ie it's not an upgrade, its an additive product) and lastly the user should stick more to the UI principles and less to OS Market analysis.