There’s no doubt that smartphones, particularly the iPhone, have contributed to a phenomenal boom in the mobile web. Before Mobile Safari the Internet was a painful place to get around. However, I would argue that the primary reason mobile browsing suddenly exploded was due to the introduction of unlimited (fair use) data tariffs.

Not having control of your monthly bill adds stress to your mobile experience. Without being able to accurately gauge how much your internet traffic is costing you, you reduce your traffic or stop it completely. Users aren’t able to calculate how much a page “weighs”. It was already difficult when using WAP-enabled phones when pages were text-only, but now that Mobile Safari offers you the whole of the web (Flash excluded obviously) page weights and data transfer becomes impossible to calculate on the fly.

I remember trying to access some of Orange’s WAP sites when I first got a capable phone. It was a nerve-wracking experience, with the over-riding thought being to close the connection as soon as possible. Try and find what you need and close the connection. It’s no surprise that I rarely looked to my phone for the Internet. Yes it was slow and hard to get around but it felt like money was, figuratively, trickling through my fingers!

With the introduction of the unlimited tariffs, suddenly I could browse for pleasure. I could idly play with the Google Map app and watch my progress on the train; skim through the headlines, reading any or all of what caught my interest; play games that connected to transfer progress and save positions – and all safe in the knowledge that it wasn’t costing me any extra. Without this piece of mind I know my web use wouldn’t have been so great.

So now we face a reduction in our use on O2s new iPhone 4 tariffs and all this is due to end. Already I’ve been semi-seriously auditing my data use. Which games connect and what do they transfer? How big is the average YouTube video? How big are the data files in Maps? Not only are these tedious things to consider; users shouldn’t have to!

O2 contends that the introduction of the limited tariffs will improve the majority’s mobile experience and I’d certainly be all for that. But that just doesn’t fit with their assertion that 97% of customers won’t notice any effect.

What I think O2 are more nervous about is the rise in popularity of mobile music streaming services like Spotify and surely more recently the launch of the iPad. With increasing numbers of people and increasing mobile usage, O2 are legitimately concerned about the effect on their network. “Total mobile data consumption on our network is doubling every four months,” a spokeswoman for O2 said. “And the number of iPhone users on O2 has grown in a year from 1m to 2m.”

However, punishing the majority in order to catch the few just doesn’t seem fair. Just look at the uproar tax on alcohol had when responsible drinking campaigns suggested we should all bear the cost of the binge-drinking few. And anyway, surely O2’s “fair use” policy should catch this 3% of heavy users out?

O2 describes their data flow as a “permanently-on tap”. But surely the answer isn’t drought?

Interestingly, I’m getting visited by something at one specific IP address again and again but apparently using different browsers and operating systems. One quick IP look-up later and it seems Adobe are hitting me about ten times a day, every day. Either someone there loves trees too or else some sort of bot is coming back again and again.

Adobe don’t have a search engine (as far as I know) so why are they bothering? Have I managed to offend them, or has a page of my site somehow got saved in their browser lab? Enquiring minds want to know! Anyone got any ideas?

The conference ran over three information-filled days. Day One was lecture day, with inspirational talks from the likes of Peter Merholz, Luke Wroblewski, Dan Saffer, Jared Spool, Jeffrey Veen and most excitingly Don ‘The Don’ Norman.

Days Two and Three went into much greater detail, with interactive workshops covering all aspects of user experience practice; from sketching lessons to learning how organisations can make better decisions through design. In fact there were too many fantastic workshops to get around all of them and there were no “fillers”.

One of the highlights for me however was hearing Don Norman speak on Day One. The author of seminal books such as The Design of Everyday Things, Things That Make us Smart and more recently The Design of Future Things, Don is well known to all user experience architects and designers alike. The fact that his comments were Twittered with the hashtag ‘#TheDon’ just goes to show the affection and regard in which he is held.

Often the contrarian (“When everyone is asking for something, I tend to take the opposite approach”), Norman has recently caused minor storms by arguing that simplicity is highly overrated and that complexity is good thing. At first this approach feels wrong: as usability people, we are often in the habit of trying to make online experiences as simple as possible. Surely complexity can only harm the experience and put customers off?

Norman offers a familiar example of simplicity: the Google search homepage – often quoted as the epitome of the simple. Without a doubt, Google is by far and away the most popular search engine. And yet Yahoo! have the most popular homepage and it’s packed with information. Yahoo! is optimised for exploration, with Google it takes a little more work. You might argue that these two homepages are different products but the sheer popularity of  Yahoo! goes some way to show that complex pages are popular.

Another example: the iPhone is often touted as simple and consumer-friendly. And yet with a new software update, Apple has added 100 extra new features. How is this level of complexity compatible with the idea of a simple product? One might argue it’s because users have learnt how to use the device and are now demanding more advanced tools (like Copy and Paste).

There appears to be a fundamental conflict here: when asked, people will demand simplicity (“Why is it so hard to use?”, “Why can’t products be simpler?” [1]) but when you watch these same people comparing products side-by-side, it is the number of features that sell a product. People want more features even when they realise this must complicate the product. People believe that as you add features you add capability, thereby making more feature-laden products more desirable. However, as user experience professionals, we believe adding more features decreases usability. Both positions, Norman argues, are wrong. “We must distinguish complexity from confusion, perplexity, and unintelligibility. The goal is complexity with order, lucidity and understandability.”

People prefer complex things. If it’s too simple, it gets boring. Once a user gains experience with a product, the user moves into a new role; that of the Intermediate and suddenly their perception of what is complex changes.

An aside – roughly speaking, there are three classes of user: the Beginner, the Intermediate and the Expert. If we plot the number of people against perceived skill level, like most population distributions we get the classic statistical bell curve, with most users situated in the middle of the curve at ‘Intermediate’. It stands to reason therefore, that these are the users we should spend the most amount of time designing for. And yet it’s often the Beginners and the Experts who get the most attention. The Product Manager demands the Beginner must be able to hit the ground running and yet the engineer or developer, if left to their own devices, designs for their own skill-level – that of the Expert.

So how does Norman suggest we solve the complexity problem? Unsurprisingly the first approach should be through well-researched design. By modularising actions we can contain the complexity and by teaching users as they go, we can help them manage complex interactions.

I’ll leave the last few words to The Don himself: “Why are things so complex? Because the world is complex. Our tools must reflect reality. Complexity can be good, leading to a rich, satisfying life, filled with rich, satisfying experiences.”

And again: “The mark of the great designer is the ability to provide what people need without excessive complexity, without feature bloat. Simplicity should never be the goal. Follow the famous Einstein quote: “Everything should be made as simple as possible, but not simpler.” Complex things will require complexity. It is the job of the designer to manage that complexity with skill and grace.”[2]

Thanks

To Luke W for his notes on the lecture that were far more comprehensive than my own.

And to the Twitterers who helped refresh my memory.

References:

1. http://www.jnd.org/dn.mss/simplicity_is_highly.html
2. http://www.jnd.org/dn.mss/why_is_37signals_so_1.html

Hotels have been welcoming guests for many hundreds of years, and certainly since the invention of the light bulb, interior and electronic design have been integral to the guest’s experience. You might have thought, therefore, that hotel designers have had a chance to get things right. However, contemporary hotels are under pressure to provide all kinds of modern convenience. From wifi and iPod connections to innovative, comprehensible controls for taps, heating, lighting, television, cable, telephones, room service and trouser pressing – all in one room, without cluttering the place up or making it look ugly.

I recently had the pleasure to stay in the centre of Barcelona, in a fantastic new hotel. I’ll leave their name out of it but in case you’re up on your “boutique” Barca residences, this one had an open air swimming pool on the roof…

The hotel was marvelous and the service was wonderful, so no complaints there. The problem came in the room when trying to control the lights. Surely the lighting problem has been pretty much solved by now – it can’t get much harder than on and off? Or can it…

The lighting system in our hotel room was incredibly counter-intuitive. Suffice it to say my girlfriend got extremely tired of me trying to work it out, especially when that meant my fiddling turned the lights on and off every five minutes. However, the usability issue that really caught my attention was this – why illuminate an off switch in the dark? Surely it would be much more useful to illuminate the “On” switch, so that when you’re desperately trying to creep to the bathroom in the middle of the night, you don’t have to fiddle about? I can only imagine the button was wrongly labelled and then wired in according to the incorrect label? Or does someone see a benefit that I’m missing?

The fixed control – pretty simple huh?

So, it was particularly interesting for me to hear the following UIE podcast upon my return, where this sensible question was asked (and I paraphrase) “If Don Norman’s seminal The Design of Everyday Things is twenty years old, why are simple design mistakes still being made?”

Let me know.

Mistake proofing

Shigeo Shingo defines three types of Poka-Yoke:

1. The contact method identifies defects by whether or not contact is established between the device and the product. Color detection and other product property techniques are considered extensions of this.
2. The fixed-value method determines whether a given number of movements have been made.
3. The motion-step method determines whether the prescribed steps or motions of the process have been followed.

A familiar example of the poka-yoke philosophy is the 3.5″ floppy disk. You probably won’t need reminding that the top-right corner is shaped in a certain way, meaning the disk cannot be inserted upside-down. Undoubtably helpful in all sorts of situations, so how can this be applied to the web and websites?

Certainly one the most frustrating objects encountered on many websites is the ubiquitous web form. Where ever you go on the web, a form is the most likely way you are going to interact, order, sign-up, comment or log in. There are server-farms full of pages on how to create html forms so I’m not going to go into code detail, but maybe it’ll be instructive to run through Shingo’s three types of poke yoka with respect to internet forms.

Our goal is to remove, or at least reduce, the number of errors occuring on our forms. Often sales or contacts are lost when users become frustrated by malfunctioning online forms. The more we can do to prevent mistakes in the first place, the less money we’ll spend fixing things later, and the happier our customers will be!

The 1-10-100 Rule states that as a product moves through the production system, the cost of correcting an error multiplies by ten. Ensuring an order is correctly entered costs us £1. Detecting an error at the billing stage costs £10. If the error isn’t caught in time and is passed on to the customer, that costs us £100 to fix. If we don’t fix it at that point and the customer complains to their friends, the cost can quickly escalate to £1000!

The Contact Method

This group of checks include “methods in which sensing devices detect abnormalities in product shape or dimension by whether or not contact is made” (Shingo). These methods are usually used to check size, shape, orientation, presence/absence, or appearance. They are also referred to as “physical poka-yokes”.

Although web forms don’t usually handle product shapes and physical objects, I think in this group we can include the following tests:

1. always make sure an email address is correctly formatted (often acheived by testing the ’shape’ of the address, in that it contains an at symbol and a fullstop)
2. required fields must always be tested for the presence of a valid string
3. phone number fields must contain numbers

The Fixed-Value Method

Fixed-value methods – or counting poka yokes – determine that a fixed number of movements have been made and is analagous to ensuring all your required fields have been completed. But they might also be seen in online games, or where users have to accept terms and conditions. A purchase-step progress bar could also fit into this category – those identifiers that graphically display your progress in an online transaction. They give you a count of how many pages you are away from completing your transaction, thereby providing some reassurance that progress is being made.

The Motion-Step Method

The third poka yoke method uses sensors to determine if a motion or step in a process has been completed. If the step has not been completed or has been completed out of sequence, a sensor signals the process should stop.

When applied to the web, you can quickly see who this can easily be enforced by a suitable linking strategy. Users should only be presented with links that are relevant to their current situation. And if they are about to make a purchase, don’t complicate matters with adverts that can distract and lead them astray.

If an email address has been entered incorrectly, ensure the user is made aware of that fact immediately, rather than when they submit the form.

Summary and comments

The concept of a poka yoke device is useful in reducing errors made by users. Reducing the occurence of errors leads to lower costs and happier customers. Poka yoke is well established in the industrial work place, applying the concepts to web pages isn’t a perfect fit but the essential principles can help establish good working practices that lead to improved user experiences.

Anything to add? Drop me a line and if you aren’t a spambot I’ll add your comment below.