smorgasbord-design

Last week I took some time out of my day to talk with Nick Southgate (Grey) and Victoria Murray (IPA) about Behavioural Economics. Specifically we talked about choice architecture and how this reflects my day-job. I could, and did, talk about default setting on transactional forms, page hierarchy, button sizes, pupil dilation and all sorts of strategies experience designers deploy to assist (and, let’s be frank, hoodwink) our users. Whilst I’ve got a good grasp of the behavioural side, the economics does continue to confound me.

I’m not great with money, but I like organising things. I can’t remember when it was that I got a copy of MS Money, probably around 2002, but it co-incided with my increasingly web-based life. At that time NatWest’s site was an HTML-skinned translation of their internal systems and as such was largely designed by bankers for bankers and consequently a grind to use. Time-outs, slow, paginated statements and nothing that could really pass for analysis or interrogation existed. To this end I’d download as much data as I could and import it into Money. Putting aside the usual Microsoft prejudices, this was actually software that worked. It allowed me to flip to and from accounts,  associate spending to categories, identify trends, budget, calculate interest and so-on and so forth.

The strongest element of this management for me was the assignment of categories. Before this point I kept this kind of stuff in my head: “I’ve taken about £50 out in cash this month”, “about £100 in groceries this month”, “£1000 of my savings is allocated for my bike/laptop/holiday” and so on.

If it turned out that I wanted a laptop costing £1200 and I had £700 in savings and £600 in my current account I could technically afford that laptop, but I wouldn’t buy it because, in my head, I only had £700 in the allocated savings account.

This was (it appeared) no different to putting coins in a piggybank – easy to deposit, hard to extract. Or allocating three cookie jars for deposit bits of your pay to rent, bills and food. If the food jar is empty, you tell yourself you can’t raid the bills jar – even though there’s money in there to feed you… This is a common strategy, it’s called bracketing (Soman & Cheema) or Envelope System (ostensibly ‘invented’ in The Great Depression)

Banks, largely, don’t let you do this online at-source. Well, they do if you count opening up different accounts. When my wife and I moved into our first owner-occupied place this year we started a process of bracketing by consolidating and creating accounts to deal with joint savings, personal day-to-day spending, personal savings and day-to-day expenses – not to mention investments and shares. This is clumsy and annoying. It means some are tax efficient, low-access accounts, some are distinctly not. What happens – even with MS Money to keep track – is that money that could be earning better interest doesn’t move in to the savings accounts until we have time to do it or – as happened recently – current accounts approach overdraft because nominally-protected savings haven’t been used to prop-up unusual spending patterns.

[Aside: I once heard a colleague talking about a sweep account where all his current account surplus was swept into his savings at the end of every month. A superb nudge, why don't all banks offer this?]

This strategy has been shown to be used by gamblers too. Winnings go in to a different coat pocket than their stake money, taking from the winnings pocket is a physical objection that increases the level of consideration to give the gambler a fighting chance of keeping their habit under control.

I’d assumed that I was doing this in MS Money too, allocating money to ‘cash’, ‘rent’ etc. surely, the same thing? A virtual cookie jar right? Nope. As Chris Peterson so rightly puts it, this is not the same. It’s ex post facto. I’m allocating the monies having already spent them, analysing and tracking my behaviour rather than using strategies of bracketing and assignment to modify that behaviour. Ok, so there’s some overlap; by knowing I’ve a budget of £100 cash per month, seeing I’ve spent £70 is helpful to reign in the next trip to the ATM but it’s still essentially a mental barrier. It’s diagnostic, not preventative.

So, on to the solution (and hat-tip once again to Chris though the answer did rather suggest itself). My online bank should allow me to allocate jars/envelopes/brackets. I could see my total account balance as allocated to my identified needs: a balance of £5000? £100o for the mortgage, £1500 for this month’s bills, £1000 for the holiday savings. So, I’ve got £1500 to play with this month? Well, last month you spent £400 on day-to day costs … so that’s more like £1100. Not quite enough for that MacBook Pro yet…

Each of these brackets could be intelligent, based on that post-analysis of your spending. They could be temporal, expiring once a time limit is reached, or rule-driven based on the achieving a financial goal. How strict these brackets are can be user-defined, do you allow one bracket to take from another under given conditions? Perhaps after asking you, ‘are you sure’? The equivalent of hiding the alarm clock on the other side of the room to make it harder to hit the snooze. This is information architecture, it’s choice architecture. It’s the taxnomic organisation of your finances and no bank is doing it in-house.

Sure, you can, of course, sign-up to online services such as Mvelopes promising this kind of functionality but who wants to do that? Why complicate the experience? I wouldn’t sign-up to Google Mail and then find a third party to manage my folders and mail rules. So, instead of blithering on about Mint et. al to financial clients and pointing out analysis capability they’re missing out on delivering, perhaps we can identify some territory that they could actually own … that is if they have the agility & appetite to do so.

Poor old Rory, he can’t seem to do anything right over at BBC’s dot.life. One minute he’s been told he’s anti-Apple and the next, an Apple fanboy. Theoretical biases aside, his report on Jonathan Ive’s RCA session, “Listening to Mr iPhone” made for further corroboration that designing for people needn’t mean asking those people what they want:

“… how did [Apple] decide what customers wanted – surely by using focus groups? “We don’t do focus groups,” [Ive] said firmly, explaining that they resulted in bland products designed not to offend anyone.

Christopher Frayling reminded us at that point of Henry Ford’s line about what his customers would have demanded if asked – “a faster horse” – and it’s surely true that the point of innovative companies is to come up with products that customers don’t yet know they need.”

Though the solution may not come out of a focus group; the problems certainly do. It is the interpretation of these problems that reflect – as human-centred designers – our skillset, in the same way a keyword reflects perhaps something of a user’s need but not necessarily their motivation. Recently I’ve begun to find myself faced with a long list of requirements from clients who’ve “listened” to their users and are passionately trying to ensure their next online release is the answer to their range of gripes and desires. It’s hard to knock these back and say “the customer’s wrong here” but in quite a few cases they will be. If we don’t we might find ourselves with the antithesis of Ford’s vision, The Ford Mondeo. A car so everyman, so bland and box-ticking that it did nothing spectacular at all – becoming a byword for mediocrity.

Ive doesn’t design mediocrity and the long-held assertion at Apple that the focus group is ineffectual hasn’t done their revenue or sales figures much harm has it?

:: More on how to interpret what users want from what they do/say, in “Desire Paths and Digital Ethnography“

Anyone who knows me, I mean really knows me, will be aware of my passion for the analogy. Explanations to clients and colleagues are, like a plasterer’s radio, liberally splattered with them. One of the strongest analogies that us Information/User-Experience Architects deploy is that of the construction industry; borrowing heavily from ‘real’ architecture. It’s how I explain what I do to my mum. I think she gets it.

So, when I picked up on Dustin Curtis‘ fine post on The Bathroom Door recently, it really struck a chord and thence went straight into my toolbox of bon mots. The story goes that Dustin visits a wealthy executive’s San Francisco house for a party. The house is “insanely beautiful … just about everything in the house is remarkable”, but this experience is tempered by the experience of using the bathroom door. A door with a knob which “felt loose and hollow, like cheap crap”. The paradox of being in a jaw-dropping house that relied – for one of the most-used interactive elements in the entire building – on shoddy ironmongery provides a wonderful analogy for interaction design.

Of course, it’s nothing new: attention to detail, red-route usability and so on. The key here is that by making some of these most-used interaction elements absolutely as solid and well-considered as possible then we go a long way to defining the experience overall. So, I ask you this: where is the bathroom door on your site, and how flimsy does it feel?

Detail of the Morzine piste map

I just got back from skiing in the Port de Soleil area in France. It’s only the second time I’ve been skiing, the first being three cold and very wintery days atop Cairngorm Mountain in Scotland. I always expected the French Alps to be a more impressive affair; freeway-wide pistes, tree-lined rat-runs and acres and acres of skiable terrain. Add to that the inumerable chairlifts, gondolas (correct plural) and tow-lifts and the region was a beffudling confusion of signage and mapping.

I’d spend a large part of my afternoons (i.e. outside of our guided lessons) fixated on a tatty series of piste maps, glancing myopically at signage and lift names and trying to figure out where on earth I was, where I had come from and how I was going to get anywhere. This confusion was made a little worse by my inability to ski much other than ‘easy’ reds and (of course) the blues.

My issue is very much with the piste map. Even before I’d ever been on skis I’d seen these posted in hotels and magazines when summering in ski regions. Generally over-stylised in water-colour or faux-photographic rendering, they attempt to do more than just locate the various pieces of infrastructure, they are attempting to visualise a mountain range in all its wintery glory.

And this is – one of the reasons – where they fail. For a start the ones I was using rarely showed the direction of the piste. Of course, you can only really ski downhill and it’s not always clear in which direction the piste descends. Secondly the intersection of piste and lift was often ambiguous. Pistes may appear to end at a lift but the reality may be that the lift is the other side of an uphill ridge. Finally, the colour and weight of the line indicating a lift was unhelpfully typically black which – as you may know – causes a little confusion when establishing if it’s something I should avoid (difficult piste) or savour (circa 10 minutes of peaceful ascent).

My personal take is that we should follow the example of Harry Beck, the chap who rather sensibly realised that the relative geographic location and landmarking on the original London Underground maps was irrelevant to the subterranean traveller and thus a schematic representation was more effective. Using landmarks is not helpful on the piste, whilst it may be effective to the guide or experienced navigator, the majority of recreational users will find the succession of snowy peaks and wooded valleys impossible to differentiate – especially in low visiblity conditions when correct navigation is even more crucial. The maps, for example, only work in one orientation due to their use of shading and perspective.

So the market is open for the production of a piste map that supports:

1 :: piste direction (and potential graduations as green, blue, red & black not detailed enough)
2 ::  lift speed and capacity
3 :: lift and piste interchanges
4 :: non-geographical routing

I’m going to take a section of the Morzine (Pleny side) map and make an attempt at this, if you know of any Piste schematics that are a little more like this then let me know. I’ve also see examples like Whistler’s and Avoriaz’s that allow you to switch features on-off but this isn’t exactly practical on the slopes…

Let’s see where this goes.

So, Nathan sent details around today of a code ‘hack’ in YouTube where you can link to a specific time code. Handy for those ‘forward to 01:24 to see the guy get hit‘ type links. I’d seen it in the comments before but didn’t know it was possible in a direct URL, viz:

{YouTube url} #t=2m14s  (which would link to 02:14)

My beef was that this was a power-user feature and that code hacks like this could be better served by a UI fix. Or, in the very least, fix it to:

{YouTube url} #02-14 (again, linking to 02:14)

So it’s hardly going to change the world but then again making a toilet door latch green and red rather than ‘vacant’ and ‘engaged’ isn’t either but it’s the kind of incremental fix with saves split-seconds and cognitive load.

Anyway, far better would be this, a contextual right-click on the playback either on the video in-situ or on the timeline. A modal call-out which enables you to share the url from the time-reference by invoking the send-mail function or pasting to clipboard. You saw it here first.

Whilst it’s true that I used to irk easily, it is equally true that my ire – more specifically my blogging of that ire – has diminished in recent months. Trains, however, have a peculiar ability to provoke a rant.

Today the train itself is fine. Heading to Newbury to talk to Vodafone about great digital work is apparently effortless; warm, empty trains, peaceful. Tranquil. What concerns me is the timing of those trains and my connections yet I can do little to easily plan and check my route.

What I really fancied was an iPhone app that pulled UK train times and allowed me to search a real-time route planner and status check. The National Rail site is a visual dog’s dinner and the mobile version basic. I eventually found a PDA version which worked nicely though.

Steve tells us that making apps is easy – he’s right, there are thousands of them and some are excellent. I can get apps to check almost and transit system in the world – including a nice London Tube one – but UK rail? Nope.

There was a great one apparently, MyRail, but that got pulled – presumably falling foul of some daft timetable copyright issue.

The fact is that National Rail should have an official, trusted app that encourages greater train usage. If they can’t do it, license it out to the likes of the MyRail team and rake in the £2+ each time it’s downloaded.

By insisting people always visit a website they’re missing an accepted trend that people are increasingly interacting online outside of the browser space: Adobe AIR and iPhone apps are just two examples of the way this is happening.

In the meantime I’ll probably sail past Basingstoke and miss my connection whilst fumbling around through radio buttons and drop-downs and negotiating ghastly tables on the National Rail site.

UPDATE (24th March 2009)

Rory Cellan-Jones has picked up on this in his blog today: “Who owns the train times – or the news?”

Turns out National Rail had the idea to monetize this data and thus have pushed the excellent MyRail out of the market by insisting that the data is not free. The National Rail app is £4.99 – an absolute rip-off when MyRail did it for free. I’m in a dilemma. I really want to see MyRail survive and yet I realise that it makes a lot of sense for National Rail to run an app too – and to make money from it to plough back into the network – but the competition should be around the user experience not the data. Open source, skinning the same data with paid-for quality Ux

I really am torn about this, data should be free but, whilst it has never been more popular, it has become harder and harder to justify providing it if it can’t be made to pay somehow. Everywhere I turn I see people retreating to positions of monetization: Stephen Fry recently conceded it was impossible to offer all his content for free, Twitter too are running out of VC cash to realistically continue to operate an entirely free experience.

I haven’t played a computer game for a long time. I used to have an Amstrad 6128,  and Amiga 500 and later 1200 and when I moved to PC I had a few games on that. The best were simple; though seduced by graphics on the back of the pack in Computers Plus (Sittingbourne High Street) it was inevitably the most playable games that had longevity. Dizzy, Sensible Soccer,  Mega Lo Mania (sic), Speedball II, Red Alert (’im sure my brother knows more). When I picked up my iPhone a few weeks ago I thought I might try out the games in App Store having seen my wife lose hours to Brick Breaker on her Blackberry.

A recommendation from Nick at work lead to me picking up Fieldrunners this morning. Fieldrunners is a classic tower defence game and in that respect learns from the classic Amiga and early PC strategy games that a cute characterisation paired to simple strategic thinking leads to something akin to crack. It’s maddeningly addictive. We’re talking hours of game-play this morning, an entire battery’s worth and I can’t see that fading for a while.

The interface too is simple, in game it makes use of gesture and touch to allow the player to drop towers onto the field of play, to zoom and scroll and that is pretty much it. The principle strategies are to prevent enemies moving to the goal position from one (or many) entrance points, the answer being to place the towers carefully to slow and eliminate their progress.

I’m sure there are games in App Store that make far greater use of the iPhone’s accelerometer, the gyro and the graphics and sound capabilities but look back on gaming history and identify the slew of pretty but dull games against the simple but addictive ones. Those you return to time and time again (if you’ve still got the hardware!) are those that the developers played more than they painted.

An interaction designer can learn a lot from this parsimony. So, Nick, forgive me if I treat Fieldrunners sessions as essential homework for large slews of my billable hours…

When I first moved to London (nearly a full year ago) and began commuting on the busy South West Trains line through to Waterloo and thence to Oxford Street, I found the whole experience incredibly frustrating. I remember almost-anthropologically observing the peculiar behaviours of the expert and the novice negotiating those sticky transit punctuation points: where to stand on the platform, how to squeeze on to the train/tube, how best to glide through the barriers and scooti up/down the escalators. Over time this becomes terribly mundane of course. The expert-commuter begins to boil over at the sight of a novice not closing the gap on the escalator, fumbling for their ticket at the barrier and dawdling along when all about them is hustle and bustle.

NY Commute: Christoph Niemann

Leaving aside the melancholic fug that has descended on the wintery commute and poured ‘Recessionary’ water on the urgency and fire of people’s routines, I’m all for insider-tips to expedite one’s passage at that all-important time of the day. Christoph Niemann is an artist whose work has graced the covers of The New Yorker and New York Times Magazine (amongst others) and Christa passed me a link to his New York Cheat Sheets in the NY Times a few weeks back. 

I’ve picked out this one to show how he uses that indigenous and well-earned commuting savvy to – in the best Edward Tufte way – illustrate how best to commute from Clark Street station: how to find a (more) quiet spot on the platform, where to deposit one’s paper and which exit to take. It’s so quaint, mildly Autistic (and, dare I say it, peculiarly British), I’m simply going to have to create my own, undoubtedly inferior, version.

In the meantime, you could do worse than check out his recent blog postings which deal with another perennial favourite of mine, the purchase and consumption of coffee.