Insanely great, or just good enough?

Apple are pursuing perfection, but should they be?—Apple’s design strategy for the original iPod, in the context of adaptive design and hackability, with particular reference to the battery life debacle of 2004.

Ed. Back in 2004, Allan Chochinov asked me to write an article about the iPod's design for the Core77 website. What follows is the original edit of that article, published on 14 February 2004. The iPod was a couple of years old at that point, but questions about the difficulty of replacing its battery had just been raised in the UK Parliament — hence the focus here. This piece articulates some of my growing awareness of adaptive design, and how Apple might find a better balance. Obviously, some of it has dated (‘social software’ is what we originally called social media, kids), but two decades later, we might still argue about whether this 'better balance' has been achieved. The article was picked up at the time by John Naughton in his column in The Observer.

---

Insanely great, or just good enough? — Apple are pursuing perfection, but should they be?

“(T)his House notes with concern the difficulty people are having in replacing batteries for iPods, the new pocket-sized walkmans; notes that iPods were a favourite Christmas present, with a value of between £250 to £400, but that when people come to buy replacement batteries they find that they cost £100 and that they have little knowledge of where such batteries can be obtained from; and calls on the supplier, Apple, to ensure that replacement batteries are plentiful in supply and priced at a reasonable level.”—Early Day Motion 447 ‘IPODS’, presented to the British House of Commons by Labour Member for Parliament Lindsay Hoyle, 19 January 2004.

“Its design is pure genius.”—Iain Balshaw, England Rugby Player, on the iPod. (The Independent, 19 January 2004)

Occasionally, design flashes into the harsh glare of the public eye, usually in relation to some very obvious failing — the wobbly Millennium Bridge, the Columbia accident, the SA80 assault rifle, and so on. However, occasionally — just occasionally — public opinion also bestows the exalted status of ‘design classic’ upon a product within its lifetime.

Right now, the undoubted contender for living design classic is the Apple iPod, hailed by all and sundry as an instant masterpiece, from the professional design community to that English rugby player. Most notably, and perhaps most importantly, the public have voted with their dollars and euros, such that the notion of Apple’s ‘traditional 5% upper limit’ market share has been obliterated in the mp3 player market. This is to the obvious delight of Steve Jobs and an Apple transformed.

But that iPod-buying public has just suffered its first serious dent to their previously unlimited confidence in the product: the Great iPod Battery Debacle of late 2003. The meme went like this: iPod batteries die after 18 months and you have to get them replaced by Apple at such great cost you might as well get a new iPod. This wasn’t actually the case as such, but that didn’t stop the meme spreading. So something approaching a backlash has occurred, as this seemingly ‘insanely great’ product was suddenly shown to have a fundamental flaw.

It doesn’t even really matter whether Apple’s battery replacement scheme is good value or not. Whether it’s $100 or $10, the point was actually the general sense of shock that the product was less than perfect. Its halo had slipped. This is a cultural field, not a simple interplay of economics. Hence the column inches. Hence the agitprop Quicktime videos. Hence some British MPs asking questions in the House of Commons.

So why the shock? Why, on the one hand does Apple engender such a positive brand experience, and on the other, garner an equal and opposite reaction when it goes wrong?

In both, I think we see the effects of a design strategy which pursues perfection.

When this works — and it has often approached perfection since Jonathan Ive took over the industrial design team at Apple — it creates a community of users who will not just consume, but actively fight its corner. This is perhaps summed up best in Mark Morford’s article, “Lick Me, I’m A Macintosh: What the hell is wrong with Apple that they still give a damn about design and packaging and “feel”?”. Actually, the title sums it up by itself, but in case that wasn’t clear enough, Morford describes the facets only Apple seems to reach:

“Detail and nuance and texture and a sense of how users actually feel, what makes them smile, what makes the experience worthy and positive and sensual instead of necessary and drab and evil.”—Mark Morford

Quite. Yet pursuing this perfection is a winning strategy, but it can also be held responsible When iPods Attack. Moreover, it could be out of step with both the broader cultural themes around technology over the last 30 years, and a way to move forward.

It’s got to be perfect

This isn’t about whether the iPod design is good or bad as such. It’s clearly a fabulous device built by what is possibly the most talented product design team on the planet. What this is about is the underlying design philosophy at Apple, and the way that culture thinks about design in general.

Most product designers, informational or otherwise, are taught to attempt to attain perfection — to create the perfect device, the pixel-perfect user interface, the perfect newspaper layout. To create something for the user which gets it so right, first time, that the user achieves some heightened state, falling in love with the product at first sight, before going on to form a lasting relationship. We covet things which attain this status: say the Citroën DS or a Charles and Ray Eames lounger … and the iPod.

More Mark Morford:

“This is what they are known for and why their design team is so famous and why they win so many awards and why they engender such passionate devotion from their adherents and why Macs are still far, far superior to PCs and always will be. It’s true.”

In my opinion, he’s right. Yet ironically perfection still isn’t good enough. Because ‘good enough’ is arguably all we should be striving for. Why? Adaptive design.

I borrow the notion of Adaptive Design from IBM’s Tom Moran and his keynote presentation at Designing Interactive Systems 2002 conference, and from Stewart Brand’s classic book on the lessons of vernacular architecture and well-understood design patterns, How Buildings Learn.

In essence, adaptive design is about designing to enable the user to change things. You strive for ‘good enough’ as a starting point, such that the user feels they have a ‘way in’, almost an implicit goal of working through the finished design themselves. It sees design as a social process, developing over time, via a relationship with the user. It draws heavily from Brand’s idea of separating a building’s architecture into six different layers, from slowly developing layers through to relatively quickly moving layers. (Ed. I would later discover that this is really Frank Duffy’s idea, adopted and adapted by Brand).

The domain of the user is in those fast layers. This is where invention occurs. in those layers best-placed for everyday learning, keeping products useful. Enabling these layers to move at these different paces is what’s actually key overall.

Ed: Interested readers may want to review these other adaptive design articles, or start with my notes on Moran’s keynote and my subsequent presentation to AIGA Experience Design forum, London, on 5 December 2002.

So the professional design team’s attempts to attain perfection actually shuts the user out of the design process. This echoes thinking around other interactive media, such as in David Weinberger’s Small Pieces, Loosely Joined where he notes “the imperfection of the Web isn’t a temporary lapse; it’s a design decision.”

This ‘design decision’ enables the Web’s users to join in, to work themselves into the fabric of the Web. It’s why it works, and it means the professional designer has to be somewhat humble in this equation.

The key factors at work here are time and learning. What adaptation enables is a device to mould itself over time in response to changing circumstances. This can enable technological and cultural change — given how rapidly technology develops, how can we be sure this particular model for portable music storage is the right one going forward? We can’t. So we must enable systems to adapt.

Plug and play versus tinkering

Regarding technological change, there are essentially two approaches: an adaptive approach whereby you upgrade bit by bit over time; or the wholesale replacement model, whereby you replace the entire machine every couple of years.

Generalising, Apple have tended towards the latter, despite positive recent signs like their G5 side doors. Their designs don’t give you the feeling that you can fundamentally change the device. It’s never been a platform for tinkering with, ever since the original Mac came cased in the all-in-one housing.

This elicits a feeling of total confidence in the design — the Apple philosophy of giving you everything you need, with no tinkering and minimal upgrades (except software). In many ways, this is a good thing. Personally, it’s what I prefer. I don’t want to have to think too much about that stuff. It’s not because I can’t— I just want it to work, smoothly, easily, beautifully, which Apple products tend to do.

And this is how much of the simpler consumer electronics world generally works—the good stuff. For example, you just buy, plug, and play the Playstation 2. Likewise radios, looking at Tivoli, Brionvega, or Roberts as iPod-equivalents. Likewise, increasingly, mobile phones. The only thing you might change from time to time on these devices, tellingly, is the battery.

Ed. Much later, I would write about interaction design for musical instruments, as inspirations in this respect.

However, the overriding hegemonic model within personal computing is the Wintel PC, which currently accounts for 90%+ of the market. And the design philosophy behind that, such as it is, is one which enables constant tinkering with the hardware — swapping in and out RAM, hard drives, graphics cards, sound cards, motherboards etc.

In many ways this is a more adaptive approach. And it’s interesting that this would appear to be the populist approach — people like to feel they can buy a machine, and then adapt it over time, rather than replacing it every few years —assuming that this practice occurs beyond the hobbyist market.

The first problem here is that technology in this market is tending to take exponential, Moore’s Law-style leaps forward every few years and there’s a limit to the effectiveness of adapting after a certain point. The second problem is that this places a huge burden on the user to engage with this relatively complex technology. It’s a pretty steep learning curve.

However, those users are there. Most normal PC users at least consider incrementally improving their machine in a way that most Macintosh users don’t, whether that’s OS, RAM, hard disks, various cards or peripherals, whatever. And that doesn’t seem to have ever been part of the Apple strategy, post-Apple II. This vignette at the excellent folklore.org makes clear the ‘Jobsian’ model of hardware upgrade:

“Steve Jobs, didn’t agree with Jef (Raskin) about many things, but they both felt the same way about hardware expandability: it was a bug instead of a feature. Steve was reportedly against having slots in the Apple II back in the days of yore, and felt even stronger about slots for the Mac. He decreed that the Macintosh would remain perpetually bereft of slots, enclosed in a tightly sealed case, with only the limited expandability of the two serial ports … Burrell (Smith) was afraid the 128Kbyte Mac would seem inadequate soon after launch, and there were no slots for the user to add RAM. He realized that he could support 256Kbit RAM chips simply by routing a few extra lines on the PC board, allowing adventurous people who knew how to wield a soldering gun to replace their RAM chips with the newer generation … But once again, Steve Jobs objected, because he didn’t like the idea of customers mucking with the innards of their computer. He would also rather have them buy a new 512K Mac instead of them buying more RAM from a third-party.”

This story may have a certain angle to it, but this is still surely the fundamental point about Apple — which has rather too many echoes of the ‘Great iPod Battery Debacle of late 2003’ for comfort. Leaving aside the economics of forcing the user to replace rather than upgrade, the design strategy means a lack of ability to adapt, to evolve. All the evolution goes on within Apple labs — there is no room for the user in all of this. But that story is from a time when computers were built in garages, following instructions from magazines. The cultural context is Atari 2600s and the BBC Micro. Times have a-changed. But the basic human desire to adapt its surroundings and tools has not.

Battery assaulted

Back to batteries, though, for that’s where we came in; not hard drives or graphics cards, but batteries. Mobile phones enable batteries to be replaced relatively easily. Also, the design of these phones suggests you can do this. My Nokia 7650 battery hasn’t needed replacing yet, but it’s clear how I might do that when I need to. The affordances are all in place. There is a removable panel, flush to the back of the casing but with a fingernail-sized catch for release.

Ed. Of course, ironically enough, this paragraph would read very differently within a few years, after the release of the game-changing iPhone in 2007.

These affordances are visual clues that you can do something, interact. Note Jeff Veen’s frustration with a product which seems to reject affordances in favour of aesthetics — the Mitsubishi Gallant. In terms of affordances implying change, cellphones have them and iPods don’t. Could it be that simple?

OK, it could be that the iPod is so fundamentally different to a mobile phone or a digital camera that it isn’t quite that simple. A comment by Andrew Barnett to an earlier post on this topic at cityofsound noted that:

“Tightly integrated (coupled) design is exactly necessary for disruptive technologies at the start, simply because performance of the technology is not yet good enough and tight integration offers maximum performance, at the expense of flexibility. Once performance, through development, reaches “good enough” levels, architectures shift to a more decoupled, component-based model.”

This may be so. Further, the iPod is basically a miniaturised hard drive and therefore has quite different constraints to work within as compared to, say, the removable storage and battery possibilities afforded within a mobile phone or digital camera. A removable battery cover may jolt a hard drive unacceptably when jogging, albeit imperceptibly to the user. However, Apple have considerable experience of designing removable battery covers from their excellent laptops, which also have to be robust and portable, albeit at larger size and not usually used whilst jogging. The battery compartment on the PowerBook does the job of both enabling the easy replacement of batteries and effortlessly indicating to the user how one might go about it. Hence the famously slim Apple ‘user manuals’. Their products are so well designed in the first place that much interaction is self-evident.

One of my team’s software engineers described how he’d actually replaced an iPod battery — “You get sent some spiky sticks by Apple, you shove them in the top, kinda near the FireWire port, and then you crack it open like a lobster!”. My less technically-bold colleague and I exchanged fearful looks — that didn’t sound straightforward. That just didn’t sound like the right thing to do to my lovely iPod. It sounded more like a physical attack than any normal resuscitating operation. There were no affordances there—shove the stick where?!—and it seemed a potentially destructive strategy in itself.

It’s hugely reminiscent of the non-adaptive architecture that Brand wrote about in How Buildings Learn, with tucked-away ‘service layers’ in buildings such as the IM Pei-designed MIT Media Lab.

“Getting new cabling through the interior concrete walls — a necessity in such a laboratory — requires bringing in jackhammers.”—Stewart Brand

The shock isn’t that you can’t necessarily replace the battery at all — as many have made clear, at great length, you can — but that the design of the iPod suggests you don’t need to worry about this. Then, when you realise you do, you’re helpless and reliant on either after-sales service pushing you into buying a new iPod, or sending your device off to strangers. Or treating your beloved like a lobster.

Because the product wasn’t designed to look and feel as if it could change, the product itself appeared to be perfect, impervious to the ravages of time. Time in this instance manifests itself in the inevitable dying of the light in the battery. The product suffers this alright. After 18 months, your iPod’s functionality — the screen, the tunes, your lived experience with your music, your ratings, your playcounts — fades away in your hands. You are left with a seemingly hollowed-out husk where once a great product lay, unable to breathe life back into your iPod, unable to see the battery, unable to see how you’d replace it. No matter how many times you rotate the thing in your hands, there is simply no way in — and so there’s no easy way out for the festering, dying battery.

Apple provided no clues that you might want—indeed need—to change some physical aspects of the device too. Hence the shock when that display begins to fade like a ghost and just doesn’t come back.

Learning from software

Apple did build in some ability for the iPod to deal with the progression of time. For instance, its software is supremely adaptable, via firmware upgrades. In this way, the software functions could adapt almost effortlessly given the brilliantly rationalised physical interface of the dial/buttons. This is refined further with the iPod mini, but still essentially built around 4 or 5 actions. Change is built into the device on some level.

This is the adaptive design principle of separate layers changing at different rates. Apple are able to change the ‘space plan’, ‘stuff’, and even some of the ‘services’, without having to touch the ‘structure’, ‘skin’, or ‘site. This works.

One expects an ability to change software due the fact that the iPod is designed to be part of a system involving your computer. You can change the tracks on the device and you can download software because there’s a wire connecting it to a computer which is connected to the Internet — it even shows up like a disk on your desktop, and we know they are changeable. The user sees those clues—the Firewire cable, the iPod icon on the desktop—and figures that they can change things.

Ironically, compared to the largely deserved adulation for Ive plus team, Apple’s software designers take a lot of flak, not least from their former colleagues where criticism is fierce even when constructive. But leave aside the aesthetics of those candy-coloured window buttons, if you can, and the system design of Mac OSX exhibits advanced behavioural features.

Steven Johnson has noted how Mac OSX is proving the value of adaptation, how a flurry of incremental improvements can be effected through a modular approach; an architecture so flexible as to enable the differing layers of change fundamental to adaptation.

“Mac OS X changes so significantly and improves so rapidly because it can. Whole subsystems are refactored, recoded, or resynchronized with work done elsewhere (e.g. FreeBSD, XFree86, OpenSSL, etc.)”—Steven Johnson

While software can be adaptable in a way that today’s physical products can’t be, this is another pointer to the benefit of an adaptive design, from within 1 Infinite Loop itself.

Whilst talking about systems changing over time, it’s also worth noting another overlooked aspect of the now-infamous iPod’s Dirty Secret movie — the affable after-sales care which suggested replacement rather than modification.

With the notion of user experience design, the design of a system should extend to thinking through the after-sales service. Again, the importance of looking at the device as a social production — whose real life begins, not ends, at the point-of-sale. Apple are generally perceived to be not exactly great in this respect, though that’s not exactly unusual for a computer manufacturer.

So, whilst I acknowledge their general understanding that design is not just about aesthetics, and that the iPod’s simplicity and fit-to-purpose is evidently part of its success, Apple still seem to be failing on two fronts: the broader user experience, and then the lack of adaptation within their product design strategy. The resulting backlash must be frustrating, but actually, the stakes may be somewhat higher.

It’s what sets us apart…

Not involving the user in the ongoing design and life of the product is not only about ensuring ease-of-use, but more fundamentally solving the problems so successfully for the users that they have nowhere to go, nothing to learn. Again, the likes of Tom Moran and Stewart Brand would probably hear echoes of buildings designed to be “just so” — that then require jackhammers to rewire an office. Whereas Don Norman hears a generation of US kids unwilling and unable to understand technology.

A central cultural theme here is of a user that wants to engage, to be part of the design process. At one end of the spectrum are the descendants of the PC hobbyists, or their analogues in the ‘modding’ community of massively multiplayer online games as described by JC Herz. At the other, the everyday design strategies carried out by all of us, from organising your MyYahoo page layout to organising your shed.

In terms of user interface, this broad cultural theme could actually imply that the development of the graphical user interface (GUI) involves a gross simplification — see Neal Stephenson’s notion of today’s Morlocks and Eloi. The GUI draws many new users in, but does it eventually create a level of abstraction too far? Would an interface which enables a steady development curve towards a fluid, powerful command line-based interface be more effective over time? Would the user understand systems better that way? Actually actively engage in building a deeper relationship with the product?

In Brand’s How Buildings Learn, the musician and artist Brian Eno wrote:

“We are convinced by things that show an internal complexity, that show the traces of an interesting evolution. Those signs tell us that we might be rewarded if we accord it our trust. An important aspect of design is the degree to which the object involves you in its own completion. Some work invites you into itself by not offering a finished, glossy, one-reading-only surface. This is what makes old buildings interesting to me. I think that humans have a taste for things that show that they have been through a process of evolution, but which also show they are still part of one. They are not dead yet.”

So, Apple have produced a series of beautifully near-perfect products — honed, polished, focused, designed to be “one-reading-only”. For this, their team is rightly lauded. However, as a result of this near-perfection, the effect of each slip will be grossly amplified, as each product feels so finished that users don’t think they can adapt; they don’t feel part of a process of evolution. Apple need to find a way of retaining their brilliant pursuit of simplicity and attention to detail whilst also enabling change in their products — visibly enabling adaptation and seeing design as an experiential process after point of sale. In this way, they’d not only engage a generation who need a way in to technology, but nurture a generation who want to get involved in the product, modifying it over time.

This points to adaptive design as a strategy potentially running across most of what we do, as a broad cultural theme.

• With the design of user interfaces, we must surely consider enabling interfaces to become complex at the behest of the user, such that they can keep on adapting the ‘fast layers’ of interface, just as they design their everyday surroundings.
• With the design of social software products, the movement of thousands of people through those spaces should erode, define, and form those environments just as the sea shapes the shoreline — as long as the ability of the product to change is built in and signposted to be part of the deal.
• As physical products and ubiquitous computing become further embedded in what it means to be us — in our communication and identity at every level — we must surely keep on striving for enabling adaptation. And since adaptation may be what largely sets us apart, adaptive design is the only design strategy which will enable these increasingly sophisticated devices to remain truly meaningful.

---

This article was originally published on 14 February 2004, and was originally commissioned by Core77. 

---

John Naughton wrote about the iPod battery issue, referring to this article, for his column in The Observer newspaper:

“These reflections are prompted by a fascinating essay by Dan Hill, a BBC designer who runs one of the most interesting weblogs, www.cityofsound.com. Apple has, Hill argues, a design strategy that pursues perfection. ‘When this works’, he writes, ‘and it has often approached perfection since Jonathan Ive took over the industrial design team at Apple — it creates a community of users who will not just consume, but actively fight its corner.’ The flip side of this loyalty, however, is that enthusiasts feel betrayed if the design turns out to embody an avoidable flaw — which is why iPod owners are so aggrieved. And this illuminates a deeper paradox, namely the tension between the quest by designers for ‘perfection’ and the fact that users have an unruly desire to modify things. This is, as Hill points out, a profound need. It’s why people are forever extending and modifying their houses and, in the process, usually undermining the architect’s original vision. There are some lovely photographs of what people living in houses designed by Le Corbusier have done to them. All of which makes one wonder if Steve Jobs missed his vocation.”—John Naughton, ‘Tinker, tailor — but not with an iPod’, The Observer, 15 February 2004