The 90/90 Rule of Design

Or the 95/95 rule, although 90/90 rolls off the tongue more easily.

We have all heard of the 80/20 Rule a.k.a. the Pareto Principle and which comes under a broader category of relationships called the Power Law. In essence the Principle is a rule-of-thumb that states that for many phenomena, 80% of consequences or events are the result of just 20% of the causes. Thus, 20% of scientists are responsible for 80% of all discoveries, 20% of actors are behind 80% of blockbuster movies, 20% of the members of a team do 80% of the work, or 20% of websites get 80% of the hits. 80% is actually only a rough measure. In an actual instance, it could be more like 95/5 or even 99/1 as in the case of weblogs where less than 1% of blogs get greater than 99% of all visitors to blogsites.

The Pareto Principle may be applied to computer applications such as Microsoft Word or Excel: most users employ less than 5% of the applications' features. The counter argument usually advanced is that the 5% of features employed by User A might be different from the 5% of features required by User B. Further, the same user might require a different set of 5% of the features at different times.

Nevertheless, we may find that 95% of the time User A uses the same 5% of an application's features. Extending this logic, if we look at what 95% of all users need 95% of the time, we would discover that most people can get by with a relative small set of features. This Son of the 80/20 Rule we shall hereafter call the 90/90 Rule: the features and functions (f&f) employed by 90% of users 90% of the time.

How dow apply this rule to the design of interfaces? As a general principle, the interface should reveal or provide direct, immediate access to, or control of, only 90/90 f&f; all others should be buried one or more layers below the surface -- in other words, the remaining f&f require the user to perform additional. The user, therefore, is not overwhelmed with interface cues and functions that exceed the 7 +/- 2 chunk cognitive capacity of her Short Term Memory and is able to easily focus on a few primary functions. Additional functions are just a click or two away.

A tale of two remotes

The remotes in question are those from Apple used with its Front Row software, and from Microsft, used with its Media Center PC. In size and complexity they occupy opposite ends of a spectrum.

The Apple Remote (AR) in keeping with that corporation's 90/90 philosophy is small and spare - it is about the length of a strip of chewing gum, and not much wider and thicker either. The Microsoft Remote (MSR) in tune with that company's All Things to All People ideology is large and feature rich. It's the size of usual remotes that come with your TV, DVD player or stereo.

MSR looks - and feels like a remote. Physically, it is well designed to fit comfortably into an adult hand, and the key controls are in the right places. Cognitively, however, MSR is just as confusing and infuriating as the average remote.

AR looks nothing like any remote one might have ever encountered. In fact, it looks almost exactly like Apple's own iPod Shuffle DAP. It has the same sort of controls plus a menu button. AR has high learnability; the controls have been carefully designed to be usable across different cultures and languages and the buttons are juxtaposed in a manner that would allow users to easily map onto their own mental models. No user is likely to feel overwhelmed by AR.

To be sure, MSR has far more functionalty than AR; an experienced and trained user can access various functions far more rapidly than with AR. The problem is, most users (other than teenage boys) are unlikely to read the manual or spend enough time to master the remote. More likely, they will learn to use the two or three things they need most and try to commit those functions to muscle memory, so that they can use the remote even in the dark, say while watching a movie.

Apple has moved most of the complexity of the interface to the device itself -- in this case, the Macintosh computer. Additional functions are accessible through a menu that pops on the screen when the menu button is pressed. While this requires more button presses than hitting buttons on the remote, they require far less cognitive effort and the process itself is perceived by the user as being less effortful and more satisfying. The Apple Remote can be learned very quickly and then operated with one hand and without requiring the user to look at the controls on the remote; any additional operations are performed by the user without taking her eyes off the screen.

MSR exposes the complexity of the entire device to the user, while AR masks it under a menu that appears on the screen. The Apple Remote is a classic example of user-centered design that respects and accommodates the cognitive limitations as well as behavior patterns of the user. The Microsoft Remote is a textbook example of traditional device or technology centered design where device functionality and control takes precedence and the user is mostly an afterthought.

Here is another example of the value of simplicity over complexity.

Functionality, Style and Usability

It turns out that Nicolaus Copernicus did more than just stare at the heavens; when he was not peering at heavenly bodies, the man had something to say about money, which the English financier Sir Thomas Gresham later formulated into what is now know as Gresham's Law: Bad money drives out good.

This blogpost has nothing to do with money, although I will paraphrase Gresham thusly: Bad design drives out good. In the mad rush to conquer, occupy or seize new territories, corporations seem to lose perspective. In particular, superficial Style, the aspect most immediately visible to users ends up trumping Usability, a quality that is discerible only through actual use by intended users in real contexts. I am going to use as examples Motorola's new line up of terminally cool four-letter-named cellular phones: RAZR, SLVR, PEBL and others that might be on their way. There's the RAZR (pronounced 'razor'). Then there is the SLVR (no, not 'silver' but 'sliver' as in 'thin'). And then the PEBL (yes, as in 'pebble'). Cute names, even cuter phones. In fact, the phone are smash hits, and every cool person wants to be seen in the company of one of these. I mean, who could resist them? They're infinitely cool. I myself have been salivating over these for a long while; with my frumpy little phone, I don't want to be caught in flagrante uncoolo.

I played with one of these puppies the other day in a store and came to the conclusion that my fingers were much too fat, and my nails were not long enough. Either that, or the phones are meant for squirrels, or anorexic heroin chic models.

Take a close look at the keys. They aren't really buttons. The surface is flat with grooves cut into it and numbers are etched onto the surface. In the SLVR model the number keys aren't even lined up making it look really radical and chic. Would I be able to punch in a number quickly? Not a chance because of two problems: a) I would be constantly hunting for the numbers, and b), I cannot use my thumbs directly because I end up pressing more than one key at a time: I need to use the tips of my fingers or my nail (if it is long enough). For somebody with a lot of time on their hands (and thin fingers) who also wishes to project an image of hipness, these phones are great.

To be sure, the phones, especially the RAZR, are chock full of functions, and are slim and light. And rather tasteful to look at. Unfortunately, style has trumped substance. Rather than focus on usability first, and allowing style to emerge from the process, the designers have gone all out to infuse the product with style. Even more unfortunate is the fact that the products have become a major market success, especially among the young and hip; the message this has sent out to the industry is that style matters most to customers, usability be damned. Products are now flying out from major vendors emphasizing, above all their stylish qualities. According to a study released by the business analysis firm Compete, form matters more than function.

"What we're seeing is that people are willing to pay pay a premium for style over functionality," says a senior analyst at the firm. "Shoppers consistently tell us that what's most valued to them in purchasing a phone, No. 1 is price, and No. 2 is form, shape, and design ... People tell us that how a phone looks is particularly important to them when they purchase. What it does is significant as well, but to be a cool device it has to look cool."

In an ideal world, a focus on style would not be necessary: a functional product, thoughtfully designed for usability would be inherently stylish in a manner that would never go out of style. Take the humble rose, for instance; it's been around longer, probably, than mankind, and yet doesn't feel a bit dated.

Too Zune to tell

Microsoft introduced its Zune platform yesterday. Zune is set to compete with -- and intended to transcend -- Apple's market dominating iPod. Now in hoary Microsoft tradition, Zune is more than just a product -- it is a strategy, a campaign, a music downloading service, a technology platform, and much more.

Microsoft is following the script burned into its DNA: wait to see what kind of technology succeeds in the market, and then come stomping all over it with a similar, or superior product. This has worked with just about every product that the company has introduced, most recently, the Xbox gaming platform. Market leaders Nintendo and Sony were left speechless by the speed with Microsoft came to claim a large chunk of the gaming market. Clearly, Microsoft hopes to do the same with Zune; after all, it did beat Apple's Macintosh platform with Windows -- a poorer and cheaper knockoff. But with Xbox, Microsoft showed that it could master a demanding market with a quality product too.

iPod users constitute a market much like the gaming market -- they are discerning consumers who are willing to pay more for quality (unlike DOS and Windows users who were willing to put up with abysmal technology in return for a lower initial cost). Apple has a dominant share of a market made up of hundreds of competiting products at every price level, most of which are priced well below the iPod and include additional features such FM radio, voice recording, digital camera, PDA functions, games, and no exclusive tie-in to a single music store.

How does the Zune stack up? Bear in mind that the Zune introduction yesterday was a pre-announcement something that Microsoft typically does to scare off competition and pre-empt potential customers from spending their money on competitors' products. Given Microsoft's repeatedly delayed introduction of its "forthcoming" Vista OS release, it is anybody's guess when Zune will really make it to the market. Nevertheless, Microsoft will likely push to release it in time for Christmas.

Which brings us to the current state of Zune. As you can see from the prototypes in the picture above -- those are officially released Microsoft photographs; no independent pictures of the device have been spotted in the wild -- it looks a lot like the iPod. And in grand Microsoft tradition, that was probably intentional, given Microsoft's me-too approach to things. It is physically about the same size as an iPod (albeit somewhat larger and heavier) and comes in black, white, and brown. Yes, brown. Go figure. It sports a 3" screen to iPod's 2.5" and what looks like a copy of iPod's scroll wheel is just that -- looks. In fact it is a four way navigation device with a center button. The device, in fact is built by Toshiba and is a variant of that company's Gigabeat DAP (digital audio player). Here's a picture of the Gigabeat. See the likeness? All they did was to mask the "plus" shaped navigation control of the Gigabeat by using a circular control instead. In my opinion the Gigabeat looks a lot better than the Zune. But Microsoft probably wanted it to look as much like iPod as possible without inviting the wrath of Apple's lawyers.

Zune's claim to fame is the extra features it sports over the iPod: it includes the usual fare of FM radio and voice recording, but in addition, it includes WiFi. Microsoft's plan is let Zuners share their purchased songs wirelessly with other Zuners who have 3 days or 3 trials to purchase the song themselves. Like the iPod, the Zune will be tied into an exclusive music store, and will be incompatible with Microsoft's Playsforsure DRM technology incorporated into its other DAP platform. So Microsoft will be competing with itself, in addition to Apple.

How will Zune fare? I have an instinctive disliking for Microsoft products so I am likely to write it off. But it is always unwise to "misunderestimate" (in the words of our fearless leader) the 800 lb gorilla from Redmond. While it hasn't always succeeded -- at least with versions 1, 2 and 3 of any technology -- Microsoft's deep pockets have shown themselves to be capable of prevailing in the end.

The screens do look nice with a lot of eye candy, but the product looks rather cheap, overall, and brown just sucks, man. What were they thinking? And will people ditch iPod's scroll wheel for something that looks like but doesn't work like one? And the proposed price of $299 for a 30 GB device with a few more features than the repriced $249 30 GB iPod gives one pause. Besides, Bill Gates' eye is off the ball, what with his new mission to save the world. Stay tuned.

Icons, Metaphors and Shopping Carts

Today, a question arose in my Interaction Design class about the difference, if any, between an icon and metaphor as used in an interface, particularly a GUI. Both icons and metaphors rely on visual symbols rather than words to convey the meaning. Here's an example that might help clarify the issue.

Many e-shopping sites employ a shopping cart icon in their interfaces. As you shop around at the site, you ADD items to your shopping cart, just as you would in a physical, bricks-and-mortar store. You may, at any time, review the contents of your shopping cart by clicking on. Likewise you can decide to abandon the shopping cart, or empty it of some or all of its contents, much as you might in the physical world.

The picture or symbol of the of the shopping cart is an icon that represents a real shopping cart. The process of purchasing goods online relies on the metaphor of a shopping cart in that you go about shopping online using a shopping cart in much they way you would in the physical world. While the shopping cart icon represents a specific object using a shopping cart metaphor in an interface would require the designer to create a process that mimics that of a shopper moving around a store pushing a shopping cart. Note that while it is probably a good idea to use a shopping cart icon as part of a shopping cart metaphor, the designer could just as well use some other icon, such as a pickup truck, a bag, or even a horse (as means of carrying goods). Further, the use of a shopping cart icon does not necessitate its use as part of a shopping cart metaphor. It could be merely one of several objects scattered about on a desktop and may possess properties that have no correspondence with real shopping cart -- e.g., a magic flying shopping cart, a shopping cart from hell that attacks people and property, or a shopping cart that grows leaves and bears fruit.

An icon, then, is a symbol, usually for an object, whereas a metaphor describes an entire world, or reality, process that the designer seeks to create as means of helping the user interact successfully with a system. Icons may or may not be employed within a metaphor, although they often are.

You can clarify these concepts for yourself by visiting an e-shopping site such as Amazon.

Some interesting input devices

CNET has been running a news item input devieces that are likely to show up on desktops and notebooks in the near future. Some of them have been developed by Synaptics, who developed and supplied various generations of scroll wheels for the Apple iPod. Here are some interesting devices:

Linux Desktop a pipedream, says former Linux evangelist

Martin Gerard makes quite a persuasive argument -- which is likely to be considered flamebait by hordes of Linux enthusiasts -- that Linux will never become popular on the desktop. He writes,

Linux is an operating system that was designed and implemented from the ground-up by geeks and nerds—let them be "technology enthusiasts" if you think it sounds better or less offensive. My point is that the brain of these people is wired very differently from the brain of the average Joe. Whether this is good or bad is irrelevant; what matters is that they are bound to have different perceptions and values, causing a rift between them.

In order to make the Linux desktop appealing to the average Joe, technology enthusiasts would have to betray their core values to such extends that it would be the operating system's undoing. In other words, they would have to make the phone's case opaque so that we can no longer see its insides, remove all the extra keys (yes, all the cool ones) and strip it of every prominent feature that makes it superior to but more complicated than the average phone. What geek would want to use—or develop—such a boring phone?
...
The truth is that most people [regular folks] cannot visually grasp the entire screen at once, let alone an interface of fifty icons or user controls. In order to be attractive to them, the interface needs to be dumbed down to the average automated bank teller's level. Yes, they find the simplicity attractive, not disconcerting or dull.
...
Geeks spend hundreds of hours tweaking their systems and reading documentation because they love to learn how it works, just like they enjoyed playing with Lego TECHNIC (they would love the MINDSTORMS now) or science fair kits when they were kids. Needless to say the crowd finds these things rather uncool (yes, even the Lego MINDSTORMS). Most people don't like to read the user manual or learn fifty new concepts before getting started. They like when it just works and it's intuitive—at least to them.
...
Get real: the Linux desktop has been designed and implemented by technology enthusiasts, for technology enthusiasts. If they were to seriously try to make it appealing to the masses, the effort would collapse halfway because they would be dismayed by the result. My take is that things are just fine the way they are, and the Linux desktop for Dummies an utopia.

Persusive argument, if you ask me. Then again, you have MacOSX, which, with its Unix underpinnings warms the cockles of the geek heart, and yet is no harder to use than the original Mac ("computers for the rest of us"). The question is, does the Linux community have in its midst the kind of enthusiasts who can build outstanding interfaces that pass every usability test out there? To my mind, the community is outstanding when it comes to adding features and functions and plugging holes, but it unlikely that you can herd together a bunch of geek cats under a charismatic if authoritarian leadership to build a unified interface.

The new iPod Shuffle

Apple is nothing if not bold. Or radical. Or ingenious. Or fickle. Whatever. Except for the one shining current example of the iPod Nano II resurrecting the design of its grandparent iPod mini, every new model that Apple introduces is typically a radical departure from its predecessor. Well, that's not entirely true. The full-sized iPod demonstrates an evolutionary rather than a revolutionary design ethos. I guess they didn't want to mess with success too much. Every bit of change that has the iPod has been subjected to has been subtle and thoughtful. Perhaps there aren't too many things one can change in the design a small and highly successful device, as long as its fundamental functions haven't changed.

The iPod shuffle, however, while its functions have not changed at all, now sports a completely new design. Take a look at the old and new Shuffles. The controls remain identical and spare, dare we say, Scandinavian, look is retained but the rest of the device has undergone massive surgery. Apart from a significant reduction in size - size reduction for its own sake appears to be a driving design principle at Apple -- Shuffle2 now sports an anodized aluminum exterior (suggesting class, compared to the glossy plastic of Shuffle1)and a built-in clip for attaching it to one's clothing, much like a tieclip or brooch. Significantly, Shuffle2 no longer has a USB plug; instead, it syncs with a computer (and recharges its battery) via a dock, like it elder iPod siblings, equipped with what looks like an audio plug -- the same jack is used to plug in the earphones, thereby cutting down on ports and manufacturing costs (and complexity).

What might have prompted this change? One can only hazard a guess. Here are more than one. For one thing, the reduced size probably prevented the incorporation of a USB plug. But why shrink the size and change the appearance of an already tiny device? Here's what I think Steve Jobs had in mind: Shuffle1 looked a lot like a flash memory drive and indeed doubled as one. From a consumer's viewpoint, Shuffle1 was a flash drive that also worked as an MP3, thereby diluting the iPod/mp3 cachet. Further, by inviting comparison with flash drives, consumers might also compare the Shuffle's price with those of flash drives, which are signficantly less expensive. The answer was to redesign it to look (and work) nothing at all like a flash drive. This way, Shuffle2 looks and works like nothing other than itself. The consumer experiences no cognitive conflict, and freed of all fear and doubt, immerses herself in the sweet music flowing through those white earbuds.

Designing Technologies in Context

An interesting thing happened today: Apple introduced the second generation (hereinafter, 2G) iPod Nano. Now let's talk about why this is interesting.

It is early 2004, and the iPod has smashed all manner of records in sales ... but the iPod is relatively high-priced and beyond the reach of a significant section of the potential market. And so, Apple introduces the iPod mini at significantly lower price point. The mini comes in capacities of 2 and 4 GB -- much less than the 30 GB iPod -- but the price is right, and the capacity of the mini is enough to carry around a couple of days' worth of music. The mini is sheathed in brushed aluminum and therefore able to withstand some level of abuse. Given its relative ruggedness (compared to the iPod) and the lower capacity, the mini becomes quite popular with younger users on the go, including some who carry it along for a jog in the park.

The mini, however, suffers from a technical problem -- since it contains a hard drive, it tends to skip when tossed around. Come October 2005, just about 20 months after its introduction, Apple retired the mini and introduced the iPod nano. The nano is created in the image of its older sibling, the ful-sized iPod, and looks just as gorgeous, perhaps even more so. It has a glossy black or white front bezel and a mirror finished stainless steel back. Watch Steve Jobs introducing the Nano. Note how he keeps emphasizing its small size and shows how it fits into the tiny inner pocket of his jeans, suggesting that owners could do the same.

The nano traded the hard drive of the mini for a flash memory drive, thereby eliminating the skipping problem of the mini. The nano could withstand all manner of shocks. Consequently, it was subjected to even greater abuse than the mini. And here's where the problem arose: the glossy, shiny exterior of the nano could not withstand bumps, scratches and similar abuse and began to look old and ragged rather quickly, angering its owners. The nano's beauty faded rapidly. Intriguingly, though the full-sized iPod is sheathed in material identical to the nano, its owners hardly complained.

It turns out that the larger iPod, being heavier and more expensive, was treated with much greater care by its owners and hence aged quite well. The less expensive mini was likely purchased by younger customers who were more likely to carry it around with them and exercise less care. The combination of a different user profile and different use scenario resulted in a similar product being perceived and treated differently, thereby delivering a different user experience. Apple seems to have learned its lesson.

Today, less than a year after its introduction, the company unveiled the second generation iPod nano which has married the dimensions of the first generation nano with the look and feel of the iPod mini. The dimensions of nano2 are almost identical to those of nano1 it is sheathed in the same sturdy, brushed aluminum skin of the mini. Apple's designers are the best in the business, but they too fall prey to seductions of looks thereby missing the importance of context. Fortunately, they learn and correct themselves. Watch Steve Jobs introducing nano2. Note how he focuses on how it is made of aluminum, and how beautiful it is, but not on its size. In its entire history, Apple has never reintroduced a design it has retired; the nano2's imitation of its mini ancestor is a first for the company. The mini was the best selling mp3 player in its time, according to Apple. Perhaps even Apple can learn not to mess with a good thing.

Although the regular iPod and the nano/mini delivered near-identical functionalities, their very physical appearances caused users to perceive -- and therefore use and experience -- them differently. The interaction designer's task goes beyond merely providing features and functions, and indeed, beyond even excellently designed features and functions. The ID's goal is to design experiences and to do this she must gain a deep appreciation of those experiences. This is why the Palm Pilot's designer Jeff Hawkins walked around for weeks with a wooden mockup of his product-to-be, pretending to use it in various relevant contexts. Only by living with his product could he fully understand what it would be.