A few months ago my sweetheart said one of those things that would make any geek start drooling: "I hate getting lost each time I drive into Cologne. Can't we get a TomTom?" I love getting a free ticket to spend obscene amounts of cash on a gadget, without all the weak, post-hoc rationalizing of why it's so useful, which is usually met with something between fury and pity, depending on its price tag, size, number of cables, and overall potential for destroying your living room's visual appearance. But I digress.
For those of you who have been living under a gadget-proof rock for the past few years, a TomTom (mine is a GO 910) is a GPS car-navigation system made by the current market leader of the same name. You stick it to your windshield with a suction cup, tap in your destination address, and off you go, hopefully in the right direction.
It has also become the technology to most profoundly influence my everyday life since my first DSL flat rate in 2001. And that's despite not being a regular driveror maybe just because of that.
Now, I will admit that I got my first mapping-software fix back in '97, when a little-known Dutch software company called Palmtop had just released EnRoute, a route-planning application for my favorite personal computing device of all time, the Psion Series 5 PDA.
But back then, of course, there was no live navigation support. GPS devices were still something supremely geeky, and well beyond the purchasing power of your ordinary computer-science grad student (i.e., me). Geocaching had not even been invented.
Also, in all honesty, only geeks had PDAs back then, so this was definitely not yet affecting the public at large. But we, the bold and fearless early adopters, could explore this strangely empowering new world of geographical information literally at our fingertips. While I hardly used it for the demanding task of live, in-car navigation, it became indispensable to quickly estimate driving times when planning trips, or to simply hide my deep geographical ignorance in a conversation on, say, the wonderful architecture of Barcelona, by discreetly checking which country that was in again.
While it did become possible later to attach a GPS to your PDA (until you realized that multiple loose devices, power adapters, and 200 feet of cable around your dashboard weren't exactly safe, and that setting up took longer than most actual trips), it wasn't until around 2004 that Palmtopnow renamed TomTom!and others started selling all-in-one devices, and live navigation support became a realistic option for the average consumer.
Boy, what a difference. Instead of having to map out each new trip in advance; write down or print out those instructions that as soon as you hit the road you realize are conveniently still sitting on the kitchen table; have your codriver call a friend for instructions on a bad cell phone connection, which he'd then repeat back to you while you're nervously peeking at each new street sign because it could just be the one where you had to make a right (or was it left?)in short, instead of this constant sense of sublime (or not so sublime, depending on the nature of your fellow passengers) tension while driving, you could now focus on traffic and your environment knowing your TomTom would alert you to each turn in time. Even male drivers have been reported to now occasionally have a few brain cycles left to follow what everybody else in the car is chatting about. In other words, here's a complete revolution of your emotional experience of driving somewhere unfamiliar.
Its real potential unfolded for me, though, when we recently moved to San Diego. It's hard to imagine the stress this saves you driving around an unknown city in a different country. It also quickly becomes hard to remember how much of a hassle it all was before. In fact, TomTom offers special computer voices with "I told you we should have taken that exit"-style instructions, should you miss that part of the classic driving experience.
Of course there are still plenty of usability problems that make you scratch your head, wondering just what the designers were thinking. City or street names are listed so close below each other that you keep selecting wrong onesFitts' law at work. I also got a furious call when my sweetheart first tried using it: Köln (Cologne) wasn't in the city list. It turned out TomTom had left out German umlauts from their onscreen keyboards, but forgot to include the standard transcriptions in their search algorithms; unable to type Köln, she'd entered Koeln, but the system was expecting Koln, not even listing the city as a close match otherwise. Dudes, localization.
Oh, and turning it on is a nightmare. Pressing the tiny, half-sunken power button briefly is happily ignored, but keep pressing it a couple times at the wrong moment and it won't turn on at all. Protecting against inadvertent operation is fine, but have these people ever heard of the inherent evil of time-based interactions, or of at least providing appropriate feedback when they're unavoidable? A short "Thank you, starting up, you can let go now" beep would have done the trick. But then again, no one with the slightest case of arthritis in their fingers will ever be able to press that button, so with our aging population the company will soon run out of customers anyway.
But I'm sure these issues will be fixed. My point is that TomTom has crossedno, jumped acrossthe "threshold of indignation," as Paul Saffo put it in Terry Winograd's great book, Bringing Design to Software. The usefulness of the device by far outweighs the remaining awkwardness for a wide range of users and their daily tasks.
Obviously, the entire user experience counts here. You can actually go and buy this thing in a department store today, stick it to your windshield, turn it on, and after making a few obvious(!) choices, enter your first destination and be on your way. This is careful design. Some companies, such as TomTom and Apple, get how important this "first-encounter usability" is, from just the right software default settings, to physical device design, to the printed quickstart, to the design of the packaging. It's no coincidence that for a brief, innocent period, Googling "iPhone porn" actually led to slideshows of devoted users unpacking their new gadget.
So what can we learn from the TomTom story? At some point the mix of features, technical feasibility, and task-centered product, software, and user-interface design came together to shape a product that could make such a radical difference to people's lives that its popularity skyrocketed. Of course this takes years of market research and iterative product development, but it creates a qualitatively new product genre that brings an unprecedented and realistic promise to the market and fulfills it. I call this moment the "sweet-spot" phase.
A telltale sign that a product has reached this stage is that people get its usefulness within 15 seconds of explanation, even though they may not know the technology yet (or even understand it afterward). Non-geeks start telling you about this new thing and begin to evangelize others about it.
Another sweet-spot indicator is that social behavior around the associated tasks changes. These days, when someone gives me driving directionsa sales clerk on the phone, or a friend inviting me to his houseI find myself politely cutting them short, just asking them for their street address, which I then write down and later type into my TomTom.
Clearly, using these devices also has questionable consequences. For one, we quickly begin to rely on them. Usually, after going to a new destination with my TomTom, I still can't go there on my own: There was no need to memorize the route. A more subtle effect is the potential loss of a mental area mapwith a TomTom, you never care to develop a picture of your city as a whole in your head. Will people forget how to describe the way to their home to others? Will real-estate owners bribe TomTom to direct traffic away from their upscale properties? Studying these effects will keep us busy for some time. But even such potentially adverse consequences show the fundamental change that a specific technology can bring about.
Now the bad news: Feature development doesn't stop at its sweet spot. Beyond the idea of providing reliable, easy-to-use directions, TomTom has since added an MP3 player, live updates through the wireless network, connections to "Buddies" (the use of which has escaped me so far), cooperative street updates, photo slide shows (I'm not kidding), and a stream of other features. Some of these are actually useful, but the original TomTom was the sweet spot.
HCI research should focus more on preparingand industry on creatingnew sweet-spot devices, rather than wasting time on baroque extensions of existing paradigms.
David Liddle, design lead for the world's first commercially available GUI computer, explains his theory of technology adoption in Bill Moggridge's wonderful book, Designing Interfaces. He postulates a first, enthusiast phase exploiting the new technology, a second, professional phase putting it to use to get work done, and a third, consumer phase when it becomes available enough for people to enjoy.
I think we should add a fourth stage to this otherwise excellent model: the "baroque phase," in which the successful new consumer product genre is then embellished with secondary features that often already existed before but are now integrated into the new product.
This phase obeys the terrible law of feature creep. Consumers, having experienced the wonderful new possibilities of the initial sweet-spot device, are hoping that subsequent products in this new genre will have an equally revolutionary and additional positive impact on their everyday liveswhich of course they don't, as they're just incremental improvements and so buy new models because of their added features. The resulting featuritis, prevalent in software, is spreading to consumer devices as they are increasingly software-controlled. (Shopping for a new toaster, I recently encountered a model that would assist me in my complex toasting tasks with an informational LCD screen. Please?)
At first sight the sweet spot and the baroque phase seem hard to tell apart: Both give the user new features, just at different levels of originality. But there's an easy test: Sweet-spot products make your life simpler, baroque ones more complex. Sweet-spot products support you in a new way, making a previously difficult or awkward task change fundamentally. Learn just a few new things, and you get an almost magical boost in productivity, simplifying your everyday life. Baroque products just tweak existing processes, trying to make them more efficient in some situations but often complicating other tasks (and sometimes the most frequent onesthink microwave ovens). And to use them, you often need to learn a fair amount of new interaction concepts, operations, and other lingo.
Let's look at some products I consider worthy of a sweet-spot award, and some technologies way in their baroque phase.
Cell phones hit their sweet spot in the mid-'90s: pocketable handsets, with several days of standby and calling charges that didn't ruin the average consumer anymore. What a change! Within years, people moved from carefully planning their evening out to "call us when you're ready; we'll tell you what bar we ended up in." Agreeing when and where to meet, which often failed before, leading to heated arguments over whose fault it was ("But I was looking for you!"), was replaced by the stress-free model of just calling if something came up, no matter where everybody was. The list goes on.
Today cell phones have moved squarely into their baroque stage. In a 2007 study we did for Germany's largest mobile-technology consumer magazine, connect, virtually all models we tested gave users problems with even the most basic and essential tasks: turn on, mute ringer, call number. Being able to browse the Web, take pictures, watch or record movies wherever you are is great, mind you, but it has overloaded the sweet-spot product and interaction design of the traditional mobile phone beyond recovery.
The only way out was to radically rethink the product. Apple's iPhone did that to a degree, removing the keyboard and its dead-end soft-key concept and introducing multitouch to more directly interact with what's on screen. It was far from perfect, but mobile browsing became good enough to become useful, giving you the tingling feeling of a new sweet-spot candidate.
Or take home internet access. After listening to our chirping modems for years, it was DSL's unlimited-time, unlimited-volume flat rate that changed how we thought about the Internet: Suddenly it was free to access the Net after paying a fixed monthly fee. Getting movie showtimes, driving instructions, or just a recipe for cranberry sauce became a snap. And flat rates made our systems always-on, with no dial-in delays. Since then providers have tried to integrate DSL, landline, cable, and cell phone contracts, leading to a maze of options with some further savings but no impact anywhere near that of the flat-rate DSL effect.
Occasionally, consumers will go as far as backpedaling to find the sweet spot again. My last microwaves were all of the one-dial-for-time, one-dial-for-power, go-bing-at-the-end variety, and I can't be alone, judging from what's in stores. On my Sony-Ericsson T630 phone, I quickly replaced the default, distracting, low-contrast ColorBombs theme with a simple black and white one that let me focus on the important stuff.
The desktop metaphor had its sweet spot with the release of the Xerox Star and Apple Macintosh between 1981 and 1984. Since then its basic idea has remained unchanged, as is often lamented, and only small improvements and secondary features have made it onto our screens. Smaller sweet spots were reached within that metaphor (full-text search or Apple's Time Machine backup come to mind), but most new, more colorful and feature-rich systems fall into the baroque phase. Sometimes I fantasize about a system that returns the desktop metaphor to its sweet spot (not that that would be very useful today), or that finds the right revolutionary approach to kick the desktop metaphor out the door.
Other examples include the original iPod, or the affordable consumer digital camera with enough resolution for standard-size prints, letting you take, immediately check, and delete shots for free. TiVos changed TV viewing habits fundamentally, and personally, I would include iChat AV, for letting me show our new kitchen to my mom some years ago, walking around with a laptop and iSight camera (okay, still geeky).
So what gives? For consumer experiences, HCI research should focus more on preparingand industry on creatingnew sweet-spot devices, rather than wasting time on baroque extensions of existing paradigms. To make history, look for that sweet spot providing the broad public with a device/application/service to which they had no usable, affordable access before.
This, by the way, is also why HCI is key to innovative products. Sweet-spot solutions are task-centered in an unprecedented way; they are uncluttered, simple, and elegant.
The other day, after Googling another nearby store on my iPhone, because the one where we were didn't have what we wanted, my sweetheart said, "You know, it's really incredible how useful this iPhone is." Now excuse me while I go and drool some more.
RWTH Aachen University
About the Author
Jan Borchers is a professor of computer science at RWTH Aachen University in Germany, where he heads the Media Computing Group, studying interaction with audio and video streams, mobile devices, and ubicomp environments. He has currently deserted his students for a sabbatical at UC San Diego to write random rants like this. He's not getting paid for this article by TomTom, Apple, or Psion, although he will happily provide them with his banking details should they feel obliged to change that. He can be found at http://hci.rwth-aachen.de.
©2008 ACM 1072-5220/08/0300 $5.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2008 ACM, Inc.