Mobile: The State of the Art

Posted by
“…it will certainly become more and more commonplace to see people transacting with their phones at arm’s length instead of speaking into them.”A few months ago, I was on my way home from the San Francisco Airport in one of those door-to-door shuttle vans. At some point I casually noticed the woman beside me thumbing the keys of her mobile phone. Dialing a number, I assumed—a common enough sight. But then she kept dialing and dialing and dialing.

I had just spent the previous three months in Germany working on a mobile messaging application. While there, I had seen many a teenager and young professional similarly engaged with a “handy” and I had become savvy enough about these things to know that the Germans were sending text messages to each other, whereas my neighbor on the shuttle van was surely just having some kind of problem with her phone. Short Message Service (SMS) is so popular in Europe, they’ve turned the acronym into a verb, but on this side of the Atlantic, our “cell phones” are for talking.

To make a long story short, my professional curiosity got the better of me. I asked, and it turned out she was indeed composing a text message to send to her friend. SMS had apparently hit the States while I was away.

In Europe, non-voice (mostly messaging-related) services account for around 10% of mobile operator revenue. The figure is much higher in Japan, where many mobile phones are full-fledged multimedia devices, featuring color displays, integrated digital cameras, and stereophonic ringtones. The U.S. market is a different story, but things are changing.

The popularity of SMS is not likely to explode here like it did in Europe and Asia. But as more powerful devices and better services become available, it will certainly become more and more commonplace to see people transacting with their phones at arm’s length instead of speaking into them. Mobile devices already represent a significant channel—and they will no doubt become the primary channel—for a number of common human-computer interactions. These interactions often take place during brief pauses in transit, in distracting environments, on devices that are difficult to use, and where mistakes can be expensive. Obviously, a highly usable interface is key, and there is a growing demand for IAs who specialize in mobile.

But try to find a good book on the subject.

The world of mobile phones is a jungle of proprietary technologies with few established standards that, in some ways, resembles the early days of personal computing. I intend in this article to paint a kind of impressionistic landscape of this world; to present a survey of the markets, technologies, devices, and key applications, along with some examples of successes and failures, a glimpse of the near future, and some thoughts on what all of this might mean for IAs.

Key markets

Nowhere has the mobile phone industry burgeoned like it has in Japan. With mobile phones, as with other things electronic, the Japanese have lived up to their reputation for embracing new gadgets as quickly as manufacturers can conceive of them. It remains to be seen whether Japan is a year ahead of the rest of the world or simply a unique market. For the moment, it is safe to assume both are true. The proven success of Japan’s most popular mobile services seems to promise their broader appeal, but the multitude of niche offerings is largely ignored by the rest of the world.

It is worth noting that in their latest generation, Japanese phones are actually a little bit bigger than their predecessors, marking the first reversal of what has been a steady trend toward smallness. The demand for processing power seems to have subjugated the demand for shirt-pocket-sized convenience. Japanese mobile phones double as portable game consoles, music players, and cameras, among other things. They employ an always-on network connection, with data speeds roughly equivalent to dialup modems, so users can download small pictures, sound files, applets, and games cheaply, quickly, and easily.


The Generations of Wireless

1(st) G(eneration): The analog radio cellular phones that first appeared in the 1970s

2G: Digital voice encoding introduced

2.5G: Increased bandwidth; packet routing

3G: Broadband data speeds; global roaming; enhanced multimedia

Europe has the highest average mobile phone penetration rate in the world, and it’s not uncommon for Europeans to own more than one mobile. European mobiles use removable SIM cards – the small chips inside the phones that store the subscribers’ personal information (phone number, contacts, saved text messages, etc.) and identify the subscribers on the network. The European SIM card is universal, meaning it’s easy to remove a SIM from one phone and insert it into another. Therefore, it’s easy for Europeans to upgrade to a new phone or own a whole drawer full of them.

With non-voice services, Europe is following Japan’s example. Operators are scrambling to introduce new color devices and accompanying services to run on their relatively new 2.5G infrastructures. “i-Mode,” the packet-based service for mobile phones offered by Japan’s leader in wireless technology, NTT DoCoMo, has had recent launches in Germany, the Netherlands, and Spain, and European operators are touting Multimedia Messaging Service (MMS) as the next big thing.

The United States
The United States has lagged behind, but the latest service offerings from the biggest American providers suggest the gap is closing—the technology gap, that is. Adoption rates in the U.S. are a different story. The mobile phone penetration rate here is about 45% (compared to about 75% in Europe and 65% in Japan). American consumers have not embraced mobile like their Japanese or European counterparts for a variety of reasons.

U.S. providers, however, are boldly charging forward. AT&T will reportedly offer i-Mode to its customers this year (NTT DoCoMo holds a 15% stake in AT&T wireless), and Sprint recently launched their “PCS Vision” service, which includes color browsing, downloadable stereophonic ringtones, and MMS.

Mobile technologies

This is where things get messy. There is little in the way of consistency at any level. Developers are faced with a huge number of unique client devices running any of the huge number of proprietary operating systems and integrated browsers, supporting any of a handful of development technologies and markup languages, and communicating with the network via any of several digital data transmission standards.

From the bottom up then…

Transmission standards
The selection of transmission technologies has been somewhat regional. The actual mess of acronyms doesn’t warrant a detailed discussion here, but suffice it to say that when it was time to migrate from analog to digital, Europe took a consensus approach. They chose a standard called Global System for Mobile Communication (GSM) to cover the continent. Japan too, allowing politics to intervene, chose a national standard.

In typical fashion, however, the United States decided to let the market drive the decision, resulting here in the deployment of a mishmash of semi-compatible standards. For voice calling, this is not an issue, but transmission of other data across the various standards has been hindered by a number of roadblocks.

Application development technologies and markup languages
With the exception of DoCoMo’s i-Mode, which uses a language called CHTML (Compact HTML—essentially just what it sounds like), WML is the markup language of choice for the wireless web. WML is simply an XML Document Type specific to mobile devices. HDML, the predecessor of WML, is still mentioned occasionally, but for all intents and purposes it is obsolete. The version history of WML can be a little confusing, especially since WML (the language) and WAP (the protocol) are often used interchangeably in the context of version support (e.g., “device A supports WAP/WML version x.x”).

Significantly, XHTML, XSLT, and even Flash have been gaining support (although Flash more slowly), and many new devices will render a familiar range of image formats. This means that from a technical standpoint, developing for mobile phones will become more and more like any other web development, so the burden will be on IAs to design channel-appropriate interfaces.

One other mobile development technology bears mentioning: Java. Sun created a slimmed-down version of its language and called it J2ME. It is designed to accommodate the limited computing power of mobile devices and allow them to run small, self-contained applications. Computing power isn’t always the only limitation to be accommodated, however. Bandwidth, as well, is an issue for applications that are to be delivered for over-the-air downloading.

Operating systems and browsers
The mobile world is in the midst of its own browser wars, and most often a device’s built-in browser is tightly integrated with its operating system. The biggest players are Nokia/Symbian, Ericsson and Openwave, although Microsoft has recently begun to move into the wireless space.

Nokia, the leading handset manufacturer, has recently begun to peddle a productized version of its software to other device manufacturers. Alternatively, Openwave, whose main business is software, has seen their browser installed in a wide range of handsets. That, however, has far from guaranteed any kind of consistency. Openwave’s software has been deeply customized for certain manufacturers, and there are even different customizations of the software for different handsets by the same manufacturer.

This means the same markup is rendered differently on different devices. It also means the interaction between the hardware, the software, and the remote application—the physical mapping of the phone’s keys to the application’s functions—cannot easily be predicted or specified.

The range of devices on the market continues to expand, with new devices being introduced much more rapidly than old devices are being retired. The best way to make some sense of it all is to take a zoological approach, to impose a classification system on the multitude of species.

There are two useful facets of such a system: degree of mobility and amount of computing power. Focusing on mobile phones, I divide these into two classes. The more common, and therefore more familiar, of these is the set of monochrome data-capable phones. The other class is the set of more powerful phones with large, full-color displays. My chosen differentiators in this case are rendering capability and navigation method (as expressions of computing power).

This classification system obviously has its limitations. Some monochrome phones, for example, support Java, and some don’t. Some color-capable phones don’t support four-way scrolling. And there are always anomalous devices that defy easy classification altogether, like the Handspring Treo 180—a powerful “Smartphone” that happens to have a large monochrome screen.

Key applications

At the moment, a common perception is that mobile computing is little more than a poor imitation of desktop computing. Critics wonder why anyone who has access to a computer would bother to agonize their way though an m-commerce (ecommerce on a mobile device) transaction. The simple answer is: they wouldn’t.

The mobile applications most likely to succeed will be those that take advantage of their mobile-ness. I have mentioned i-Mode as an example of an application that has succeeded, but it’s useful to focus on the broader categories that this example and others represent.

By far, the most successful non-voice application in the mobile world is SMS. Remarkably, it was originally conceived not as a consumer product but as a way for mobile service providers to send data—anything from promotional messages to technology upgrades and patches—to their subscribers. These subscribers quickly embraced it as an inexpensive way to send short messages (originally 160 bytes maximum, at about 15 to 25 cents each) from mobile to mobile. According to the GSM Foundation, Europeans send as many as a billion text messages every day (compared to 12 million in the United States).

More recently, various enhanced messaging services are gaining popularity, including different mobile implementations of popular Instant Messaging services like Yahoo! Messenger and AOL Instant Messenger, and Multimedia Messaging services.

Outside Japan, browsing content via the wireless Web has arguably flopped. “WAP is crap” goes the saying. However, the introduction of new color devices and the rollout of higher-speed networks have brought renewed hope for the future of mobile browsing in general. Most people believe that the browsing applications most likely to succeed are those that provide targeted, on-demand information (e.g., sports scores, stock quotes, and weather reports) quickly and easily, and obvious and immediate utility (e.g., travel and event booking, auction bidding, and gambling).

Research has shown that people who use their phones for non-voice applications often do so as a way of killing time while commuting, for example, or waiting in line. Games provide an ideal distraction. Some amazingly simple games have been a hit with mobile phone users, demonstrating that people who expect their PCs to immerse them in minutely-rendered 3D worlds are nonetheless willing to spend 15 minutes a day playing “Snake” while they ride the bus.

Games can be delivered in several ways. They can ship with the phone as built-in applications; they can reside on the network to be played during active sessions; or they can be delivered as complete applications via one-time over-the-air downloads.

Personal information management (PIM)
Most phones on the market today include a suite of built-in PIM applications such as an address book and calendar. Some phones also include email and synchronization support for Outlook or other PC clients, and WAP (Wireless Application Protocol) portals like Yahoo! and MSN provide mobile support for their popular Webmail clients, as well as POP support. Mobile PIM applications show special promise for enterprises looking to support a mobile workforce, and PIM applications are primary candidates for full, frequent multichannel use.

Location awareness
Location awareness is an application enhancement, not an application category. The architecture of the mobile telecom environment makes subscribers locatable geographically, though not with GPS-like precision. Operators are adding location features to messaging services and games, as well as to more utilitarian applications like restaurant and club finders. Obviously, privacy protection is a key concern for services that incorporate locatability.

The Role of the IA

All the basic tenets of our profession certainly apply to the discipline of mobile user interface design, but these underlie a number of unique considerations.

Mobile usage patterns are distinctly different from what we associate with the desktop PC. Mobile sessions often occur in public places, during brief pauses. Unless the user is idly browsing or playing a game to pass the time, she is probably seeking a piece of very specific information or trying to accomplish a single very specific task. Time is often—literally—money, so there is effectively no margin of error. If the user navigates down the wrong path or downloads the wrong file, she pays for the mistake.

IAs obviously need to understand the contexts within which a given application will be used, to understand the physical environments, the motives and circumstances, and the target devices. There are many questions that apply especially to mobiles: Will the user be moving or standing still? Will he be operating the device with one hand or two? What are the most likely distractions or obstacles? What if the connection is dropped?

It is important to remember that in many cases, users’ attention will be divided. They will interact with an application while walking down a flight of stairs or while half listening for their flight number to be called.

Because of the variety of device capabilities currently in use, IAs must frequently decide whether to exploit the advantages of a given device or to design something more generic to accommodate a broader range of devices. Screen sizes on mobile phones range from the tiny to the miniscule, so IAs must abandon notions of point-and-click in favor of click-and-flow.

Since users are presented with so little information at any given point, it becomes especially important for them to know where they are within the system (and where they were, and where they can go). Wireless data speeds are usually equivalent to a dialup connection or slower, and devices have very little storage capacity or processing power. Finally, users don’t have the luxury of familiar input devices like a mouse or alphabetical keyboard, and many users are only roughly familiar with the behavioral quirks of their chosen clients.

Most mobile phones currently in use support only one-color graphics. This severely limits visual branding opportunities, and while it may be possible to an extent to brand interaction design, it is more important to stick to familiar user interface conventions and metaphors as much as possible. Users are likely to encounter more than enough uncertainty without our help. We don’t need to create more uncertainty in the pursuit of distinctiveness or innovation.

We declare all the time that less is more. With mobile phones, one would think perhaps we don’t have a choice. Even so, the maxim applies. The simplest interfaces are the most successful. Wizards, for example, generally work better than forms because of their one-step-at-a-time simplicity. A login process requiring a username and password, then, should be a three-step, three-screen process (1. username 2. password 3. submit). On the other hand, perhaps such extreme simplification would be maddening to power users. There’s only one way to find out…

Test. Conducting usability tests on mobile applications is difficult. There are few software or hardware tools designed for testing mobile phones, and there are few documented guidelines or best practices. But that’s also part of what makes it exciting. As with all frontiers, we are required to imagine, to innovate. I worked quite a bit with a firm that used an awkward-looking setup involving a miniature spy camera and duct tape, but it gave us exactly what we needed.

Any of the points above could of course warrant at least an article all its own. I look forward to the opportunity to discuss in much greater detail some of the particulars of mobile user interface design.Acronym Soup

ARPU Average Revenue Per User
CDMA Code Division Multiple Access (a digital voice encoding format)
CDMA-2000 The broadband CDMA standard developed by Quaalcom and Lucent for 3G
GPRS General Packet Radio Service (a packet-switching protocol designed to improve data speeds on GSM networks)
GSM Global System for Mobile (the most common worldwide mobile communications standard)
HDML Handheld Device Markup Language
J2ME Java 2, Micro Edition
LBS Location Based Services
MMS Multimedia Messaging Service (mobile-to-mobile transmission of images, video, sound)
OTA Over-the-air
PCS Personal Communications Services
SIM Subscriber Identity Module (a sometimes removable microchip that stores a subscriber’s personal data and the information necessary to identify the subscriber on the mobile network)
SMS Short Message Service
T9 Text input on nine keys (a text-input helper application that employs a database of commonly-used words)
UMTS Universal Mobile Telecommunications System (a 3G transmission standard)
WAP Wireless Application Protocol
W-CDMA Wideband CDMA (a 3G standard)
WML Wireless Markup Language

Standards organizations

Developer sites

Other resources

Shawn Smith has worked as an IA and user experience designer since 1996. Currently he develops applications and UI standards for Vodafone, the world’s largest mobile operating company.


  1. Excellent overview of a complex domain. I’m curious: What’s your take on why market penetration has stayed relatively low in the U.S.? Are competing standards entirely to blame, or is there something cultural going on?

  2. There are two books that are worth mentioning when it comes to mobile interaction design/IA/usability.

    Information Appliances and Beyond. Eric Bergman ed.
    Has a couple of detailed case studies on the design of Nokia and Palm interfaces that deal with contextual design issues and interface trade-offs: a really good resource.

    Handheld Usability. Scott Weiss.
    Haven’t read this expensive tome, but it sounds interesting, but likely too focused on rapidly changing standards.

  3. At the last minute, I had to trim some of my thoughts on market penetration out of the article to keep it under the maximum length, but it’s worth discussing.

    Essentially, there are two factors that have slowed the market penetration rate in the US. First and most significantly, the mishmash of transmission standards chosen by US operating companies meant that in many ways they couldn’t share the burden of infrastructure development. Instead, they had to build large, expensive networks almost entirely independently of each other. As a result, Wireless coverage in the US is weaker than it is in Europe, especially in rural areas. Coverage is still a prime differentiator for customers shopping for cellular service in the US, which is not the case in Europe or Japan.

    Secondly, and this has more to do with data transmission than voice, PC penetration is much higher in the US than it is in Europe or Japan, so there is not the same market here for high-end mobile phones with all kinds of data capabilities.

  4. I have been doing some casual research in the UK, Norway and Finland looking at the patterns and interaction flows within different handsets, and handset modes.

    A key observation is that when connecting to a WAP service a Nokia phone becomes highly unusable – the WML encoded sites cannot make proper use of the keypad or other buttons.

    For example the ‘soft keys’ under the screen usually act like back and forward buttons in a browser; taking the user in and out of menu items for normal phone functions. But in ‘WAP services’ mode they are can be reversed or become entirely redundant.

    It makes me a little depressed to read the end of your article, thinking about the dozens of handset designs currently in development for 2.5 & 3G use, and wondering how many site versions we are going to have to develop in order to take advantage of these different interaction possibilities, or to just make them barely usable… This model also seems to put far too much power in the hands of developers who will constantly have to reinvent the wheel with inconsistent interfaces.

    Can there not be a standard for hardware interfaces as there are for WML stacks or xhtml divs?

    On positive note: on a recent trip to Tokyo I found the best features of iMode phones to be the *heart* and the *car* symbols, for ‘courtesy mode’ and ‘driving mode’ respectively. The ability to turn off all sounds or divert all incoming calls at the press of one button seems like such good contextual design.

  5. It appears you have done some extensive research. However I would ask you comments on the possibilty of a newer idea in the using of standard webpages via a mobile device. Instead of creating seperate copies of wesites (HTML) for device types. Would it be beneficial to have a fully engage HTML parser that only displayed usable content for that device? This is a newer concept from what I can tell is being introduced via smaller fish in the market such as and parts of Ericcson. Do you think it might stick around? How does similar areas such as Japan, europe handle such? Thanks,

  6. As in the early days of web development (remember Netscape Vs. Internet Explorer?), we’re facing a new two-contendents match: Microsoft Vs. the rest of the world (now embodied by Symbian OS inside Nokia and Ericsson mobile devices).

    The main issue is the lack of CSS and XHTML support by Microsoft Pocket Internet Explorer for Smartphones and Pocket PCs. Obviously we can still operate server-side transformation with XSL, but the IA suffers: we should finally deal with a consolidated workflow, and again this is not possible, although the W3C efforts to establish standards in this chaos.

    Our only possible commitment will be to design consistently across devices, keeping in mind on one hand the two or three different technical implementations, and on the other hand the digital identity of the project, making large use of liquid design. Otherwise we will loose our sleep chasing all the esagonal screens that mobile industry will provide…

  7. There are several philosophies that can guide the development of applications and content for different devices…

    The comments so far have focused on the maddening challenge of developing device-specific “alternative” versions of what we develop for the PC. There are three basic approaches I have used:

    1. Custom coding – Custom coding for each device class (and sometimes for specific devices) is the most difficult and time-consuming approach, but it offers the most control. It’s analagous to coding for a wide range of PC browser versions, but it goes a few steps further. Developers have to accommodate a high degree of both software and hardware variation.

    2. Transcoding – At the other end of the scale is transcoding, an automated process by which web markup and scripts are parsed and processed into WML or other device specific markup. Developers have very little control, and since the transcoding engine doesn’t *know* anything about the content, usable and useful output is far from guaranteed.

    3. Recently, a few developer tools have appeared, which automate parts of the process and allow control over other parts. One I have begun to use requires a developer to translate HTML (or other) markup into a proprietary XML-based markup that essentially makes every element an object. Then, using a simple GUI that references the markup and a library of objects defined by the markup, the developer can build custom layouts per device or device class.

    So far I have been satisfied with this tool, but it amounts to a workaround. At some point the mobile world will settle into some sanity and our lives will be made easier by established standards and more powerful development tools.

  8. In regards to SMS in Singapore, Its a big hit, although the interface is hard to use. I would however say that its very popular only to the younger generation. Teenagers to young adults. I don’t see SMS being used by the older generation.(This is my opinion on observation for many and not validated research.)

    These are the reasons as to why its a big hit in Singapore (In my opinion):

    Motivation: Teenagers and Young adults per se love talking on the phone to Instant Messaging and now you can carry on this using SMS while mobile.

    The love for communication between friends no matter where you are. The motivation to learn the hard to use interface through trial and error so that you can chat while you are in the bus/ MRT ( sorta like the BART in San Francisco) and whereever its inconvenient to talk on the phone.
    Also, when you do not want to converse with the person but rather just inform the person a specific comment, for example, “meet me at 2p.m.”, if they called the person, it might lead ot a conversation and SMS solves this problem. The motivation factor is a great reason as to why people learn to use these hard to use interfaces.

    Price: The cost of sending SMS messages are way cheaper than talking on the handphone. This factor is an important factor for the teenagers and young adults whose budget is small or for most people who choose to use SMS.

    Society: The teenagers and the young adults in Singapore have grown up in a society where they are introduced to technology in the early age. Kids are introduced to computers as early as 6 years old and high school students use wireless devices for various purpose and majority of singaporeans have a computer at home. With this influence, they are also introduced to interfaces which are often hard to use. They have come to expect interfaces to be hard in any IT product/service. Although, there has been a slight awkening to need for easy to use IT products and services.

    In Singapore, most people blame themsleves for not be being savvy enough to know how to use these SMS or any other IT products or services. They learn to be “Savvy” enough to use these products/services.

    These are some reasons as to why I believe that although the interface is very hard to use, its popular in Singapore.

  9. Timo wrote:
    > There are two books that are worth mentioning when it comes to mobile interaction design/IA/usability.
    > Handheld Usability. Scott Weiss.
    > Haven’t read this expensive tome, but it sounds interesting, but likely too focused on rapidly changing standards.

    I have read Scott Weiss’s book, and I highly recommend it. He does a good job of including current technology and standards without making them the focus of the book. The focus is solidly on conceptual principles that will persist even when the technology standards have been rendered obsolete.

    He also has an excellent website:

    The book is expensive, but if you are working professionally in the design or evaluation of interfaces for handheld devices, it is a worthwhile investment.

    Disclaimer: I know Scott and think he’s a great guy, so I may not be entirely objective about this.

  10. I think the popularity of SMS is shown here(UK) by the fact that you can send SMS from your PC. I think it’s quite interesting that the SMS thing worked and the WAP thing hasn’t really. SMS negates the idea of checking your email via a mobile phone. The screens are too small aswell.

    Incidently the BBC Radio 4 here are the moment they are having a competition to decide the best and worst invention ever – some one nominated the mobile phone as the worst. You’ll be able to tell when SMS has truly taken off because you’ll hear those irritating beeps of messages being received wherever you go.

  11. Anyone interested in mobile devices should look at the Danger Hiptop / T-Mobile Sidekick. It has a form factor that is much more suited to mobile data communication than any of the phones currently on the market and the web browsing, email, messaging, PIM apps on it are incredibly well designed – very elegant and easy to use. The major handset manufacturers have a lot of learning / catching up to do.

  12. Another book that might be of interest is DoCoMo: Japan’s Wireless Tsunami by John Beck and Mitchell Wade. It’s not a book on usability…more of a boosterish business book (that idolizes the management of the company)…but it does provide some interesting insights here and there on why i-mode is such a hit in Japan.

  13. I cam across this site searching “cost of sending SMS”+discussion (on Google) while working on a project for my business. I would be interested in your critique, comments or questions about what we’re doing with SMS at my link below…

    I love the article. I assume you’ve read “Smart Mobs” by Howard Rheingold? Great stuff as well.

  14. I cam across this site searching “cost of sending SMS”+discussion (on Google) while working on a project for my business. I would be interested in your critique, comments or questions about what we’re doing with SMS at my link below…

    I love the article. I assume you’ve read “Smart Mobs” by Howard Rheingold? Great stuff as well.

Comments are closed.