|
|
Main
 
I'll be teaching a "WAP Design" full-day course in London and New York this December.
This course is for designers who want to know how to design for today's phones, using today's mobile browsers. It is filled with examples and exercises geared to ready designers for immediate success.
WAP Design includes XHTML, CSS, and .mobi issues in a designer-focused course covering the following topics:
* WAP Overview
* WAP, including XHTML, CSS, and .mobi
* Successful and Unsuccessful WAP Sites
* Requirements Gathering for Design
* Competitive Analysis of Desktop and Mobile Web Sites
* Mobile Design Patterns
* Expert Design Guidelines
* WAP Design Toolbox
* Wireframing
* Paper Prototyping
More Info
Register for London: 3 December, 2007
Register for New York: 11 December, 2007
 TBWA\Chiat\Day outdid themselves for Nissan's Altima Coupe ad, featured at left. It's beautiful, fascinating, modern, and makes you want to buy a cellphone that looks like a bunch of little cars! I loved it so much that one of my staff got permission from Nissan for me to use the image here on Handheld Usability. I think it indicates how technology can inspire art--just like art can inspire technology. Bang and Olufsen is the firm that most often comes to mind when one thinks of art-inspired technology.
The artwork in the Altima Coupe ad inspired our graphics for the WAP Design course. Molly Bowman created the logos for London & New York.
What is it about the cell phone that inspires art? Is it that the device keeps us in touch, is ever-present in our pocket or purse, and has so many sense-inspiring features? We feel the devices, see them (and their screens), hear them... We don't taste or smell them thankfully, but they satisfy the other three senses beautifully.
It's interesting that touch screens lack that sense of feel--except those that use Immersion's VibeTonz technology, which makes touch screens "feel" like they've been pressed when buttons are activated.
 At Mobile HCI in Singapore in September, Songyee Yoon from SK Telecom keynoted about digital convergence, and service proliferation. She presented the T Interactive solution that you see above. Apologies for the scratchy pictures, which are from my digital camera from the audience.
T Interactive pushes content to the idle screen, with overnight downloads of content, free of charge. The content is then shown throughout the day without requiring further downloads. One example is Cizle, a movie ticketing service that shows advertising for new movies for which customers can buy tickets.
With search, search results are providing a single result, the nearest, instead of a list of many. Searching for Starbucks in their old design provided 169 results on 19 pages. The new design provides a single result with a map to it, with links to other locations.
The "After" solution looks a whole lot better, though it actually looks as if one of the links in the "Before" image was clicked. I would prefer to see a set of information above the map, with links below it. Above the map, I'd list the address and cross street, business hours, and a short review. Below the map, I'd list the phone number (linked for a call), the directions links, and "More search results for 'Starbucks'".
 "Handheld Usability" was obviously the predecessor to "Designing the Mobile User Experience." The shame of it is that Barbara Ballard and Scott Weiss (your humble blogger) did not collaborate to produce a second edition, sharing both of their insights. Instead, Ballard writes the same book, but in 2007 instead of 2002, with many of the same weaknesses, and many of the same strengths. Readers who liked "Handheld Usability" will enjoy "Designing the Mobile User Experience," but will be frustrated by the lack of detailed WAP design advice. Also missing is detailed advice for FlashLite, UIOne, SVG, tat, Java, and other environments. Ballard is a very strong writer, with very strong opinions. Fortunately, she is very smart and knowledgeable, with extensive experience working at Sprint and since providing services to Sprint. Wiley did her a disservice by printing the book in black and white, and does a crazy disservice to readers by charging so much for the book. However, designers who want to learn about mobile will benefit from this book. Experienced mobile designers will be frustrated, as they were with "Handheld Usability." Even experienced mobile designers should consider purchasing this book, as a reference and as a supplement to their own knowledge.
Buy Ballard's Book
 Widgets are small, single-purpose, highly graphical and interactive Internet-enabled applications. Apple's Dashboard for OS X was the commercial pioneer of the widget phenomenon, though MIT's Project Athena was the actual inventor, back in the 1980s. (See the Wikipedia entry on GUI_Widget for more information.)
Widget engines are the technology platforms on which widgets are deployed, the most well-known of which is Widsets, a well-funded Nokia spin-off with no monetization model in sight. While Widsets is available for free download, it only works on the latest smart phones. Furthermore, Widsets requires download and installation procedures likely to baffle most mobile phone users, and which could be costly for mobile phone users without all-you-can-eat data plans.
The one operator-deployed widget solution is Celltop from Alltel, a regional US operator. The left softkey label on the idle screen is "Celltop," which launches the widget engine, reputed to be UI One from Qualcomm. The engine starts with a branded animation and sound effect, lasting several seconds. Celltop emerges, with two side-by-side applications, each laid atop a subtly three-dimensional gradient bubble See http://mycelltop.com for a sexed-up demo of the experience, though you should know that though Celltop looks like the demo, it's not nearly as responsive.
Each Celltop widget is a single data application. Several "cells" ship with a new Celltop handset, including Inbox, Call List, Weather, and different sports, including Rodeo. Celltop has two modes, Navigation, and Application. In Navigation mode, one can slide the display left and right between the subscribed widgets, with an associated "whoosh" sound effect. Pressing the center select button enters Application mode, where the selected widget expands to take up the entire display. Widgets update automatically in either mode, though in Application mode, more data is obviously visible. In an as-yet unreleased Usable Products Company study of Celltop, updates were sluggish, and the data was sometimes stale. The Search cell is not yet available.
Celltop sets a low standard, being the first deployed solution. However, there are non-deployed solutions from several providers, including Nokia's Widsets, Mobidgets, BluePulse, Openwave's MIDAS, Opera Widgets, and the just announced Microsoft spin-off, Zen Zui. All have received some amount of hype.
Come to MAPOS in San Francisco 26 – 28 June, 2007 to learn more about mobile widgets, and attend a very special workshop devoted to widgets that I will be moderating. This workshop is an exciting full-day strategy session that starts by teaching the background, advantages, and challenges of each of the major platforms. It continues with group design sessions for widgets on today's—and tomorrow's phones. It wraps up with a round table discussion on policies and pricing: what should widgets cost, and also their use on a regular basis. The brightest minds in mobility will be there, so please join us! Email me to receive a 25% discount from Informa on the MAPOS US event.
 Snow will be the new screen candy on mobile phones, since there are few options for broadcasting content to phones: place shifting, one-to-one, one-to-many, and RF broadcast. RF broadcast, good old TV without cable, appears to be the most likely distribution mechanism, since with the digital broadcast, only a few simultaneous streams per cell can be transmitted. And with such a tiny device, there's not much of an antenna. Ergo, snow. Presentations from Informa Telecoms, Qualcomm, and Texas Instruments informed this entry.
Mark Burk, a Research Analyst at Informa Telecoms, suggested that handsets for mobile TV will cost $150 more than existing 3G handsets. He said that 43% of the cost to consumers for mobile TV will be the handset for the 18-month period of the typical contract. He also said that consumers were willing to spend $10-15 per month for these services. I ask, what’s Mark been smoking?
Place shifting is the technical name for video podcasting, as users would choose what programs to watch and simply download them to their phones. This choice is a bandwidth constraint that will ultimately fail, unless it’s done through a USB cable or Bluetooth/WiFi to the home or work broadband connection.
One-to-one and one-to-many are cellular transmission of mobile content. This strategy could fail, since cellular sites can at most handle three or four simultaneous broadcasts, even with EVDO Rev. A and HSDPA deployments.
Broadcast using radio frequency (RF) technology will turn mobile phones into television receivers, giving them two radios. This concept makes the most sense, as it doesn’t impact the cellular network in any way. The challenge here is broadcast quality and spectrum allocation, both of which will be tremendous issues. The program guide and other content will stream via the cellular network. I see snow on screens in the future—isn’t that a step backward?
The program guide can also combine the three broadcast methods, not necessarily even indicating which is in use. The biggest usability challenge stated by all of the speakers is the channel change time, which is longer than three seconds.
Paul Scanlan, COO & Co-Founder of MobiTV spoke about the future of mobile TV.
2004: <1 FPS with MJPEG, then 5-7 FPS with MJPEG, GPRS & 1X-RTT
2005: 12-17 FPS, MPEG, GPRS/EDGE/EVDO
2006: 15-30 FPS, MPEG, UMTS/EVDO
I’m looking forward to 2007. MobiTV has a new, sexy program guide (EPG), which has up-sell area above the programming content, which they showed on the Sprint service. While watching content, they scroll up-sell opportunities for ring tones, voting (using premium SMS), and what appears to be sponsorship banner ads.
MobiTV monitors the broadcast traffic, down to each cell site. The Michael Jackson verdict was the most popular day in their history.
MobiTV believes in WiMAX as the future.
Cost Implications of Mobile TV & Video
David Carey, President and CTO of Portelligent gave a fascinating presentation about the cost implications of mobile TV and video on handsets. It seems to be about $60 for the Bill of Materials, though Mark Burk earlier said $150 to the end user. Portelligent creates "tear down" studies of handsets, where they "kill them gently," taking them apart and lovingly photographing every aspect of the design. The biggest costs didn't appear to be the chips, but the complicated double hinging mechanisms for the displays.
There are no written standards for SMS application design, so here are some guidelines. Please comment here or email me your thoughts.
1. Paginate clustered messages. See the Pagination section, below.
2. Answer ‘?’, ‘help’, and ‘info’ messages with helpful information. Including a link to a WAP or HTML web site with more detailed information.
3. Never use blank lines, which might confuse readers when they appear just above the fold. Use dashed lines instead to separate content.
4. When returning an error message, identify the problem, offer steps to correct the problem, and a link to more information (WAP or HTML).
5. Support common misspellings of commands, especially missed characters.
6. Always allow ‘Stop’, ‘Quit’, ‘Cease’, and ‘End’ to be accepted as a command to discontinue sending additional messages, such as to end a subscription service. Upon receipt of a termination command, it is acceptable to send a confirmation message, that may include instructions to restart the same service. However, never require a final confirmation, since the ‘Stop’ message is clear enough.
Pagination
It is not unusual for text message creation clients to support more than 160 7-bit characters, splitting outbound messages into chunks of 160 or fewer characters. Some user interfaces indicate the number of messages that will be sent, but none effectively convey the cost to the sender. Most clients simply indicate the number of characters that have been entered, and the consumer learns how many messages are sent after committing the Send action.
In your own applications that send messages in clusters, it is important to put page header information at the start of each SMS, indicating the current message and the total number. Ending messages that continue with “…” is also a good idea. AIM for SMS begins clustered messages with “FromName: -X/Y- ” and ends messages that continue with “…”
Since SMS delivery is not synchronous, “-2/2-” could arrive before “-1/2-”, potentially confusing users. However, delivery timing is out of the application designer’s control.
Designers have opportunities for creativity with pagination, but keeping it simple conserves characters. AIM for SMS uses seven characters to display “ -1/2- ”. The extra spaces and hyphens (‘-’) took up four out of 160… 2.5% for visual clarity does not seem like a bad tradeoff to me.
The eponymous book for which this site is named has just about sold out! Wiley have contacted me about writing a revision, which I am carefully considering.
Meanwhile, if you have any suggestions, please submit them here. Either based on the original work, or things you think should be included.
I'm currently planning to include several case studies, which will illustrate design areas, such as SMS, photo, music, video, and designing for an older audience.
In Usable Products Company’s Mobile ESPN Usability single-product study, self-professed “sports nut” respondents struggled with the handset’s navigation features, despite enjoying the gorgeous screen and video quality. Mobile ESPN created a service that delighted respondents with its wealth of sports news and statistics, but bewildered them with an inconsistent user interface that proved hard for respondents to follow.
In Q2, 2006, we conducted a suite of six usability tests with mobile phone-using sports fans, whom we recruited in New York City. Each of these people used mobile media services frequently and did things typical of sports enthusiasts: attending events, watching sports on TV, visiting sports web sites, etc. We asked each participant to find leading sports news, stories about boxing, and their favorite sports team’s standings and statistics. We then asked participants to locate and play a sports highlights video, and to set alerts about upcoming game scores. We asked participants what they thought about the handset and service throughout the process.
 In our study, participants struggled with the Sideline, which is a lot like the Orange signature handset UI on steroids. The Sideline is a vertical icon menu, engaged by pressing left on the five-way directional keypad. I’m no fan of these five-way controllers, which are quite hard to use. I’ve watched usability respondents initiate Enter when they want Up or Left, for example. While respondents had little difficulty discovering the Sideline, some respondents had difficulty using it consistently: actions mapped to the direction keys were dependent on the screen state, especially when used with the ESPN Menu, which appears as one of the Sideline elements. 
You can see from the image of the Mobile ESPN UI that there are a lot of interactive and information components. The designers missed the point of higher resolution displays: increased pixel counts should be used to display content more attractively, not to stuff more content into the UI. The text displayed on the Mobile ESPN handset was particularly small. Our respondents were skewed to younger people, who had no problem with the text size.
We asked respondents’ opinions of the handset at the outset of the interview and after they spent nearly an hour using it. Respondents liked the Mobile ESPN handset more after they used it than beforehand, a good sign. Average final reaction was 4.2 out of 5, while initial reaction was 3.7, a 0.5 increase. 5 was equated to “I like it a lot,” while 1 was equated to “I dislike it a lot.” Respondents were especially pleased with the display and video quality, as well as the selection of sports news and statistics available. However, respondents were disappointed by the dearth of video content, slow download speeds (despite being a 3G service on Sprint’s EVDO network), and the lack of clarity of the organization and menu nomenclature. Who knew that sports nuts cared about usability?
At Marek Pawlowski’s MEX (Mobile User Experience) conference this week in London, I saw Zi Corporation’s Qix product, an in-handset interactive search user interface that is compelling. As users type on the numeric keypad, Qix utilizes predictive text analysis to produce optimized search results in the form of sequential menu choices on the phone. Qix differs from T9 and similar predictive text technologies, which display only the best word guess rather than the digits typed. In T9, “84” appears as “Th”, as the predicted combination of one element of each set, (T, U, V) and (G, H, I). T9 has many usability problems, but is well suited to text message entry.
In the example photo for Qix, typing “93” is interpreted as any combination of one element of each of the following two sets: (W, X, Y, Z, 9) and (D, E, F, 3). The result set is displayed in a list, with a logical order: phone numbers, contacts, handset features, optimized results, and then everything else. In the example, a phone number in the contacts list, “+1 33 935 60 68 55” includes “93” as part of the digit string. Next in the list is a contact, “Wendy Crétien”, which contains “WE” as part of the string. Zi optimized the search results further for the “Web Browser” feature, which also contains “We”. I was unable to witness the performance of off-handset searches, which could be frustrating due to data service connection speeds.
As it was demoed to me, Qix optimizes all of the phone’s key features, like the camera, clock, and alarm. They go further to monetize off-handset terms, such as song titles and artist names. The fundamental link to external search is promising, and allows the operator to monetize keywords by prioritizing off-handset results.
Performance in the demo I saw was terrific, on a Nokia Series 60 phone. I’d like to see Qix on a Fastap or other QWERTY phone, but I was quite impressed with the utility and apparent simplicity.
 Game development is purely design-driven. Ultimately, it is this concept-made-reality design that consumers use to interact with the virtual game world. Game design seems easy, but is actually quite demanding and raises many questions. Will players notice a certain important mission-critical object in a room? What if he doesn’t and can not proceed? The player might get frustrated to the point of giving up on the game, and in the extreme circumstance, swear off the brand due to the bad experience.
With skyrocketing game development costs, it is quite evident that the user’s perceived experience of the game’s design can ultimately make or break a game development company. However, very few development studios can afford to run many usability/playability prototype tests on their game concepts before starting development. All too often, such companies find out right before the launch deadline that their concepts are either dysfunctional or just not fun at all.
To combat this rising trend, Giles Schildt, formerly of Steve Jackson Games, is pushing the concept of paper prototyping in game development as a low-cost alternative to costly computer-based prototyping. Gamasutra recently posted an excellent article on a lecture Schildt gave at Austin Community College about the strengths and opportunities of prototyping in general, and how paper prototyping in specific is a very cost-effective strategy for game designers to employ.
At Usable Products, we've used paper prototyping extensively to test out different variations of user interfaces. Scott Weiss introduced me to the concepts and elegance of paper prototyping, and is quite knowledgable about the subject (as one can gather from reading this article). Paper prototyping is ideal for situations (or even mediums) that have to change on the fly, and is often more efficient than drawing in a program than, say, Microsoft Visio. Also, as Schildt is quick to point out, the flexiblity of paper prototyping allows it to adapt to virtually any situation or ruleset, as well as any medium, very quickly. In such cases where changes happen on the fly (usually during the very early conception stages), pen and paper are often more powerful and elegant than the sophisticated computer machine.
Contributed by Nick Hernandez.
[Publication: Mobile ESPN Usability Study]
 We've just completed a usability study of the mobile sports content components of the Mobile ESPN service. The study was a small sample, qualitative usability test meant to explore this new MVNO. We have studied other MVNOs, Virgin Mobile in the Media Download Usability Benchmark and Amp’d is part of the ongoing Video + Music Usability Benchmark study. Since this ESPN study is independent of larger usability benchmarks, we were not restricted to studying the aspects it shared in common with the other services, and we were able to focus on ESPN's unique audience. We had tasks for accessing news and sports statistics as well as setting alerts and watching video clips. Our respondents were real sports fans. Despite the focus on the sporting content and the fan audience, the usability findings were of a wholly different nature and stemmed from the unique set of interactions introduced in this service.
Continue reading "Mobile ESPN: a step off the beaten path" »
 Arlene Harris is the CEO of GreatCall, who market the Jitterbug phone.
Jitterbug is what you see at the left. It's a simple phone with large buttons, large type, and few frills. It's targeted to older people or anyone who wants a basic handset. Arlene explained to me the history behind Jitterbug. In the 1990s, Arlene and several partners founded SOS Wireless with the SOS phone, a three-button emergency phone, pictured below.
The SOS phone had a simulated dial tone, to make it more familiar to people new to mobile phones. Arlene told me an anecdote from a customer who went on hike with some friends. They all brought their mobiles, but Arlene's customer was the only one with an SOS phone. He said that his was the only phone that had any signal, since when he opened it up, he got a dial tone. At the time, the simulated dial tone did not mean that signal was available, but now the Jitterbug phone's dial tone indicates signal availability. Obviously, GreatCall did the right thing.
The GreatCall model is not just to create what looks like a terrific phone. GreatCall plans to offer web-based services for provisioning that are particularly easy to use, so that customers who so choose will not need to call Customer Support. Of course live support will still be available, but the web site should be able to do most of the heavy lifting. GreatCall has a tough nut to crack: none of the carriers I've tried to use for purchasing or provisioning handsets have been even remotely usable, and I've tried several. On the other hand, the obvious need for a phone with large buttons has so far been overlooked in the US.
You'll notice some differences between the Jitterbug and the Simply, another "simple phone." Simply has softkeys, and Jitterbug does not. All navigation is accomplished with the up/down and Yes/No buttons. I'm eager to try this phone out, and to watch members of the target audience try it too. The phone was produced and designed by Samsung, with Arlene's direct input. I suspect they did a pretty good job, but I look forward to putting it through its paces.
Motorola has leaked photos of the upcoming RAZR "successor" via Engadget. The important thing to notice on this phone (shown above) is the lack of a menu button. The menu button is a real source of usability challenge for people, and Motorola must have figured that out along the way.
No one is truly impressed with the industry's snail's pace of innovation in user experience, but eliminating the menu button is a step in the right direction. You can see the menu button in the existing RAZR model's photo at the right, above the five-way directional control. If you're wondering why the menu button has no label, you're not the only one.
In Usable Products' Keypad Trouble video retrospective, the problems with the menu key are made truly apparent through the use of video clips with participants complaining how they don't understand what "that middle button" is for. Participants also complain about the proximity of the center OK button to the four directionals and how all the button surfaces are flat, thereby making it difficult to press a direction without pressing the center OK button. Heck, Moto can't fix everything in one release, now can they?
There are several platforms usable for mobile deployment prototypes.
1. SVGt 1.1+ and the Ikivo Animator, using XHTML as a wrapper
2. Adobe Macromedia FlashLite, using the Flash authoring system
3. WAP by itself
 SVG is the Scalable Vector Graphics format. Ikivo Animator is an animation authoring package. You can create graphical files, representing user interfaces, in any graphics editor that outputs SVG. Adobe Illustrator outputs SVG. Ikivo Animator supports SVG animations of these files with an intuitive graphical user interface. Combining the animations with XHTML enables the prototyper to simulate more elaborate user interactions. Getting the content to the phone depends on the equipment, accessories, and any data plan set up for the equipment. With the Sony Ericsson W600i, we were able to transfer files using the included data cable.
Adobe Macromedia FlashLite player, available on many phones, enables Flash authoring to be deployed on handsets in a similar manner to the SVG described above.
Lastly, WAP can be used as a prototype delivery mechanism. Coding the WAP pages and serving them is up to you.
We have been prototyping handheld applications for some time. Flash is the easiest, but it’s far from easy.
|
|
