How can developers create good user interfaces
when the world around them keeps changing? There are a few
simple strategies to follow, but there are no guarantees.
Easy navigation, speedy performance and uncluttered designs are some of the goals designers strive for when making user interfaces for mobile phone applications. Meeting those goals wouldn’t be difficult if only one kind of phone existed. But even when choosing a single platform, there are still hundreds of different devices. Each is typically conceived and built in six months, running on several different major carriers in the United States alone. How can designers be expected to make application user interfaces that aren’t terrible?
In the process of balancing usability against complexity, it’s vital for designers to sincerely bounce ideas off real-world users before a product leaves the conception stage, not just to give lip service to usability by running a few marketing-oriented focus groups when a phone is nearly shipping, said Scott Weiss, executive director of consultancy Human Factors International and author of Handheld Usability.
 |
| Weiss: Even the iPhone isn’t perfect in user interfaces. |
|
Most people automatically look to the elegantly designed Apple iPhone for the answer, but Apple differs from other handset makers in that it has full control of the software and hardware; it can focus all of its resources on a single product and it needs to partner with a single carrier in each country. Part of the iPhone’s cachet is not in its software code, but in Apple’s intangible corporate mystique – what detractors jokingly call the “Reality Distortion Field” of Co-Founder and CEO Steve Jobs.
Looking past that aura, “The iPhone gets it part of the way there,” Weiss said. Despite the smartphone’s sleek hardware design and impressive screen, “it has an inconsistent user interface that is really bandaged by fancy graphics and animations. Text entry is a true and total nightmare. The obvious reason is, trying to type on a piece of glass is a problem. Every time I write, ‘B, T,’ it corrects me with, ‘B, Y,’ and it doesn’t just ask me if I want that correction, it assumes that it’s right,” he said.
“The back button is sometimes there and sometimes not there. Sometimes it’s a cancel button that moves to the right side of the second row. The edit button [for contacts] is only available in one segmented portion of the user interface,” Weiss continued. Moreover, the iPhone was announced nearly two years ago, but still lacks a cut-and-paste feature, he added.
Designers at independent software vendors are best served by either settling for an average user interface, not a great one, or by focusing on a narrow selection of phones and operating systems. “Joe ISV who wants to write a program and have it work on all of them at once hasn’t got a prayer… he might be able to leverage the backend, but the whole user interface is going to have to be rewritten,” Weiss said. Another possible answer is to focus on Web-based applications, but in that situation, there are also problems as browsers and programmers alike are often inconsistent at following industry standards, he said.
In some cases, the answer of balance is best solved by considering user demographics, said Martijn Van Tilburg, senior design director at Artefact. “We focus mainly on next-gen user experiences. We do that by following a user-centered design process, where we go out and talk to the users to uncover what they need,” he said.
For example, in a software tool aimed at professionals, the interface just needs to work effectively and quickly, not be attractive. Conversely, in an interface designed for consumers via carrier decks, there might need to be a variety of branding and personalization options. But lessons are learned from dealing with customers like Apple, HTC, Intel and Microsoft. The process will only get more complicated, as new smartphones by the end of 2009 will almost universally employ sensors and touchscreens, becoming virtual “social graphs” of a person’s total communications in 2010 and beyond, Van Tilburg and director of product strategy Agnieszka Girling predict.
CHANGES SINCE RAZR
Such ideas are also floated at Motorola. Roger Ady, director of engineering in the consumer experience design group, acknowledges that sleek hardware, not usability, was the success driver of the RAZR – Motorola’s last big hit. As Motorola now is trying to return its handset division to the A-list, “I think definitely within Motorola there is a recognition that the user interface is the centerpiece of the product. So much has changed since the RAZR. The biggest change since then is the display has gotten a lot bigger. You were certainly constrained back then in what you could do,” he said.
 |
| Ady |
|
Motorola is not planning to scale back its use of multiple phone platforms, Ady said. But the company does plan to continue researching UI concepts such as ModeShift, introduced this summer and launched to consumers in the ROKR E8 handset. It extends software user interface ideas to data input. Rather than having physical keys, the keyboard area uses haptic sensors and piezoelectric buttons that can change based on software mode. So different keys virtually appear, even though they feel like real keys to end users and change to completely black when power is switched off.
“We kind of took a shot in this product; what are all the user interface issues we can resolve?” Ady said. He added that it’s important not to make people think too hard when deciding which buttons to touch and to consider how features such as accelerometers might react when a user is driving or otherwise in motion. By the end of 2009 and into 2010, Motorola may be preparing for another big leap forward in mobile user interfaces, by combining touchscreens with a third dimension of zooming interface navigation, he said. “The big challenge is, you’ve got to really keep it focused and the end result is something that is useful for the user,” not merely cool in the eyes of engineers.
Google’s approach, not surprisingly, builds on the Web-is-king idea of application user interfaces. Even the best mobile browsing software today falls far short of the desktop browsing experience, especially lacking in technology such as multithreading, noted Scott Jensen, mobile UI team lead. But it’s more efficient to design mobile software for a few browser types than for a few dozen platforms and a few hundred different phone models. “It’s not like the Web browser is my perfect, preferred way of talking to users. It’s the only freaking one that won’t break the bank,” he explained.
Jenson’s understanding of what works in mobile application interfaces is unique, as his professional background includes several years working on the Apple Newton interface. “There are mobile platforms that are trying very hard to mimic what’s happening on the desktop and what a surprise... they continue, no questions asked,” he said. “So much of the crud you have on the desktop, it’s kind of coming along for the ride in the mobile and it just doesn’t make any sense whatsoever.”
TOUGH CHOICES
Until someone perfects a fold-out screen, no amount of scrolling or zooming or 3D interfaces will be able to let developers avoid the ultimate sacrifice – which is having to make tough choices about how much content to include on a mobile page. More isn’t always better, but usually mobile designers opt for it anyway at the detriment of usability, he added.
In the end, much software goes through carriers before it can be sold to customers, either through a phone portal or just for approval. Sprint product manager Jason Cole works in exactly that role. The company this summer introduced its “1-Click” system of modular tiles for phone homescreens, in which customers can perform tasks such as entering a Google search or a URL without having to open the equivalent of a new window. Next up for Sprint is a way to let developers create new tiles and brainstorm for new applications that may exploit existing homescreen features, similar to how widget applications work on a desktop computer. The idea was under development four years ago, but Apple’s iPhone “certainly did raise the consciousness,” Cole said.
Weiss, of Human Factors, added a note of caution to all mobile software developers. Just as in mobile security, where changes occur faster than security companies can hope to keep pace, so too will mobile phone features continue evolving faster than user interfaces can reach nirvana. By the end of 2009 alone, he predicts the world will see iPhone copycat phones from every major manufacturer, a return to QWERTY keyboards, 8-megazpixel cameras with autofocus and optical zoom, much more sophisticated mapping and location-aware features, not to mention ubiquitous application stores. The best thing mobile user interface designers can do, he said, is think before acting.
| Smartphones May Become Social Communicators |
The rules for designing mobile user interfaces are about to become much more complicated because of the growth of social networking.
|
| Today, social networking systems such as Facebook are available as downloadable applications for smartphones |
The typical "social graph," a map depicting one person's various spheres of contact, is quickly expanding in the mobile world, according to Agnieszka Girling, Artefact's director of product strategy.
Today, social networking systems such as Facebook are available as downloadable applications for smartphones, but within a year or two such applications may be directly integrated into a phone's menu and navigation system.
For example, today users can call, text, e-mail and set person-specific ringtones. But in a smartphone that's more like a social communicator, users could check on a contact's availability, location and mood. Users also could review previous communications with the contact and find new contacts through existing ones.
Scé Pike, co-founder of consultancy Quantum Mobility Solutions and of software company DialPlus, put it more simply: "Let users find content through their contacts and contacts through their content." DialPlus' software can change a phone screen and offer new information based on a wide variety of contextual situations.
The company hopes to announce a beta test with an unspecified Tier 1 wireless carrier this quarter, she said.
Cell phone, smartphone, iPhone, Google phone - will the social phone be next? One company that currently makes a social-oriented phone is INQ Mobile, backed by Hutchison Whampoa. Its INQ1 handset accomplishes many of the same ideas suggested by Girling and Pike, such as having an address book that incorporates Facebook photos and states updates. The phone is based on Qualcomm's BREW platform, adding Java and a proprietary APIs. INQ sells the phone in Europe and Asia and plans to come to the United States, although no timeframe has been announced. There are no plans to make INQ's product software-only.
The deep integration requires custom hardware, company officials said. —Evan Koblentz
|