User-centered design in IT services: End-to-end user experience and conclusion

This is the final article in the "User-Centered Design for IT Services" series. These articles give an introduction to a lightweight process for including user needs, goals, and feedback into a development process.

In this final article, we discuss larger issues that affect the usability of IT services, and provide a number of resources to learn more about the topics we have covered in this series.

Mashups

To design a truly successful service, it is necessary to broaden your focus beyond the specifics of your website or application and focus on the overall, end-to-end user experience. This is especially true in cases where systems are being created by pulling together software, services, and data from multiple providers, both on and off campus. These new techniques (often referred to as mashups) offer many advantages, including reduced costs and increased speed of development. Unfortunately, keeping the design of a system focused on a user and maintaining its usability often becomes more difficult. In particular, components from different sources often use different terminology or user interfaces for the same functionality, or use the same term or interface for different functionality.

As we have discussed before, one of the major goals of user-centered design is to increase user productivity by reducing the percentage of mental effort users expend on operating a system, so that they can concentrate more fully on the task they are trying to get done. Forcing users to figure out these inconsistencies, particularly in different parts of what — to them — appears to be a single system, greatly increases the mental effort necessary to operate the system.

It's not always easy to resolve these issues. If you are developing a system where you are only integrating one component from another source, matching the terminology of the rest of your system to that component is often helpful. In the case where you are pulling together two or more systems, services, or data sources, this can be significantly more difficult. Tools such as CSS and XSLT can help reduce this confusion. Specifically, CSS can help you change the way items on a web page are displayed to maintain consistency. XSLT can help you manipulate the XML data sources often used in mashups in many ways, making it possible to increase consistency in terminology and formatting.  Alternatively, you may decide to design your user interface so that it's clearer where terminology and behaviors change, and why.

User expectation: Standard terminology and interactions

Though your service may be extremely well designed and very usable, it exists in a larger context of personal, organizational, industry, and Internet standards. As you design your service, be sure to pay attention to these standards. For example, here at UC Berkeley, there is a standard screen that requests a user's CalNet ID and passphrase to log onto a service. Asking for the CalNet ID and passphrase in a different way may confuse users, possibly cause them to mistrust your site, and likely be a violation of the campus' security standards. Jakob Nielsen mentions a more humorous, but very revealing, example in one of his columns:

Eric Davis, an Information Architect with Resource Marketing, recently reported on a usability test of shopping cart terminology. The draft design featured the term "Shopping Sled" since the site (selling winter sports products) had a desire to stand out and avoid standard terminology. Result: "50 percent of users did not understand The Sled concept. The other 50 percent said that they figured out what it meant because it was in the same location as a cart would be. They knew that you had to add to something, and the only something that made any kind of sense was the Sled." Lesson: Do not try to be smart and use new terms when we have good words available that users already know. [1]

Wide compatibility

If the service you've designed requires users to upgrade or change the software they have installed (or even their operating system), they're unlikely to comply. Rather than going through the pain of an upgrade, most users will instead try to find an alternative or workaround (such as calling you!)

Put another way, the service that users can't access is not usable at all.

It's important to consider the whole service from the users' point of view, starting with where and how they will be accessing it. So, for example, for a web-based service, you will probably want to check the following:

• Multiple browsers on multiple platforms. At a minimum, you should test your service with Firefox and Internet Explorer on Microsoft Windows, and Firefox and Safari on Mac OS. Beyond that, you may want to emulate Yahoo and support the "A-Grade" browsers listed at http://developer.yahoo.com/yui/articles/gbs/. Cross-platform compatibility is especially important in a university environment — a survey of nearly 40,000 computers on the UC Berkeley campus network in July 2006 showed 68 percent of the computers running Windows, 22 percent running Mac OS, 6 percent running a Unix variant, and 4 percent other/unknown [2].

• Older browsers. Most people do not have the most current browser with the most current plug-ins installed. Therefore, testing with somewhat older versions of browsers will ensure compatibility. Evolt.org maintains a comprehensive archive of older browsers to install and test with, and browsershots.org will take screenshots of your site with multiple browsers and post them for you to retrieve.

• Multiple connection speeds. Many people still access the Internet through dialup (or relatively slow "broadband") connections. If your service has a wide target audience, especially if it's a worldwide audience, make sure it's usable over these slower connections.

• Multiple devices. Smartphones and PDAs that can connect to the Internet, such as the iPhone, are becoming increasingly popular. As these devices tend to have very small screens, they can often cause interesting problems for sites that you design. For example, on one recent site we've seen, all of the text is rewrapped into a single one-character-wide column when viewed with Windows Mobile's Pocket IE browser.

• Your server logs. Your web server logs are a great way to get a sense of which browsers your users are using to connect to your site. This can give some good guidance about where to concentrate your development efforts.

Conclusion

We hope this series has piqued your interest about user-centered design, as well as given you some initial tools to begin practicing it in your work. Although in most organizations it's nearly impossible to build a user-centered culture overnight, now you can begin taking steps to build awareness about the importance of user-centered design. Understanding your users' needs, tasks, and goals is a critical first step in building websites, applications, and IT services that fade into the background or even delight users as they do their work.

Incorporating user-centered design into your work

If your goal is to incorporate more user-centered design into your work and move your organization forward in Jakob Nielsen's "Corporate Usability Maturity" model [3], there are some concrete actions you can take. The first step is often to prove the value of spending time on user-centered design activities to others in your organization. A good way to do this is to successfully utilize user-centered design techniques on a project, and to clearly show how they helped make the project successful.

Choosing a project

When choosing a project in which you'd like to incorporate user-centered design techniques, select carefully in order to set yourself up for success. In general, a good project is usually one that you can get involved with early, as opposed to one where you are asked to "usability it" at the end of the process after most of the design and development has already taken place. However, for your initial forays into user-centered design, you may want to select a short project which allows you to show some quick wins, such as improving an existing project. You should try to select a project with users you will be able to access fairly easily, and you will definitely want to choose a project with a team that is receptive to incorporating user input and feedback into the process.

Perhaps the most important thing to remember when selecting a project in which to incorporate user-centered design, is learning when to say "no". If it's important to show how user-centered design can improve your projects, you'll want to make sure not to spread yourself so thin across so many projects that you are unable to show measurable results on any of them. Make sure the scope of the project and your efforts within it are manageable, and that spending time with users will generate significant returns on your investment.

Choosing techniques

When you begin using user-centered design techniques in your projects, as we've mentioned previously, you really won't want to use every tool in the user-centered design toolbox. Especially if you are in the early stages of incorporating user-centered design, you will want to choose your techniques carefully. At first, it's probably best to choose the techniques which offer the most return for the least investment. As you show your organization how much value user-centered design can add to your projects, you will likely receive more support for the more time-intensive activities such as user observation.

Which user-centered design techniques to use will depend on the context of your site/application and resource constraints. A good place to start may be by doing a heuristic evaluation of an existing site, and suggesting improvements. If you are involved with a system that clearly isn't working for users, you can illustrate this to other members of your team by doing user testing. If you're designing a website, application, or system with lots of information, card sorting can give you a lot of helpful user input without a huge investment of time. Finally, when you are ready to progress to a project you can design from the ground up, be sure to begin with a user needs assessment. Techniques ranging from interviews to contextual inquiries to user observation can all be effective depending on the complexity of the project and the depth of the knowledge you need to gain about your users.

Selling user needs assessment

More information on selling user needs assessment in your organization can be found in the PDF presentation Why Your Sponsors Know Too Much and What to Do About It: Selling and Performing User Needs Assessment.

Where to start learning more

At UC Berkeley, we've organized a campus user-centered design group which meets monthly to discuss user-centered design issues. We also organize other activities, such as viewings of user-centered design related seminars. You can find out about all the group's events by subscribing to the list. To subscribe, simply go to https://calmail.berkeley.edu/manage/list/listinfo/ucd@lists.berkeley.edu and follow the directions there.

There are also many great books, special interest groups, and courses that can help you continue to build your knowledge and start to use more advanced techniques and practices. Enjoy!

Recommended books

• The Inmates are Running the Asylum and About Face 3.0 by Alan Cooper
• The Design of Everyday Things and Emotional Design by Don Norman
• Designing Interfaces by Jenifer Tidwell
• Usability Engineering by Jakob Nielsen
• Don't Make Me Think by Steve Krug

Recommended websites

• The Fluid Project's User Experience Toolkit
• The Association of Computing Machinery's Computer-Human Interaction Special Interest Group and the San Francisco Bay Area chapter
• useit.com: Jakob Nielsen on Usability and Web Design
• Usability.gov
• Usability Professionals' Association
• User Interface Engineering
• Cooper
• The Society for Technical Communication's Usability & User Experience Community
• Usability First
• Usable Web
• Usability Net
• Design @ IBM

Recommended Courses

• Cooper Interaction Design Practicum
• UC Extension Courses, such as User Research: User Needs and Usability Assessment for Web and Software Products

Notes

[1] From Jakob Nielsen, Do Interface Standards Stifle Design Creativity?

[2] Based on SNS security scans in early July 2006, scanning 39,219 computers over a 9-day period. The precise breakdown of platforms was 68.24 percent Windows, 22.24 percent Mac, 4.45 percent Linux, 2.93 percent Unknown, 1.27 percent Solaris, and 0.77 percent Other.

[3] From Jakob Nielsen, Corporate Usability Maturity: Stages 1-4, and Corporate Usability Maturity: Stages 5-8.