During a field study that included a major financial institution’s contact center, a senior manager grabbed my attention with this nugget: “every second is a full-time employee in our world” [1].

He went on to describe how, if the average-call time increased by 20 seconds per call, his center must hire 20 new agents at a loaded cost of almost $1 million every year. His math was based on a contact center with about 240 agents who handle the average call in less than two minutes. Your math might be different, but the principle is the same: faster is cheaper.

Faster may be cheaper but it’s not always better. Rushed calls can result in poor customer satisfaction, lowered first-call resolution, missed cross-sell opportunities, increased agent stress, and higher agent turnover. The well-known challenge of contact center management is the challenge of managing these often-competing priorities.

Faced with demands from head office to support a much more complex product offering while maintaining current call-handle times, the senior manager quoted above turned to the same process that Apple uses to develop its products, including the phenomenally successful iPod. The process is called usability engineering and is intended to enhance product usability, defined in ISO 9241 as “the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency, and satisfaction.” Usability engineering (also known as user-experience engineering [2] or user-centered design) is a best practice among the world’s leading high-tech companies including IBM, Motorola, Google, Amazon, and, yes, Apple.

Apple’s usability engineering process led to innovative features like the iPod’s signature Click Wheel and more recently its touch interface. Perhaps more importantly, the process helped Apple’s engineers get the details right, leading to a product that “just works.” If the iPod was as difficult to use as a typical VCR, Apple’s share price might still be around $8—where it was before the iPod’s introduction—instead of today’s $180.

How can a process used to design consumer technology be applied to a contact center? In other words, what does the iPod have in common with a contact center? The short answer is they both have a user interface with which users control their technology. In the contact center, the main user is the agent, and the main user interface is the agent-facing software. This could include the database front-end used to retrieve and enter customer data, the knowledgebase that contains product and policy information, and any other software the agents use to do their jobs. Unfortunately, agent-facing software is often as difficult to use as a poorly designed VCR.

For example, agents struggle with entering and retrieving data from several different applications or having to drill deep into an application to get what they need. An agent interviewed during a field study of agent-facing software put it this way:

Why do we have to go to all these places? Log here, look for here, oh I didn’t find it here, so I have to go to there. Why can’t they integrate everything in one place?

This is the problem that the unified agent desktop software attempts to solve. A unified desktop software can improve the usability of agent-facing software, but to maximize agent productivity gains, components of the desktop must be designed to integrate with the agent’s workflow.

Other productivity drains include poor text labels and obscure abbreviations. One agent complained:

It’s very, very convoluted. We’re basically looking for data in fields with names that are totally unrelated to what you’re actually looking for… It’s just not intuitive at all. Not intuitive. And the abbreviations don’t make sense at all.

These are only a few of the problems that impede agent productivity. Other issues are subtle and agents may not report them. For example, often user-interface text is presented in all uppercase. Unfortunately, uppercase text has been shown to take 13% longer to decode than mixed-case text. This will have an impact on agent productivity although agents are unlikely to notice it.

Improving the usability of agent-facing software has a direct and inverse relationship with call-handle time: the more usable the software the shorter the average call-handle time. But usable software provides other benefits: it reduces agent stress and agent training costs by simplifying the agent’s workflow, supports cross-selling efforts by putting sales information at the agent’s fingertips, and improves customer satisfaction by helping agents serve clients faster and better. Imagine what would happen if your agent-facing software “just worked” as easily and elegantly as an iPod? Even if you couldn’t make it that user friendly, making it more agent friendly is worth the effort. How can the usability engineering process make this happen?

Usability engineering is not about technology. You can apply it to any type of agent-facing software on any platform that can be custom configured. Usability engineering is about observing and listening to contact center agents and incorporating what is learned into the design of their user interface. It is highly iterative: data is collected, designs prototyped, prototypes tested, all in several cycles until a productive design is delivered.

Usability is engineered in five iterative phases:

1. Business objectives for the engineering effort are defined, for example, to reduce average call-handle time by 20% or to increase cross-sell rates by 10%.
2. User research is completed by watching agents work and noticing pain points and time wasters. Agents are also interviewed to more fully understand the thought processes and motivations that lead agents to interact with the software in a particular way.
3. User requirements are defined by identifying areas where business objectives and user research overlap. For example, if a business objective was to increase cross-sell rates and user research showed that agents took an average of ten seconds to find the appropriate pricing information, a user requirement would be to reduce average find time to not more than three seconds.
4. User interface design is completed by developing a series of increasingly detailed models or prototypes, iteratively evaluating and redesigning each type before moving to the next. These would begin with the information architecture in the form of a diagram showing how menus or data are to be structured, then progress to a sketch of the screen design, and end with an electronic prototype of the software. This iterative design process contains development costs because a mockup is much less expensive to redesign than a finished suite of agent-facing software.
5. Usability evaluations are implemented by testing models and prototypes with agents as they perform simulated tasks. These tests generate new design ideas and determine when user requirements have been met.

By following a usability engineering process, the financial institution described above dramatically improved the usability of their agent-facing software. In the words of one agent: “Everything is user friendly. After a couple of months you’re a pro.”

Because of this enhanced usability, the contact center was able to maintain its call-handle times despite a significant increase in the complexity of the product it supported. Had the agent facing software remained the same, call-handle times would have likely gone through the roof.

Of course, usability will not solve all your contact center problems. However, in most cases it will improve performance. Sometimes a little, sometimes a lot. And in a world where faster is cheaper, every second counts.

1. How may I help you? An ethnographic view of contact center HCI, Howard Kiewe, International Journal of Usability Studies, February 2008
2. User Experience Engineering Essentials: Series Introduction, Howard Kiewe.com, August 2007

This post is the first in a series of 16, originally published on April 15th, 2006 as the “UXE White Paper: User Experience Engineering Essentials."

I coined the term "User experience engineering (UXE)" to describe a structured research, design, and evaluation process whose goal is to make user interactions with a product or service easy, efficient, and enjoyable. It evolved from usability engineering and applies psychological principles and methodologies.

Our experience of a product can be divided into three primary levels: visceral, behavioral, and sociocultural. The visceral level is an immediate and instinctive reaction that is greatly influenced by the appearance of the product. The behavioral level develops through our interaction with a product and is greatly influenced by its usefulness and usability. The sociocultural level appeals to self-identity and personal symbolism. UXE addresses these three levels of experience in product and service development.

The user experience is engineered in five iterative phases:

1. Business objectives for UXE. Define the business or organizational objectives for the UXE effort, so UXE produces maximum impact.
2. User research. Collect targeted data on users (prospective or actual), their tasks, and their environment through observation and discussion.
3. UX requirement analysis. Define specific product and UX requirements for areas where business objectives and user research overlap.
4. UX design. Develop a series of increasing detailed models or prototypes (conceptual, architectural, interactive, aesthetic), iteratively evaluating and redesigning each type before going on to the next.
5. UX evaluation. Evaluate and test models and prototypes to generate new design ideas and to determine whether UX requirements have been met.

Each phase has its own work products that capture knowledge generated in that phase and help designers apply the knowledge in the next phase.

Most major technology companies including IBM, Microsoft, Samsung, eBay, Amazon, and Apple have invested heavily in UXE. UXE is also rapidly making inroads into established industries, with companies like GE Healthcare, Whirlpool, and Procter & Gamble reaping huge product successes from it. Despite this momentum, there is still room for improvement. A review of 700 corporate Web sites, gave only 3% a passing grade, and most high-tech gadgets remain too hard to use. This means UXE can still provide a competitive advantage to any company that chooses to invest in it.

Best-practice companies invest from 10% to 19% of the total product development project budget in UXE, which offers an impressive return on investment (ROI) of $2 to $100 for every $1 invested, accomplished in four main ways:

1. Increasing sales by improving product appeal, making e-commerce sites more usable, stimulating cross-channel activity, and improving customer retention.
2. Enhancing productivity by increasing employee efficiency.
3. Decreasing costs of development, maintenance, support, and training.
4. Reducing risk of project cancellation and litigation.

For maximum ROI, UXE methods must be used from the beginning of the project so project requirements are based on marketable user needs and development efforts achieve maximum efficiency. In today’s networked, interconnected marketplace, consumers will shift their loyalties to the companies and brands that can provide them with the user experience that they value, so UXE methods offer a competitive advantage with immediate tactical value and long-term strategic benefits.

• Series Introduction
• What Is User Experience Engineering?
• What Are The Levels Of UX?
• How Is User Experience Engineered?
• Who Uses UXE?
• Why Use UXE?
• When To Use UXE?
• User Experience Articles and References

User experience engineering (UXE) is a structured research, design, and evaluation process whose goal is to make user interactions with a product or service easy, efficient, and enjoyable. UXE methods can be and are applied to the development of virtually any product or service, from mops [142] to surgical devices [98], to retail environments [32]. However, the term is most frequently applied to digital products such as software, Web sites, and electronic devices. This paper focuses primarily on UXE as it is applied in the digital context.

UXE has been influenced by the fields of human factors and ergonomics and is considered a practical application of research in human-computer interaction [95]. It evolved from, and shares much in common with, usability engineering [78; 866] whose central concern is product usability, defined in ISO 9241 as “the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency, and satisfaction” [18]. A user’s experience of a product includes its usability and extends to the user’s emotional interaction with the product and the organizational channels that promote, supply, and support it. UXE addresses these additional elements. User experience professionals typically work closely with marketers, software engineers, graphic artists, industrial designers, customer support staff, and other professionals as good user experience results from a team effort.

If engineering is the application of scientific knowledge to practical problems, the science from which UXE draws most of its primary theoretical and methodological inspiration is psychology. UXE applies experimentally validated principles [22; 84; 95] from the fields of sensation, perception, and cognition [20], emotion and motivation [92], as well as social psychology [38; 39]. UXE also adapts research methods from experimental psychology [116] and related social sciences such as anthropology and sociology [81; 140].

Our experience of a product can be divided into three primary levels [92]: visceral, behavioral, and sociocultural (elsewhere referred to as reflective [92] or symbolic [107]). The visceral level of experience is an immediate and instinctive reaction that is greatly influenced by the appearance of the product. The sleek lines of a classic automobile appeal to this level of experience. The behavioral level of experience develops through our interaction with a product and is greatly influenced by its usefulness and usability. A cup holder in an automobile need not be attractive because it is useful. If you drive your car along a winding road and can drink your beverage without spilling it on yourself, it is usable. This combination of usefulness and usability appeals to the behavioral level of experience. The sociocultural level of experience appeals to self-identity and personal symbolism and is influenced by cultural norms. Driving a mini-van says something different than driving a sports car. Most importantly it says something to the driver, it is an act of self-definition [36; 108].

• Executive Summary
• What is User Experience Engineering?
• What Are The Levels Of UX?
• How Is User Experience Engineered?
• Who Uses UXE?
• Why Use UXE?
• When To Use UXE?
• User Experience References

Although certain personality traits moderate visceral influences [17], the visceral level of experience is largely pre-wired and biologically determined. We have evolved visceral responses that attract us to objects and environments that provide a survival advantage and repel us from those that pose a survival risk [66]. We like flowers because they signal forthcoming nourishment, vistas because we feel safer if we can see danger coming, and lakes and rivers because the water they contain quenches our thirst and implies additional food resources [5]. We like symmetrical features in a potential mate because they suggest health and therefore fertility [19]. We dislike dark environments and sudden loud noises because they predict danger [48]. However, we can learn to like some stimuli that are naturally aversive. For example, we can acquire a taste for bitter or spicy foods if we are regularly exposed to them and receive social reinforcement for their consumption [143].

Visceral preferences are based on our sensory impressions and are largely influenced by the aesthetic qualities of the perceived object. They are immediate, emotional, and pre-cognitive. Neurophysiological evidence shows that emotionally evocative stimuli directly impact the amygdala—the part of the brain responsible for emotional experience—without involving the higher cognitive centers of the cerebral cortex [e.g., 70]. When we see or hear something that is potentially pleasurable or threatening, we become aroused—measured as increases in heart rate, sweat, and startle responses [14]—and are either drawn to the stimulus or away from it, all this before we can consciously analyze or sometimes even notice the stimulus. Based on our first impression we either like something and want it or do not like it and want to avoid it. This first impression biases any subsequent cognitive evaluation, sometimes reducing a logical analysis of product benefits to nothing more than a justification for a preference already formed at the visceral level [143].

Since the emergence of industrial design in the early 1900s, aesthetic appeal has been used as a powerful marketing tool [35; 101]. Apple Computer’s 1997 launch of the iMac is a recent example of a successful product that relied primarily on aesthetics to drive sales. The iMac’s curvy all-in-one form factor and colorful translucent casing stood in stark contrast to the bland putty-colored box-like computers available from PC manufactures, leading one reviewer to say “Wow! I gotta get me one o’ those right now!” [117, p. 22]. Brisk sales in the first year increased Apple’s market share by 66% (from 3.5% to 5.3%) [138], this despite the fact that iMac offered only incremental functional improvements over earlier models.

While the iMac is one of many examples in which compelling aesthetics led to a popular consumer product [35], the conventional view of the importance of the appearance in business-to-business, industrial products has been that “business products do not necessarily make anyone look or feel better, and they generally do not have significant aesthetic value” [15, p. 123]. However, this perspective has been shown to be incorrect: a study of engineering, marketing, and purchasing personnel who evaluated industrial products like motors, oscilloscopes, and pumps preferred those with attractive visual aesthetics and in some circumstances the influence of visual ascetics on purchase intentions exceeded that of product performance and price [141].

The influence of aesthetic appeal extends to Web sites and software. In a study that looked at user’s first impressions of Web pages, pages high in the subjective attribute of “beauty” gave the best overall impression [118]. Another study measured how persistent users were in browsing for information and found users significantly more persistent on sites whose color scheme they rated as attractive [82]. And in several studies of software for automated teller machines [64; 128; 129], industrial control [52], and MP3 players [51], users who perceived the software as beautiful and appealing also perceived it as useable during their initial evaluation. Perceived usability is important because software that is perceived to be usable is likely to be accepted and adopted [29; 133]. In some cases the initial perception of usability persists even after users interact with the software that is designed to be hard to use [129].

If the visceral level of experience is so compelling and colors perceptions of behavioral attributes like usability, is it the only level of experience we should be concerned with? It is true that visceral experience greatly influences our first impressions and biases our subsequent judgments. However, first impressions do change: for example, one study showed that users modified their original evaluation of the software usability after interacting with it, showing that the behavioral level of experience is influenced by product interaction [51].

• Executive Summary
• What Is User Experience Engineering?
• What Are The Levels Of UX?
  • The Visceral Level
  • The Behavioral Level
  • The Sociocultural Level

• How Is User Experience Engineered?
• Who Uses UXE?
• Why Use UXE?
• When To Use UXE?
• User Experience References

Designing for the visceral level of experience attracts attention and may motivate a customer to buy a product. But consider what can happen when the product does not deliver at the behavioral level what it implies at the visceral:

Shortly after midnight, a resident of a small town in southern California called the police to report hearing a man inside a house nearby screaming “I’m going to kill you! I’m going to kill you!” Officers arrived on the scene and ordered the screaming man to come out of the house. The man stepped outside, wearing shorts and a Polo shirt. The officers found no victim inside the house. The man had been yelling at his computer [39, p. 89].

The description above, based on a police report from Seal Beach, California, illustrates the intensity of frustration that can develop when the behavioral level of experience is not addressed in software design. And this example is not even the most extreme: a man in Lafayette, Colorado was arrested on suspicion of felony menacing, reckless endangerment, and the prohibited use of weapons after he shot his laptop four times in a bar and hung it on the on the wall like a hunting trophy. He never explained what prompted his actions, but told police that it seemed like the right thing to do at the time [24]. These are extreme examples of what is referred to as “computer rage” [25]; computer rage is widespread, for example, 70% of computer users surveyed in the UK admitted to shouting, swearing, or being violent to their computers [16]. The cost to British industry is a staggering £25,000 ($40,000) per employee per year in lost productivity as users work to troubleshoot computer problems [25]. As astounding as this estimate is, controlled research backs it up, one study showing that 38% of computer user’s time is lost to frustrating experiences [23]. Chief among these frustrations were user interface issues such as poor error messages, bad URLs, system responses that are inconsistent with user actions, and annoying, missing, or hard to find features.

For companies that use software and intranet information sources—whether externally or internally developed—this sort of frustration carries a huge cost in lost productivity and dampened employee moral. For companies that sell software and Web services, this sort of frustration provokes customer defection, damages hard-won reputations, and reduces word-of-mouth referrals, the consequences of which are well documented: reducing customer defections by 5% can boost profits by 25% to 85% [110] and retaining existing customers is six times less expensive than attracting new ones [114]. Add to this the escalating support cost when users vent to support staff, and you begin to understand the business impact of poor behavioral design.

A positive behavioral experience results from a combination of product usefulness and usability. A product is useful when it meets the day-to-day needs of its users. In other words, most of its features are needed or wanted by most of its users. This is a delicate balancing act: missing features will frustrate users, while too many features will make the product unwieldy and cumbersome. Observational methods outlined in the User Research section  (see p. 15) can help you discover the features that have significant user appeal and avoid investing resources to develop features that will simply clutter your product.

A product is usable when it is effective, efficient, engaging, easy to learn, and error tolerant [106]. These “Five Es” of usability are explained below:

• Effective. Affords goal accomplishment with minimum effort.
• Efficient. Allows rapid task completion with few errors.
• Engaging. Offers enjoyable day-to-day operation.
• Easy to learn. Supports rapid initial skill acquisition and expanded skill development with experience.
• Error tolerant. Prevents errors and supports error recovery.

A product with good usability will have all the above attributes. Combined with usefulness, usability provides a consistently positive behavioral experience that inspires high-levels of product loyalty. Behavioral and visceral appeal together create a product experience that is attractive on first impression and remains enjoyable over the long haul, making a compelling product offering. However, there is one more level of experience to consider during product development, the level.

• Executive Summary
• What Is User Experience Engineering?
• What Are The Levels Of UX?
  • The Visceral Level
  • The Behavioral Level
  • The Sociocultural Level

• How Is User Experience Engineered?
• Who Uses UXE?
• Why Use UXE?
• When To Use UXE?
• User Experience References