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.

Usability Matters

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?

Five Ways to Get Better Designs

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.


The Bottom Line: Better Usability = Better Performance

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.

References & Links
  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