Almost every technology-based system today claims to have a full set of features and functions and to be easy-to-use. Even those systems that are truly engineered to be straightforward and quickly learned, however, often miss the mark because they lack essential capabilities or information, or they are not compatible with the intended user community’s actual needs.
Systems are purchased to satisfy at least one business case that justifies the expense of buying and deploying them. Buried in the business case is an expectation that target users will be able to readily and completely accomplish an anticipated set of tasks or carry out a critical workflow. If there is more than one role, then the system is expected to support each role effectively. Often key roles are completely overlooked, forcing some users to complete critical tasks without adequate information or functions.
Ease-of-use generally describes how well interfaces work in terms like intuitive, self-describing, forgiving, consistent, etc. An elegant and easy-to-use interface that does not do what the user really needed is not a success. From a broader perspective, an interface is a way to view, manipulate, or navigate an already defined set of information or controls. Is it the right information and are the controls effective for the intended users? How can you be sure?
Usage-centered starts with roles and defined user communities who need to accomplish a set of tasks ranging from research and analysis to a well-defined workflow. Only by eliciting requirements from each of the target communities can the real technology usage be understood. What data needs to be gathered and how is it framed into information? What tools are needed to manipulate the information to gain understanding and knowledge? What parameters are needed to control an operation? What kind of feedback will allow fine tuning of controls? How does the technology serve as a role enabler?
Systems architects need to make key decisions about access to information stores long before end user interfaces are fully designed. Mathematical models, rule-based decisions, smart searching and filtering, etc. may require development. The fundamental architecture and design must consider key quality attributes ranging from security and reliability to performance and adaptability over time.
Both hardware and software choices for the end users must be optimized to support the expected usage. User (not system)-based essential use cases need to identify key interactions and pinpoint the frequency and priority of each usage to facilitate tradeoffs, including identification of the most critical tasks.
The success of the system requires that major stakeholders identify every important interaction. Critical concerns about information access and protection must be identified and appropriately handled. Changing business rules and future needs should also be factored into any design. Influential thought leaders need to buy into such a plan, and understand where and how the technology will accomplish specific goals or requirements. The system context within the larger world needs to be clearly articulated, including dependencies on potential technology fads and methods for minimizing the impact of rapid changes in both technologies and business domains.
Usage-centered thus means understanding, architecting, and engineering the needed usage within each user role, and the systems usage within the context of an always evolving business and technology arena. The approach need not be cumbersome. A lightweight process first concentrates on essential use cases for critical tasks. With rapidly changing technology, success often means getting the crucial usage paradigms right quickly. Once the key usage patterns are checked out, there is time to improve less critical areas during each product iteration.