Usability requirements: how to specify, test and report usability

Guide for purchasers and procurers

Why does usability matter?

Inadequate usability can lead to significant additional costs associated with poor productivity and increased support overheads. It can also significantly increase the risk that implementation of a new system may fail. It is therefore important to include usability requirements in procurement specifications, and to test usability prior to purchase.

The potential benefits of increased usability are:

Usage savings

  • reduced task time and increased productivity
  • fewer user errors that have to be corrected later
  • fewer user errors leading to increased quality of service
  • less training, support and documentation is required
  • reduced staff turnover as a result of higher satisfaction and motivation

Support savings

  • reduced costs of producing training materials
  • reduced time providing training
  • reduced time spent by other staff providing assistance when users encounter difficulties
  • reduced help line support

Improve the quality of life

  • less stress from frustrating software
  • users are more satisfied

Health and safety legislation

  • the European Directive on Display Screen Equipment (implemented in the national legislation of EU countries) requires that software is suitable for the task and easy to use

See the case studies for examples of the benefits that can be obtained by consumers.

Usability as part of a contract with the supplier

A potential supplier may be concerned at the increased risks and costs associated with developing a system to meet specific usability requirements. A conservative strategy is to require that the new system must be at least as usable as an existing system (see UsabilityNet for more details). This protects against the risk of the costs and overheads associated with reduced usability. If the existing system has known problems, some improvement may be required. If a satisfaction questionnaire such as SUMI or WAMMI is used that has population norms, it is possible to require that user satisfaction is at least as great as the industry average.

To be meaningful the requirements should specify task scenarios which are a realistic representation of the expected usage.

Usability as part of acceptance testing

Ideally testing should involve at least 8 users to assess whether predefined requirements have been met. However, even if no specific requirements have been defined, usability testing will still identify any areas where users have difficulties with the tasks. If the main objective is to identify problems rather than obtain measures, it is more effective to test more tasks with fewer users (4 or 5 per task).

In both cases it is important to use representative users carrying out realistic tasks.

Copyright © 2002 Serco Ltd. Reproduction permitted provided the source is acknowledged.