Administrators Anonymous - UCD, TCO, LCM, and other interventions (07 Oct 2005)
The lack of focus on administrative interfaces often comes from management's mandates to prioritize end user facing screens ahead of anything else. The number of "eyeballs" is greater for the end user screens than admin screens. It is also easier for all stakeholders, including interaction designers to understand the domain of an e-mail application than it is to grasp things like complex system monitoring, visualization of clickstream data, or the tools needed to bridge interdependent systems. However in more complex software, this initial emphasis on the end user turns out to be a short lived priority. The more significant costs of running software are often associated with installation, configuration, deployment, maintenance, and upgrade. Often, this is referred to as LCM, or Lifecycle Change Management. Industry estimates state that the budget for LCM can be 2-4 times as large as the initial license cost of enterprise software. More information is needed about how administrators work to manage these systems and what business and integration problems they are trying to solve. UCD, or user centered design can help answer these questions.
The administrator is often a misunderstood user type. Experts have recently started to shed light on this subject. As reported in this forum by Rob Barrett of IBM Almaden, administrators cling to their shells, scripts, and other command line utilities. We create GUI tools for them, but is that what they really need to lower the Total Cost of Ownership (TCO) of software, or to be more efficient when communicating with each other? What are the real frustrations in a given admin's day and how can Human Computer Interaction practitioners help them?
This talk will focus on better understanding administrators and what design solutions and techniques work. I will present more information about the latest user classifications I found in mid-to large scale companies. Their user profiles vary in terms of technical skill, breadth, and responsibility. I will also talk about a study of only DBAs, where my team saw a large difference between self reported and observed data concerning their real tasks.
Practical case studies of software ecosystems that these admins inhabit will be showcased throughout this talk. While sometimes unexpected, the tools and techniques that an HCI practitioner, or any software designer, can deploy will decrease the frustration, complexity, and cost associated with enterprise software deployment. The bridging of technical and organizational silos will often be required, and designers will need to extend their role in order to bring about any significant change and make life (and command line withdrawal) easier for Administrators Anonymous.
Luke Kowalski is the Corporate UI Architect at Oracle Corp. His role serves to bridge the user interface design groups at Oracle and he works as an evangelist for effective UI technology, on legal aspects of user interfaces, business context, and cross-divisional information architecture integration. Prior work includes begging for money at startups as director of UI and Web, a longish stint at Netscape's Server User Interface Group, as well as other odd HCI jobs to fill out the 15 years in practice. He holds several UI patents, a CPE Certification, and two Masters Degrees, one from Pratt Institute, and another from Columbia University. His PhD in Divinity cost him $5 from http://www.ulc.org/ and allows him to park in the minister spot in front of hospitals. He has also published book chapters and articles and spoke at the Nielsen Norman Group conference in Sydney in 2002.
Article URL: http://hci.stanford.edu/seminar/abstracts/05-06/051007-kowalski.html
(Luke Kowalski, Corporate Ui Architect, Oracle Corp)
Read 153 more articles from Stanford HCI Seminars sorted by
Next Article: Ceatec shows off future gadgets