Fresh: food for thought

XII.4 July + August 2005
Page: 8
Digital Citation

Why doesn’t SIGCHI eat its own dog food?

John Morris

back to top 

Investopedia says a company that "eats its own dog food" sends a message that it considers its own products the best on the market. This slang was popularized during the dot-com craze when companies did not use their own software and thus could not even "eat their own dog food." An example of not "eating your own dog food" would be a software company that creates operating systems but uses its competitor's software on its corporate computers.

As a discipline, we believe in putting the user at the center of our designs—involving them in the process whenever possible, and evaluating the results of our designs and design modifications with real users, preferably during the development stages of the product. Why, then, don't we follow this same approach with respect to the development of the "products" of our organization? Does user-centered design only apply to computer-based products, or should we, as an organization, demonstrate the strength of user-centered design by practicing it for the development of our Web sites, publications, and conferences?

While I believe that we can and should practice user-centered design on all of our products, this rant is going to focus on the annual CHI conference. The CHI 2005 "product" shipped, and we have announced that the overall design for the next "product" in the series, CHI 2006, will incorporate several significant changes. The purpose of this rant is not to discuss or criticize those changes or any of the volunteers who give of their time and energy to benefit the community and the field. This is a rant about our process of design, and why I believe a more user-centered approach is in order.

User-centered design (UCD) is a process by which the user of the product is the focus of the design rather than the product itself being the focus. We believe that involving the user in the design process results in better products that are more usable than products designed without user involvement. We espouse the testing of product designs, features, and interaction techniques with users early in the development process. We encourage the study of users in context before we have firm product directions. We caution developers and designers not to make assumptions about users without testing those assumptions.

Let us turn now to the use of UCD for the development of a new release of a successful product. Ideally, we would have two separate pieces of information as input to our design: ongoing studies of our target user community and evaluations of previous versions of our product based on surveys and user testing. Management has set specific goals and constraints for the design, including development time, financial goals, marketing goals, etc. Whenever possible, we would derive metrics to evaluate our product based on our goals and constraints. The design process might include a period of brainstorming to search for innovative new aspects to the product based on our assessment of the user's needs and management goals. Innovations that introduce significant changes to the product would probably be labeled as "risky" and would need to be prototyped and explicitly tested against real users. In the end, the entire product would be tested and evaluated against the metrics and goals established by management.

One of the major products of SIGCHI is the CHI conference. The above outline bears little resemblance to the actual design process for each year's "product." We generally assume we are the user, seldom evaluate, and rarely do formal testing of our prototypes. We have articulated goals for the conference series, but we don't tie those goals to our evaluation of the conference, or our design changes to the conference in any formal manner. We point to conference evaluations and attendance patterns as rationale for changes without testing our assumptions or our proposed designs against our target audience in any formal manner. One of the constraints we have is that this process is managed by volunteers who have real jobs and lives outside of SIGCHI and the CHI conference. All design processes have constraints, but we cannot afford to allow those constraints to force the user out of the design.

Eat Your Own Dog Food, an expression describing the act of a company using its own products for day-to-day operations.—

User-centered design, or more broadly, a focus on the user, is at the core of the discipline of HCI. The annual CHI conference is arguably the premier technical conference for HCI professionals, yet we seem to approach the creation of the conference without applying any of the tools or techniques we espouse as critical to good design. Isn't it time for SIGCHI to "eat our own dog food" and practice user-centered design in the ongoing evolution of the CHI conference?

back to top  Author

John "Scooter" Morris
University of California, San Francisco

About the Author:

A recipient of the SIGCHI Distinguished Service Award, Scooter has worked tirelessly for SIGCHI and the CHI conferences—he was the first AV chair (ever) at CHI `85. Scooter is currently the executive director of the Resource for Biocomputing, Visualization, and Informatics (RBVI) at the University of California, San Francisco, which develops software tools and Web sites to support the analysis and visualization of a variety of complex biological problems and interactions. Scooter serves on the ACM Council.

back to top 

©2005 ACM  1072-5220/05/0700  $5.00

Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.

The Digital Library is published by the Association for Computing Machinery. Copyright © 2005 ACM, Inc.

Post Comment

No Comments Found