CHI 97 Electronic Publications: Tutorials
CHI 97 Prev CHI 97 Electronic Publications: Tutorials Next

Design Ethnography: Using Custom Ethnographic Techniques to Develop New Product Concepts

Tony Salvador
Intel Corporation JF3-210
2111 NE 25th Ave
Hillsboro OR 97124 USA
+1 503 264 6455
tony_salvador@ccm.jf.intel.com

Michael Mateas
Carnegie Mellon University
2230 Shady Ave.
Pittsburgh, PA 15217 USA
Michael.Mateas@cs.cmu.edu

ABSTRACT

Design Ethnography is a set of data collection and analysis perspectives, assumptions and skills that can be used effectively and efficiently to understand a particular environment, or domain, of people for the express purposes of designing new technology products. Working from the data one forms models of the environment explicitly considering the peoples' relationship to other people, space, time, artifacts, activities and nature. The models, graphically represented, are used explicitly to derive and test product concepts.

Keywords

Ethnography, consumer market, home, teenagers, business communication

© 1997 Copyright on this material is held by the authors.



INTRODUCTION

In this tutorial, it is our intention to share with our professional community the perspective, methodologies, results and underlying assumptions of the work we call Design ethnography. This work was initiated in the previous three years at Intel Corporation and Tektronix Laboratories, and in the last year, has blossomed into a valuable element of the development process for new products at Intel. We have presented the results of our studies to senior management, have affected the long term and strategic planning process, and have assisted in redirecting product marketing efforts.

Much of the history of CHI has been the movement of focus from later stages of design into earlier stages. Energy spent testing an almost finished product was redirected towards iterative, prototype based design and cognitive walkthrough techniques. This energy was eventually refocused on requirements elicitation techniques for determining the abstract features and functions of a product. While having CHI professionals intimately involved during the requirements phase is certainly a positive step, many strategic decisions have already been made. Someone (marketing? engineering?) has already decided what general class of product (e.g. something to do with publishing on the web) will be built. But what if this general class of product just doesn't fit into the lives of the people in your market? All the requirements analysis, iterative design and prototyping, and usability testing in the world won't make this product successful. Design ethnography continues this evolution of CHI's focus to even earlier stages of design; in this case to the point when a general market demographic has been selected as important (e.g. families with children) but before any product concepts have been conceived. Just as CHI professionals should be heavily involved in requirements, they should also be intimately involved in the strategic decision process by which a company decides which markets to enter or create. Design ethnography is an approach by which a CHI professional can make valuable contributions to the strategic process of defining a company’s products.

TUTORIAL CONTENT

We will briefly define what design ethnography is and what it isn't. We then present three case studies of work (families with young children, teenagers and business communications) we’ve done in which we describe the process of conducting the study from start to finish and the methods we used and the results of the studies. The balance of this half day tutorial is comprised of discussion with the audience.

In the CHI community, “ethnography” has come to mean any sort of qualitative field work used to gather requirements for a product. This is an unfortunate dilution of a very powerful set of concepts and while field work is certainly valuable, most work we’ve seen claiming to be ethnography is simply not. Traditionally, ethnography refers to a set of methods and techniques used primarily by anthropologists in their field work. In our approach, we usurp many elements of traditional ethnography, but bend, twist and transform them to suit our purpose: to define and design new products for our company.

Therefore, we strive even further to distinguish our approach from that of the traditional ethnographer, at least so much as it can be distinguished. We discuss some of the assumptions we make and how they differ from more traditional ethnographic approaches. In addition, through out the three examples, we will point out where these differences are meaningful in the analysis.

We also distinguish design ethnography from requirements analysis, which is seen particularly well in the business example. The main difference is that methods and techniques more in line with design ethnography are better employed before there is product concept in mind, whereas requirements analysis techniques are more suited for exploring a particular product concept.

Qualitative field methods produce an embarrassment of riches. How does one take the hours of video tape, photographs, field notes, drawings, and artifacts produced by informants (e.g. friendship networks), and synthesize this into a model of a domain? Analysis is in fact the crux of ethnography and is what separates it from field work. The notions of intermediate representations and notations, tentative hypotheses, and model search guided by design efficacy are defined and described. In addition, the important role played by graphical models in our analysis method is described and illustrated.

All through the tutorial, general issues in designing an ethnographic study, designing elicitation methods, and analyzing the field data to produce a model are illustrated through the three case studies.

Example 1. The family model describes a typical day in the life of a family with respect to the analysis dimensions of space, time and social communication. New product concepts which fit into family life are derived from the model. In addition, a critique of the “box” model of computing (box, keyboard, screen) with respect to family life is offered.

Example 2. Teenagers (seniors in high school) are described as belonging to one of three communities: the One Community Teen, the Two Community Teen and the Fringe Community Teen. These communities are defined by the difference in teen life along the dimensions of space, time, people and activities. This model has an efficacy for developing specific teen design concepts which the standard marketing taxonomy lacks -- usually some form of early adopter, conformer, and "doesn't care".

Example 3. The business communication model describes the basic communication patterns of groups working together in organizations. It demonstrates high and low probability communication events and helps guide product development appropriately.


CHI 97 Prev CHI 97 Electronic Publications: Tutorials Next

CHI 97 Electronic Publications: Tutorials