Ethnography comes from social research and anthropology. Instead of asking people, what they believe or what they have done as we do in survey research, we watch what they do. Ethnography is a popular technique for the study of human-computer interaction (HCI). There are variations in how data is collected: 1) sending an observer-researcher into stores, workplace, and homes to observe real behavior; 2) giving subjects a video camera to record their interactions with the target product [photo-ethnography]; or 3) having subjects record their behavior in a diary. Some researchers will use such diary recording as an opportunity to tap into consumer attitudes and buyer beliefs as well. As with any research technique, ethnography has its measurement validity problems.

Ethnography advocates say that watching behavior is preferred to that of asking questions. The argument is that there is higher response validity, and item validity, when we watch rather than ask. The ideal in survey research sometimes preferred and meetings of the American Marketing Association, the Advertising Research Federation, QRCA, MRA, and others is that the ideal survey would be "getting the answer" without asking the question. Again, ideally, this would avoid interviewer bias and enhance research validity. Likewise, with ethnography market research, the ideal would be to watch people use products or respond to ads without them knowing they are being watched. Giving them cameras, as we do with photo ethnography, does not solve the problem: consumers or product users "know they are watching themselves." Therefore, we have found that nothing is perfect, even ethnography. We have the potential for certain problems with "observational validity" with this and most other observational methods. The problem goes away when the observations are fully unknown by the research subjects. In some respects, long-standing store-aisle hidden observation research is by far purer than ethnography in that ethnography usually has a known observer showing up, living with subjects, or taking pictures. Such intrusion will create unknown researcher-subject interaction bias. Yet, perhaps, we're pole-vaulting over a marshmallow here.

Properly used, ethnographic studies can yield exceptional insights useful for product redesign, new product concepts, discovering alternate product uses, and profiling true customer satisfaction. On balance, we view ethnographic studies a useful tool. Power Decisions Group offers this observational technique as a part of a wide range of market research tools to use to help our executive and managerial clients make important strategy decisions.

See also our "Tools" discussion.