Predicted Preference Conjoint Analysis
In this paper we propose a new method of eliciting market research information. Instead of asking respondents for their personal choices and preferences, we ask respondents to predict the choices of other respondents to the survey. Such predictions incorporate respondents’ knowledge of their peers, which presumably reflects their social network. The effectiveness of this approach has already been demonstrated in context of political polling. Here we extend it to market research, specifically, to conjoint analysis. A theoretical argument demonstrates that predictions should yield utility estimates that are more accurate, and less noisy than estimates based on own stated preferences. In addition, we show that our results remain valid even when the information from social circles is biased, provided that the net bias across individuals is zero. These theoretical results are confirmed in three online experiments.