Well, at this point you are left guessing how respondents would react to levels never shown them in the conjoint exercise. Maybe the client could provide a panel of experts at the client company that could take the CBC questionnaire that includes the full list of levels. Or, maybe you could show them the utility scores for the other levels in the study and they could tell you their best guess about the relative position of the missing levels. That way, the information about the missing levels could be provided by a "panel of experts" at the client firm? You could then use that information to impute new utilities for the existing respondents?
