
Don't limit your evaluation of customer satisfaction to a few representatives -- evaluate all influential people. 

Don't add too many questions to the questionnaire, so as to keep it easy for the stakeholders to respond. 

Even when you do your best to keep the questionnaire anonymous, if there are only a few participants, some of their responses may be identifiable. If this occurs, try to ignore those personally identifiable items, and never use them in the future. Otherwise, some stakeholders may not trust the anonymity of the evaluation. 

To ensure that you won't see personally identifiable information in the event that it does exist, focus on aggregate data and do not review the answers individually. 

