For surveys to reflect customer opinion accurately, they need to be reliable and valid. Surveys that are reliable get equivalent, consistent results on repeated trials. Equivalency shows the correlation of the variables in the trial. Consistency is the extent to which the results correspond. The responses of repeated trials indicate the stability of the test. Surveys that are valid show the accuracy of the survey’s measurements. External validity indicates terms could be understood or interpreted by the respondents in the same way. Internal validity takes into account the lateral decisions that were included or omitted in their measurements.

It is important that survey instruments are reliable and valid so that the surveyors focus on specific perceptions of the respondents. The implementation of a survey is much like the production of a product. There are several steps required in running a survey, including identifying the customer requirements, developing the survey, implementing it, and analyzing it. Since these steps are linear, the results of the following stage are dependent upon the quality of the product produced in the prior stage. As analysis of the survey results follows the implementation of the survey, the reliability of the analysis is dependent upon the implementation. A poorly implemented survey would gather irrelevant or inaccurate information. The analysis of the feedback would be skewed and useless to the goals of the survey. A poorly implemented survey may be the product of the prior stages in the survey cycle. The survey questions may not have been formulated properly, or the customers’ requirements may not have been identified accurately. Each component in the survey process, including the reliability and validity of the survey instruments, contributes to the success of the survey.