However, surveys as a measure of performance have always been a controversial issue. There are solid reasons for the same. In any case, simply conducting a survey is just an activity and does not add to performance. Therefore, it cannot be an effective key performance indicator (KPI). So, to say the least, improvement in survey results may be taken as a measure of performance and not merely administration of a survey. But that is not enough. Do you know that there are a few more careful considerations required at the initial stage to convert surveys into effective KPIs?
The first and the biggest issue with surveys is periodicity. When surveys are conducted, say once in a year and the outcome rating taken as a measure of performance, it serves absolutely no purpose. It gives you no time and opportunity to make amends and course-correction to improve performance. In other words, it is “too little, too late”. The solution lies in making it a more frequent exercise, which may not always be possible or desirable. However, administering surveys at least twice in a year if not more may be a workable option. In case even that is not possible, forget about it and look for other measures!
Cascading down exactly the same performance measure should be resorted to only by exception. Fortunately, there is no such problem in case of surveys since these always consist of a number of parameters. For cascading down, you can always pick the most relevant depending on the functions, levels, roles and responsibilities. Tremendous flexibility and options are available on this account.
Another argument against a survey may be that it is a measure of perception and hence subjective. If it was really so, any survey would lose its meaning. While it is true that individual respondents would have biases and perception, this is also a reality that when a survey is conducted across a large population, collective subjectivity leads to objectivity. As the population size goes up, any subjectivity gets evened out. It is also said that in case of voluntary participation, only the people with strong views only participate. That should not be a cause of concern even if majority of the respondents consist of the people who are either highly satisfied or highly critical. That should be acceptable because the consolidation gives you enough indications whether the majority is on this side or that. And that itself is the measurement of performance. It gives you enough inputs to make an effective action plan to shift in the desired direction.
One real issue can be that of the sampling. Quite often, straightway some percentage is chosen as the sample size. Statistical principles suggest that the ideal sample size, if the population is small, would be equal or almost equal to 100%. Therefore, it is desirable to use sound statistical principles for determination thereof rather than blindly picking up say 10% or 20% as the sample size.
By taking care of these important issues, you will find that surveys can certainly prove to be a very effective measurement of performance.