When creating a survey, we recommend you consider whether or not you will be using the survey several times in order to benchmark the results.
For example, Enalyzer's employee engagement survey will be launched in 2016 and 2017 and we want to benchmark those results in the same report. To be able to do so, we have make two separate launches in the same survey for 2016 and 2017, instead of creating a new survey or a relaunch for each year.
After the initial launch there is a few alterations you can make to your survey, therefore the decision to benchmark survey results needs to be made before you start designing your questionnaire. These alterations are changing wording, adding a new question, or hiding existing questions.