A proper, easily-flowing sequence of questions is another important characteristic of a good survey and should be labored for. Finally make sure that you pretest the questionnaire with a small group of people so that any flaws or glitches can come to light before proceeding on to the actual survey.
Medium to Use
The medium used is also just as important as the survey instrument. Each has its advantages and disadvantages. Email or an online survey for example is advantageous in the sense that it is cost effective and can reach thousands of respondents in very little time as compared to mail surveys. However email or online surveys are for those who have internet access and that might not be true of some representatives of the population you wish to survey. Moreover, only people who are free or feel strongly about the subject might fill it out leading to skewed results.
On the other hand, posting surveys via mail can be costly and might not turn in as many respondents even if the return postage is paid by the surveyors. Phone surveys again are not as costly and give immediate responses but these are to be used with caution. If for example you want the opinions of working people and you call them in normal working hours, they will not be available. For a random sample the bias of unemployed people participating in telephone suveys exists as they do it to kill time.