One of the toughest things in developing a good survey is making sure it isn't too long. When you decide to do a survey and others in your company find out about it, you'll find that you have lots of requests for questions to be added. What starts out as '10 Questions to Ask Customers' can quickly become '83 Questions to Ask Customers'.
And that's a problem. Too many questions result in people dropping off (so do difficult or poorly worded questions - a topic for another day). Or they get tired and don't think clearly about their answers, and rush through the responses just to be done.
Here are some guidelines for how long your survey should be. Keep in mind it varies based on the relationship you have with the people you're surveying, the types of people they are, the frequency with which you survey them, and the way you're doing the survey (online vs telephone vs in-person):
- 5 - 7 questions is a fantastic length for a survey, if you can stick to it. You'll have almost 100% completion.
- 8 - 15 questions is good. You'll usually have 5 - 10% dropoff. (And by 7 - 15 questions, that includes all the sub-questions that might not get numbered separately, but require a response).
- 16 - 30 questions is getting more serious. You'll see 10 - 20% drop-off.
- 30 - 50 questions is serious. You'll need a good relationship with your participants or be compensating them in some way;if not you'll see 20 - 30% drop off.
- Above 50 questions, the drop off rate increases exponentially. By the time you reach 100 quesions, you'll hit about 70% drop off (again, unless you are compensating them or you have a significant relationship with them).
When you're developing your survey (especially if there are multiple people with input to it), the very best sanity test is to ask yourself the simple question: would I complete this survey? And be honest with your answer. If the answer is no, keep chopping questions. Your survey will be better as a result.