Best practices for survey questions
Here are a few best practices to get you headed in the right direction when building your surveys:
1. Have only one well-defined objective
What feedback do you really want from your customers? Narrow down what you want to know to a few well-phrased and specific questions. Take a few minutes to consider what you want to get out of the results before building your questions. This will help you avoid the “just one more question” syndrome when building your survey.
2. Keep your survey short
A survey should take 5 minutes or less to complete in order to get the best response rate. Avoid questions with more than 10 answer options. Too many answer options will confuse the survey taker and slow them down. If you find you need that many answer options, break the question into multiple questions or consider another way to phrase or answer it.
3. Formulate your questions for measurable results
The key is to ask closed-ended questions that generate results that are easy to analyze, and from which it's easy to spot trends and set baselines.
Using too many open-ended questions that require respondents to type out responses (i.e., “What do you like about our services?”) will take a lot of time to complete and evaluate.
4. Ask about one thing per question
Avoid such inquiries as “Do you like oysters and baseball?” These are two unrelated subjects and should be broken into separate questions that will allow survey takers to answer each independently. Give respondents the option to answer the question with “Other”, “Prefer Not to Answer”, or “Not Applicable” where necessary. These answers, while vague, will provide better insight than a non-answer.
NOTE When giving an option of “Other,” it’s useful to provide a text box for write-in answers.
5. Don't bias the response
Ask questions in a manner that doesn’t trend answers in a particular way. For instance, when gauging a customer’s likelihood of referring friends to your business, use a scale that ranges from “not very” to “extremely” with a few options in between. Avoid using “Always” or “Never” extremes as they can bias responses in the opposite direction.
6. Limit required questions
Too many required questions in a survey can ruin the flow and are likely to decrease your response rate. The required questions should only be those that you absolutely need in order to make business decisions.