Questionnaires are effective business tools for gathering demographic information, facts, personal opinions and attitudes from your clients and customers. One of the greatest benefits of questionnaires lies in their uniformity — all respondents see exactly the same questions. Whether you are compiling a questionnaire to measure sentiment for a new product rollout or gauging consumer satisfaction at a retail store, understanding the attributes of a good questionnaire is a necessary first step to producing the responses you need.
A good business questionnaire should be no longer than necessary to obtain the information you want to collect. Consider the delivery avenue — a business reply card, an online survey or an extensive multi-page form — and where the respondent will complete it. The topic of the questionnaire should be relevant to the respondents and their experience with your company or product, otherwise there is little incentive to complete it.
Your questionnaire should have a clear purpose, directions that are easy to understand and well-crafted questions that make sense to the respondent.
Questions should be general at first and progress in detail. The content should transition logically from one topic to another. Avoid asking your clients sensitive questions; if you must, position them toward the end of the questionnaire and make the questionnaire anonymous to get truthful responses.
The questions you ask can take two forms. Restricted questions, also called closed-ended, ask the respondent to make choices — yes or no, check items on a list, or select from multiple choice answers. Unrestricted questions are open-ended and allow respondents to share feelings and opinions that are important to them about your products or services. Restricted questions are easy to tabulate and compile. Unrestricted questions are not, but they allow respondents to reveal the depth of their emotions.
If your objective is to compile data from all your respondents, then stick with restricted questions that are easily quantified. If you wish to study degrees of emotions or depth of sentiment, then develop a scale to quantify those feelings.
Each question should address only one topic. “When you go out to eat, do you have wine and after-dinner coffee?” from a restaurant owner does not differentiate between those who chose one or the other but not both. Be careful to address all options with multiple choice questions. With a question like “Do you plan to try our product,” a yes or no answer choice does not give the respondent the option of saying that he might under certain circumstances — and those circumstances are important to you.
Multiple choice answers must be unambiguous and mutually exclusive. There can be no overlap that could blur the choices. If you ask respondents to rate situations, be sure to give them a reference point. “On a one-to-five scale, how do you rate the customer service response you received from our company compared to your previous provider. 1 is the lowest and 5 is the highest."
Pew Research Center, which compiles many questionnaires, sometimes pre-tests questions open-endedly to see what answers people give, before using the most common answers as multiple-choice options. For example, you could ask employees, "Who do you turn to most often with questions about your job?" If the most common answers were a co-worker, my immediate supervisor, the person in the next cubicle, and my spouse, those answers would be A, B, C and D options for that question on the questionnaire. Order of answers can influence response so you may want to vary up answers just in case.
Discuss and share the questionnaire with co-workers so your work will produce the results your business needs. Determine the information you seek before formulating the questionnaire. Because you are compiling answers to gain useful information, give consideration to how you will tabulate the data you receive. If the questionnaire will be machine-read, then the layout must conform to the machine's capability. Open-ended questions require your subjective interpretation.