Your research survey data is only valuable if the questions you ask reflect exactly how your respondents feel, without qualifications or reserves. Asking the right questions seems easy on its surface, but remember that you have to accommodate the unique predispositions of all of your respondents in order to get valuable feedback.
From question type to the language you use, every component of your survey question has an impact on the customer and thus on your feedback. Even when navigating seemingly straightforward and default questions about your respondent’s age and gender, you must be extra tactful and sensitive. Without taking the time to carefully plan your survey questions, you run the risk of creating survey fatigue for your respondents or, even worse, offending them.
The following 8 do’s and don’ts demonstrate how you can create excellent questions that deliver actionable results.
1. DO use scales where possible
Sometimes “yes or no” doesn’t give you the full picture. Alternatively, asking your respondent the extent to which they agree with a given statement or how likely they are to recommend your business to a friend can give you greater insight into the intensity of their opinions. There is a big difference between the respondent who says that she “Strongly Agrees” that their last purchase was satisfactory versus one who only “Slightly Agrees.” Because scales allow you to expand beyond basic percentages, you can apply more complex analysis concepts like t-tests and ANOVA.
2. DON’T stuff multiple questions into one
Even though you want to avoid burning your respondents out with a 20-question survey, the solution is not to ask multiple questions at once. This results in potentially confusing data. For example, consider the following question: How likely are you to purchase our product again or recommend it to a friend? A respondent could plan on purchasing the product again without recommending it to a friend or vice versa, thus making your data virtually useless.
3. DO be sparing with open-ended questions
Qualitative data can be just as valuable as quantitative data, but more often than not they produce inactionable results or exacerbate survey fatigue. Don’t ask an overly generic question like, “What kinds of products would you like to see us provide in the future?” This not only will deliver myriad different results, it could also raise respondent expectations beyond what you can reasonably provide.
4. DON’T underestimate psychographics
Psychographics uncover the “why” consumers behave the way that they do. It asks bigger questions about their needs and personalities. These kinds of questions enable you to make deeper, more meaningful connections with your target market. Buyers typically aren’t just looking for a product or service; they want to be a part of something bigger than just material things.
5. DO use clear, straightforward language
Even though your target market might be an intelligent bunch with a large vocabulary, avoid fancy, high-level language when it comes to surveys. Additionally, make sure to avoid words with multiple meanings to avoid confusion. Keep it simple and direct.
6. DON’T forget to test your survey before you send it out
Your survey should take between 5-10 minutes to encourage more respondents and to prevent survey fatigue. Testing your survey first ensures that it not only falls within this time frame, but also generates the responses you’re looking for. This is also an excellent method for avoiding confusion and/or outright errors on your survey.
7. DO consider how you will report and export the data ahead of time
How will you analyze and report your data once the surveys are all complete? Sometimes the results themselves will answer this question for you, but you always want to have a backup plan. Set up your questions with a specific kind of analysis in mind, and make sure that each of your questions contributes to this. You can include a few wildcards in your survey to explore new areas, but keep the majority focused on your pre-planned analysis.
8. DON’T overlook the bias of question order on your research survey
Sometimes the order of the questions asked can skew your results. To avoid the bias that can emerge from question order, randomize your survey questions so that they appear differently for each respondent.
Now that you have all the answers, it’s time to write the questions. Which one of these tips will you incorporate in your next research survey?