Whether you are designing surveys using our library of existing templates or starting from a blank page, your goal is to get as many qualified responses as you can. We have gathered our best practices for survey design below, to help you ensure that your survey hits the mark.
Define clear question objectives:
Without a purpose for your questions, results become unfocused and difficult to act upon. Therefore, start your survey design process by formulating 4 - 6 study objectives containing clearly defined goals such as: 'What positive and negative aspects do customers highlight when navigating the website?' Or, 'What can be done to improve the website?'.
Make sure your objectives are specified as much as possible. For example, an objective like: 'What drives customer satisfaction?' is too broad. Instead write: 'What are the 3 main factors contributing to customer satisfaction?' This will help you focus your answers on 3 main action points.
Start with ice-breaker questions:
A key component of designing good online surveys is to get respondents to open up - especially since an unmoderated study can feel a bit odd to some people. This can be achieved by starting your survey with a few easily digestible questions such as: 'What are your main hobbies?', or, 'How does a typical every day for you look like?'.
In addition, make sure that the purpose of the survey is clearly described and that respondents are made aware of what their answers will contribute to.
Start broad and drill down:
Questions are best posed if you let respondents talk freely about what is most important to them without your intervention. This is done by starting your survey with broad questions such as: 'What is your immediate impression of the website?', or, 'looking at this marketing material, what are the first 3 things that you notice?' This allows you to pinpoint what respondents themselves think is important. If you want feedback on specific elements of your website/campaign - then save it for later in your survey.
Keep your language simple and specific:
Your goal is to have every respondent interpreting the questions you ask in the same way. If your language is too complex or your questions are too broad, the answers you get may end up in several different directions. Always read a question out loud to yourself and others to ensure that it addresses what it should.
In addition avoid the use of jargon language, lingo, or any specific wording that might confuse the respondent.
One idea = one question:
By asking several questions in one sentence, you risk confusing the respondent and/or getting incomplete answers. To avoid this, always consider if a question is trying to achieve too much, and strive not to use the conjunction 'and' when formulating questions.
Bad example:
'What is your immediate opinion about the website and what elements do you like the most?'
Good example:
Question 1: 'What is your immediate impression of the website'?
Question 2: What elements do you like the most?'
Avoid leading questions:
Leading questions lead respondents to provide certain answers by including non-neutral wording or preconceived opinions. As a general guide to avoid leading questions make sure to:
Do not ask affirmative questions:
Bad example: 'Do you agree that the webpage is easy to navigate?'
Good example: 'How do you find it navigating the webpage?
Do not plant ideas in respondents' heads:
Bad example: 'How positively would you rate the customer service at company x?'
Good example: 'How would you rate the customer service at company x?'
Do not assume that you know how the respondent is feeling:
Bad example: 'What was difficult about completing this task?
Good example: 'When completing this task, what was happening?'
Ask open-ended questions including 'why', 'what' and 'how' etc:
The idea behind doing qualitative research is to identify new opportunities from respondents. You do this by asking open-ended questions that allow respondents to elaborate on their opinions. As a general rule of thumb, if a question can be answered with a 'yes' or a 'no' it is not open-ended. Always phrase your questions using words like 'why', 'what', 'how', describe', and 'explain' to get in-depth answers.
Bad example:
‘Has Covid-19 affected your work tasks, and do you think about it during your workday?’
Good example:
'How has Covid-19 affected your work tasks? (if it has affected your tasks in any way). Please elaborate and explain what has happened'
Keep your survey brief:
We recommend your survey include up to a maximum of 30 questions (20 - 30 min. approx.) to reduce the risk of drop-offs and respondents providing less elaborate answers towards the end of your survey. If your survey is longer, consider if you can leave something out by asking yourself the following questions:
Are all of my questions relevant to my study objectives?
Are there overlaps between some of my questions?
Should I run two separate surveys?