Survey layout
Introducing the survey
Whether you’re talking to a customer face-to-face or using an online tool, always introduce your survey. Make sure the participant understands what the survey is about, but also explain the nature of the survey.
You should at least mention:
What your survey is about. Don’t suppose people immediately know what you’re talking about. Be very clear and precise about the product, service or category. You’re not asking about milkshakes, you’re asking about the milkshake that the customer is drinking during her morning commute.
Who the survey is for. Customers have different roles. One day the same person is taking her kids to a restaurant for breakfast and the other she stops by on her way to work. Be precise in what role the customer should imagine themselves in when answering the questions.
Indicate what aspect of the product you’re asking customers about so that they are mentally at the right level of abstraction. Introduce the survey by saying that you are thinking of making the milkshake longer to finish. Even if a customer doesn’t like that a milkshake takes longer to finish, knowing that you are only investigating ways to do so will make her answers reliable.
Never try to sell the product in your survey. Be as factual as possible; you must not influence your customer’s attitude.
So, don’t do this:
We’re working on an exciting new dairy product and we’re sure you’ll love it! Please answer these questions:
But do this:
We want you to enjoy our breakfast milkshakes longer during your morning commute, so we would like to hear your thoughts on these ideas for improvement.
A Kano survey looks odd to participants who haven’t encountered one before. Take your time to explain that:
Each question is in fact a set of two questions, each relating to the presence or absence, or sufficiency or insufficiency of a feature;
The order of the answers are not a ranking or a scale. Clearly indicate that the default answers reflect a classification. If necessary, explain the exact meaning of each answer by giving example answers.
You should not explain what you’ll do with the answers (i.e. the categorization); you want to avoid participants overthinking things or trying to fit their answers to what they believe is the right category.
Get to know the participant
While you are introducing the survey, get a feel of the participant's relation to your product or service. Find out what they aim to get out of it, and find out about the broader context too.
For instance, when surveying distributors about a B2B commerce platform, find out about their own business. Is theirs a large business? Who do they sell to? How do they sell? Are they a technologically advanced operation?
Next to that, find out what the participant's skin in the game is. Their skin in the game is the combination of what they aim to achieve using your product and their level of involvement.
Distributors may be looking for efficiency gains, better traceability of their orders or being able to order things outside of office hours. Skiers may be looking for more speed or enjoyment.
Next to their goals, find out about the participant's level of involvement. Are they a frequent user? Are they experts or novices? Expert skiers will have different expectations from skis than novices. Distributors who work with many suppliers have more experience with ordering platforms.
How important do they think their reason for using your product is? People who really want to lose weight will have different attitudes towards appetite-suppressant capsules than people who aren't serious about it.
Knowledge about the participants' skin in the game will benefit your study. You'll better know why features are in the categories they are. You'll also be able to detect what segmentation attributes are relevant.
If you cannot do the survey live, add a section to the survey form where participants can tell you who they are. Formulate your question so that you will understand why the participant uses your product (their "consumption goal" as Mittal (2001) calls it) and what their level of involvement is (e.g. usage frequency, importance of the consumption goal, level of expertise, ...).
Question and answer layouts
The most used survey layouts are:
and
These layouts still give the impression that the answers are part of a scale. This can be mitigated by explaining to the surveyee that they aren't, but still. Unfortunately, I haven't found a perfect layout yet. I once tried dotting the answers along a wheel shape, but customers found that confusing.
Order of the questions
You don't have to keep a strict order to your questions. Some researchers (like Lee & Newcomb, 1997) shuffle the functional and dysfunctional questions so that they are not in sequence.
I even encountered one study (Chen, 2020) where
The functional questions were asked first and then all the dysfunctional version;
Different labels were used for the functional and dysfunctional answers;
The order of the answers was reversed for the dysfunctional questions.
It looked something like this:
A "pre-test questionnaire was administered to ensure respondents understood the content, and sentence- or word-level modifications were then made to improve its clarity. The final questionnaire had content that was more appropriate and is considered to have significant reliability." (Chen, 2020).
Although I haven't seen it elsewhere yet, as long as respondents understand the survey as they should, Chen's approach is perfectly suitable.
Do not number the answers
You are aware by now that the possible answers reflect feelings, not ratings. It bears repeating that you should make sure the survey participant understands that too. Therefore, never number the answers on your questionnaire.
Last updated