The ultimate guide to the Kano Model: value through the eyes of your customer
How often does your product team assess the value of a feature through something other than business value? And how do we relate business value back to the end-user’s perception of value? Often we weigh business value versus development effort (think: value vs. effort). But how can we include the end-users’ perception in this to make sure we are doing the right things?
As our product design team was about to embark on a design challenge. We realised we needed better insight into what brings real value to our end-users. So we can guide the product team into a direction of customer-centric solutions.
Let’s enter into a case study of the Kano Model and how it blends the user’s voice into determining value and prioritisation of proposed features.
What is the Kano Model
The Kano Model, also known as the kano scoring model, is a framework designed to prioritise features based on the degree to which they are likely to satisfy and delight customers. It was developed by Dr. Noriaki Kano, a Japanese researcher, and consultant in 1984.
The Kano Model is used to analyse consumer preferences for different features and group them into multiple categories. So, when we’re developing a new product we know what features to focus on. This technique weighs a high-satisfaction feature against its implementation costs to determine whether to add it to the product roadmap, disregard it entirely, or adjust its priority.
It also measures how customers will react to certain features. This distinguishes it from other prioritisation frameworks like story mapping, weighted scoring, RICE, opportunity scoring, and the MoSCow method. As Jared Spool points out in this article, users’ expectations change over time. They are exposed to better and more sophisticated digital experiences as companies prioritise user experience and technological capabilities grow.
Case Study; how we applied the Kano Model
Now back to the design challenge. CarePay’s design team wanted to improve the way users can browse and filter healthcare insurance products on our retail website. As the amount of products by different insurers grew, it became clear that we needed a new way for users to find products easily based on their individual needs. You might think of this as a straight-forward challenge with many best practices available. This is true. But we wanted a user’s perspective to guide us in prioritising features.
Through previous qualitative studies (semi-structured interviews), we learned that users are interested in product specifics that are not commonly found on insurance-comparing websites. This showed us that we might need to go beyond the ‘usual features’ to serve their needs. So, we did additional UX research using the Kano Model to ensure we move forward with the right ideas.
Creating the Kano survey
Let’s discuss the basics of creating a Kano survey. It consists of a functional and dysfunctional question to understand how a user would feel if a feature would be available versus if it wouldn’t be available. This way you’ll measure their response and know how valuable this feature is from a user’s perspective.
Our first step was to get stakeholder input to make sure we’re asking the right questions and capture any ideas top of mind to use in our Kano survey. The design team organised a co-creation workshop and invited business stakeholders. User related findings from previous qualitative studies were used as input for the workshop. From there, ideas were generated and categorised by topic and then prioritised (through dot voting). As we didn’t want the survey to have too many questions, we needed to carefully weigh which question was really worth asking. The previous qualitative studies helped us in doing that.
After the workshop the product design team summarised and analysed the findings. From there the proposed features were chosen and the questions formed. The most important proposed ways of users browsing healthcare insurance products are listed below.
Target audience
The online survey was solely targeted at past visitors of CarePay’s health insurance retail website. We did this for several reasons; mainly to make sure that:
- people have previously been exposed to health insurance as a concept.
- people have recently been in the state of mind of orienting themselves around health insurance products.
- people could relate it back to their own experience during their past visit of our web site and think about what could make it better.
We randomised the order of the questions in case participants would drop off prematurely and collected 60 responses in total.
As responses were coming in, the drop off rate actually positively surprised us as most people completed the entire survey. Although we can only guess why — we had a hunch that this might be because they’re familiar with our brand name and organisation based on their past visit to our retail health website.
Analysing the results
Analysing Kano survey responses is not simple. This is mainly because each response needs to be individually weighed in the context of the paired question (i.e. functional and dysfunctional variation of that question). Answers where only one of two are answered, cannot be weighed.
To help with analysing the responses, we used a free Kano survey Excel template. We filled in the responses and the Kano categorisation of features would appear — i.e. Must-have, Attractive, Performance, Indifferent, Questionable.
When looking at the survey findings we found a lot of inconclusive results. A lot of features were marked as ‘Questionable’, as shown below.
After analysing the data we found that a handful of respondents with vague answers created a lot of noise in the findings. So we asked ourselves. Would the results shift significantly if we simply removed those responses from the data? This is what happened next.
- From 9 features marked as ‘Questionable’ to only 1 feature.
- From 4 features marked as ‘Indifferent’ to only 1 feature.
- From 3 features marked as ‘Performance’ to 8 features.
As you can see, the clean up of the data made a huge difference in the findings! We started out with inconclusive findings. But by removing a couple of ambiguous responses, we got better insight into which features were perceived as valuable by the user.
For this reason I would recommend taking a closer look at the individual responses and removing data points which cause noise and disturb your findings. In our case there were respondents that selected the first answer for all the questions. So liking if a feature would be available, while also liking it if it would be unavailable. In a pool of only 60 respondents. This can make a huge difference.
In our case it was surprising that being able to filter/browse by ‘Price’ was a feature that left users feeling indifferent. This is not even considered as a ‘must-have’ when you look at the results. On the other hand, being able to reach customer support directly online during the orientation phase of the purchase process is really important to them. It also confirmed that customer reviews or ratings on products are really important to people. They would like to use it during the orientation phase. These findings really made us reconsider which features to prioritise on our roadmap.
Key takeaways from our Kano Model analysis
The Kano Model is simply the categorisation of attributes. It generates output from discrete analysis, as well as continuous analysis.
- The results from discrete analysis are calculated by looking at each individual’s responses to the questions. From here we can see how each attribute is categorised on average. If ambiguous responses create noise and lead to inconclusive findings, you might want to consider removing these data points.
- Continuous analysis graphically classifies each feature. Do this by plotting the average functional and dysfunctional score (between 0 to 4) for each attribute. The higher the functional score, the more important the feature is to the respondents. The higher the dysfunctional score, the worse it is for them if the feature is not present.
- The Kano Model can help product teams get more insight into the user’s perception of value and use it as input during the prioritising process of features. This opens up opportunities to go beyond ‘gut feeling’ or a business stakeholders’ perspective when defining value. In our case it meant being surprised by the results, as we discovered our end-users prioritise differently than we expected.
Thanks for reading through our journey on how we applied the Kano Model as a UX research method. Please leave a clap 👏 if you found this article helpful and comment how you would plan to apply this method yourself. We would love to hear from you!
This article was written by Eva Hörner, Associate Director UX at CarePay & Joshua Ariga, UX Researcher at CarePay
