The Kano Model is a framework for classifying product features based on how they affect customer satisfaction. Developed in 1984 by Japanese professor Noriaki Kano, it challenges the assumption that every feature improvement leads to a proportional increase in satisfaction. Some features delight customers when present but cause no frustration when absent. Others are the opposite: invisible when they work, but deal-breakers when they don’t. The model gives product teams a structured way to tell the difference.
The Core Idea Behind the Model
Most product decisions assume a linear relationship between quality and satisfaction: the better a feature works, the happier the customer. Kano’s insight was that this only describes one type of feature. In reality, different features follow entirely different curves. A hotel room having a bed is not the same kind of feature as a hotel room having a welcome basket of local chocolates. Improving the bed won’t generate excitement, and removing the chocolates won’t generate outrage. They operate on different psychological tracks, and treating them the same leads to poor prioritization.
The model maps features along two dimensions. The horizontal axis represents how well a feature is executed, from not implemented at all to implemented extremely well. The vertical axis represents customer satisfaction, from total dissatisfaction to total satisfaction. Different categories of features trace different curves across this space, which is what makes the model useful.
The Five Feature Categories
Must-Be (Basic Expectations)
These are the baseline requirements customers take for granted. They fulfill the basic functions of a product, and customers treat them as prerequisites. A phone that makes calls. A car with brakes. A website that loads. When these work, nobody notices. When they fail, satisfaction drops sharply. You can’t win customers by excelling at must-be features, but you will absolutely lose them by falling short. On the Kano graph, this curve sits mostly below the satisfaction midpoint: increasing execution only brings you from dissatisfied to neutral, never to delighted.
One-Dimensional (Performance Features)
These are the features where more really does mean better, in a roughly proportional way. Battery life on a laptop, storage space on a phone, fuel efficiency in a car. Customers can articulate these needs clearly, and they actively compare products on these dimensions. The relationship is straightforward: better execution produces higher satisfaction, worse execution produces lower satisfaction. These features can both satisfy and dissatisfy depending on how well they’re delivered.
Attractive (Excitement Features)
These are the surprises. Features customers didn’t know they wanted, but that generate outsized delight when present. Their absence causes no disappointment because nobody expected them in the first place. Think of the first time a phone unlocked with your fingerprint, or the first time a streaming service auto-played the next episode. On the Kano graph, this curve sits mostly above the midpoint, rising steeply with execution but never dipping into dissatisfaction. Attractive features are what generate word-of-mouth buzz and set products apart in competitive markets.
Indifferent
Some features simply don’t move the needle. Customers don’t care whether they’re present or absent, well-executed or poorly executed. These are important to identify because they represent wasted development effort. If a feature falls into this category, resources spent improving it could be redirected toward something that actually matters to users.
Reverse
Occasionally, a feature that a product team assumes customers want actually causes dissatisfaction when present. This can happen when a feature adds unwanted complexity or when different customer segments have opposing preferences. A power user might love advanced settings; a casual user might find them confusing and frustrating.
How Kano Analysis Works in Practice
The model isn’t just a conceptual framework. It comes with a specific research methodology. For each feature you want to evaluate, you ask customers two questions: a functional question (“How would you feel if this feature were present?”) and a dysfunctional question (“How would you feel if this feature were absent?”). Each question offers five response options, ranging from “I like it” to “I dislike it,” with neutral options in between.
By cross-referencing the answers to both questions in an evaluation table, each response gets classified into one of the five categories. Someone who says “I like it” for the functional question and “I’m neutral” for the dysfunctional question is signaling an attractive feature. Someone who says “I’m neutral” for functional and “I dislike it” for dysfunctional is pointing to a must-be. The classification emerges from the pairing, not from either answer alone.
You repeat this across your full sample and tally up the results. If 70% of respondents classify a feature as must-be, that’s your signal. Typical Kano studies use between 50 and 300 participants, which provides a margin of error between 5% and 9%. The right sample size depends on how much precision you need and how many distinct customer segments you’re trying to understand.
Reading the Results
Beyond simple category tallies, teams often calculate two coefficients to quantify the impact of each feature. The “Better” coefficient (sometimes called the satisfaction coefficient) estimates how much satisfaction increases if the feature is well-executed. The “Worse” coefficient (or dissatisfaction coefficient) estimates how much satisfaction drops if the feature is missing or poorly done. Plotting features on a grid with these two values makes prioritization visual and intuitive. Features that score high on both coefficients are one-dimensional performers. Features high on Better but low on Worse are attractive differentiators. Features low on Better but high on Worse are must-be basics you can’t afford to neglect.
Features Decay Over Time
One of the most important and often overlooked aspects of the Kano Model is that categories are not permanent. Features migrate downward over time. What starts as an exciting surprise eventually becomes an expected standard, and what was once a performance differentiator eventually becomes a basic requirement. The American Society for Quality describes this as an arrow moving from top left to bottom right on the Kano diagram: “wows become wants become musts.”
This happens because of competition and rising expectations. When one company introduces an attractive feature, competitors copy it. Customers begin to expect it. The novelty wears off, a phenomenon psychologists call the hedonic treadmill. Touchscreens on phones were once thrilling. Now they’re a must-be. Free shipping was once a delightful surprise from online retailers. Now customers are annoyed when they don’t get it.
The practical implication is that you can’t coast on a single innovation. Attractive features generate buzz and outsized satisfaction initially, but in competitive markets, they get copied quickly. The investment that once produced delight starts producing diminishing returns as the feature slides toward must-be territory. Teams that understand this build a continuous pipeline of new attractive features rather than over-investing in yesterday’s excitement.
When the Kano Model Is Most Useful
The model shines in specific situations. If your team is debating which features to build next and resources are limited (they always are), Kano analysis gives you a structured way to prioritize. It’s particularly valuable when stakeholders disagree about what customers want, because it replaces opinion with data directly from users.
It’s also useful for understanding why customer satisfaction isn’t improving despite feature investments. If your team has been pouring effort into must-be features that already work adequately, you’re unlikely to see satisfaction scores rise. The model would reveal that those features have a ceiling on satisfaction impact and redirect attention toward performance or attractive features that can actually move the numbers.
Product teams in software, consumer electronics, hospitality, and healthcare have all applied the model successfully. It works for physical products, digital products, and services. The key requirement is access to real customers willing to answer the paired questions for each feature under consideration. Without that direct customer input, you’re just guessing at categories, which defeats the purpose.