Kano model: how to prioritize features by what customers actually value

Most product teams have more feature ideas than they can ever build. The hard part is not generating ideas, it is deciding which ones actually move customer satisfaction and which ones quietly waste a quarter of engineering time. The Kano model is one of the most reliable frameworks for making that call, because it does something most prioritization methods skip: it distinguishes between features that delight customers and features that are simply expected.

This guide explains what the Kano model is, the five feature categories, how to run a Kano survey, how to read the results, and the mistakes that quietly ruin Kano studies. It is written for product managers and researchers who want a prioritization method grounded in real customer data rather than internal opinion.

What is the Kano model

The Kano model is a feature prioritization framework developed by Professor Noriaki Kano in 1984. Its core insight is that customer satisfaction and feature functionality do not have a simple linear relationship. Some features delight customers when present but are not missed when absent. Others are barely noticed when they work but cause real frustration when they are missing.

Because of this, treating every feature as equally important leads to bad roadmaps. Teams over-invest in basics that customers already expect, or they pour effort into capabilities customers do not care about. The Kano model fixes this by sorting features into categories based on how each one affects satisfaction, so prioritization reflects how customers actually respond rather than how excited the team feels.

Unlike pure opinion-based scoring, the Kano model is grounded in survey research. You collect structured responses from real target users and let their answers reveal where each feature belongs. That is what separates it from frameworks like RICE or simple stack-ranking, which often encode the team’s assumptions rather than the customer’s reality.

The five Kano categories

Every feature in a Kano study lands in one of five categories. Understanding what each one means is the foundation for reading results correctly.

Must-be features

Must-be features are the basic expectations. Customers assume they will be present and do not give credit for having them, but their absence causes real dissatisfaction. A banking app that does not let you check your balance, or a checkout flow that does not save your cart, fails on a must-be feature. These rarely belong at the top of a roadmap for their delight value, but they are non-negotiable: shipping without them sinks the product regardless of how strong the rest is.

Performance features

Performance features have a linear relationship with satisfaction. The more or better you deliver them, the more satisfied customers become. Speed, storage capacity, battery life, and price competitiveness are common examples. These features are where competitive differentiation often lives, because customers consciously compare options on them and reward improvement.

Attractive features

Attractive features, often called delighters, increase satisfaction when present but do not cause dissatisfaction when absent because customers do not expect them. These are the features that generate word of mouth and a sense that the product is ahead of its category. Over time, today’s delighters tend to become tomorrow’s expectations, which is why continuous discovery matters: the category boundaries shift as the market matures.

Indifferent features

Indifferent features have little or no effect on satisfaction either way. Customers do not care whether they exist. Identifying indifferent features is one of the most valuable outcomes of a Kano study, because these are the ideas that look reasonable in a planning meeting but deliver almost no return. Cutting them frees capacity for features that matter.

Reverse features

Reverse features actively reduce satisfaction for some customers when present. A common example is added complexity or forced personalization that a segment of users finds intrusive. Reverse results are a signal to either make the feature optional or rethink it entirely. They also often reveal that your audience contains distinct segments with opposing preferences, which is worth investigating with concept testing before committing.

A sixth outcome, questionable, is not a category but a flag. It indicates a respondent gave contradictory answers, and those responses should be reviewed or excluded.

How to run a Kano survey

The Kano model relies on a specific two-question structure for each feature. Getting the mechanics right is what makes the results trustworthy.

The functional and dysfunctional question pair

For each feature, you ask two questions. The functional question asks how the respondent feels if the feature is present. The dysfunctional question asks how they feel if the feature is absent. Both use the same five-point scale:

• I like it
• I expect it
• I am neutral
• I can tolerate it
• I dislike it

The pairing is what makes the method work. A respondent who says they like a feature when present and would dislike its absence is describing a performance feature. Someone who likes it when present but is neutral about its absence is describing a delighter. A respondent neutral when present but who dislikes its absence is describing a must-be. The two-dimensional answer reveals the feature’s true role in a way a single satisfaction rating never could.

Choosing features to test

Keep the feature list focused. Kano surveys ask two questions per feature, so a list of 20 features means 40 questions, which causes fatigue and lowers data quality. Most well-run studies test somewhere between 5 and 15 features. If you have a larger backlog, do a rough first pass with internal scoring or product feedback survey tools to narrow the list before committing to a Kano study.

Recruiting the right participants

This is where most Kano studies succeed or fail. The model measures how your actual target customers weigh features, so the sample has to genuinely represent them. Responses from people outside your target profile, or from fraudulent survey takers chasing incentives, will systematically distort the categories and push you toward the wrong roadmap.

For B2B products in particular, the qualifying bar is high: you often need respondents in a specific role, industry, company size, or seniority. General consumer panels rarely supply that precision. This is the part of pricing and feature research where a verified B2B participant panel matters most. CleverX addresses this directly with an 8M+ verified B2B and B2C panel across 150+ countries, where participants are identity-verified and screened on professional attributes, so a Kano study reaches the actual buyers and users whose preferences should drive the roadmap.

Aim for at least 100 qualified respondents for a stable read at the overall level, and more if you want to analyze results by segment.

How to analyze Kano results

Once responses are collected, each respondent’s answer pair for each feature maps to a category using a standard Kano evaluation table. You then aggregate across respondents to see how the market as a whole categorizes each feature.

There are two common approaches to aggregation. The simpler method assigns each feature to its most frequent category, the mode, across all respondents. The more rigorous method calculates satisfaction coefficients: a satisfaction score that captures how much a feature boosts satisfaction when present, and a dissatisfaction score that captures how much its absence hurts. Plotting features on these two dimensions gives a clear visual of which features are delighters, which are must-haves, and which are indifferent.

Translate the categories into roadmap decisions like this:

• Deliver all must-be features to an acceptable standard. They are the price of entry.
• Invest in performance features where you can credibly lead competitors, since these drive conscious preference.
• Choose a small number of attractive features to differentiate and generate delight. You do not need many, but you need at least one.
• Cut or defer indifferent features without guilt.
• Make reverse features optional or remove them.

The output is a prioritization that reflects customer reality, which is far more defensible in a planning discussion than a ranked list built on internal enthusiasm.

Kano model versus other prioritization methods

The Kano model is not the only feature prioritization framework, and it is strongest when used for the right job. It excels at answering which features matter and in what way, early in planning, when you have many candidates and limited evidence about what customers value.

It is different from conjoint analysis, which measures the relative value of feature combinations and trade-offs, often including price. Conjoint is more powerful for optimizing bundles and pricing once you know which features are in play, and it pairs naturally with broader pricing research. A common sequence is to use Kano first to identify and trim the feature set, then conjoint to optimize how those features are packaged and priced.

It also differs from internal scoring frameworks like RICE or weighted scoring, which are fast but encode the team’s assumptions. The Kano model’s advantage is that it forces those assumptions to meet real customer data. The trade-off is that it requires fielding a survey to qualified participants, which takes more effort than a planning-meeting exercise but produces far more reliable input.

Common Kano study mistakes

A few recurring mistakes undermine otherwise well-designed Kano studies.

Testing too many features at once causes respondent fatigue and degrades the quality of later answers. Keep the list tight.

Sampling the wrong people is the most damaging mistake. If respondents do not match the target profile, the categories are meaningless no matter how clean the math is. Screening and verification are not optional steps.

Writing vague feature descriptions leads respondents to imagine different things, which scatters responses. Describe each feature concretely enough that everyone evaluates the same thing.

Treating categories as permanent ignores that delighters decay into expectations over time. A feature that is attractive this year may be a must-be in two years, so Kano studies should be repeated periodically rather than treated as a one-time verdict.

Finally, ignoring segments hides important differences. A feature that looks indifferent on average may be a delighter for one valuable segment and a reverse feature for another. When the overall result looks muddy, analyze by segment before concluding the feature does not matter.

Conclusion

The Kano model gives product teams something rare: a structured, customer-grounded way to decide what to build first. By sorting features into must-be, performance, attractive, indifferent, and reverse categories, it replaces roadmap debates driven by internal opinion with decisions driven by how real customers respond.

The method is only as good as the people you survey. A clean Kano analysis built on the wrong respondents produces confident, wrong answers. That is why the recruiting step deserves as much attention as the survey design. With a well-scoped feature list and a representative sample of qualified, verified participants, the Kano model turns a crowded backlog into a clear, defensible plan for what customers will actually value.