Skip to main content
← All guides
Tool + guide · September 2026 · 9 min read

The Kano model

Two questions per feature — what if it were there, what if it were not — and the pair tells you something a single importance score never can. Paste your answers below.

Short answer

What is the Kano model?

A way of sorting features by how they affect satisfaction. Each feature is put to customers twice — how would you feel if it were present, and if it were absent — and the pair of answers classifies it as must-be, one-dimensional, attractive, indifferent, reverse or questionable.

The short version

  1. Published by Noriaki Kano and colleagues at the Tokyo University of Science in 1984.
  2. Must-be features buy nothing when present and lose the deal when absent. Attractive features do the opposite.
  3. A single “how important is this?” scale cannot separate the two — which is the whole reason the method asks twice.
  4. Categories migrate: today’s delighter is next year’s expectation, so a Kano chart has a shelf life.

Free · No signup · Nothing leaves your browser

Paste the answers, get the categories

One row per person per feature: the feature name, the functional answer, the dysfunctional answer. Separate with tabs, semicolons or commas — not spaces, because feature names contain them.

1I like it
2I expect it
3I am neutral
4I can live with it
5I dislike it

Words work too — “I like it that way”, “I can live with it” — so a survey export can go straight in.

Example: four features, eight people each. One of them lands on a genuine tie.

32answers read

4features

AttractiveOne-dimensionalIndifferentMust-be0%50%100%0%50%100%Dissatisfaction if absent →Satisfaction if present →Offline syncFaster exportDark modeCustom themes

Reverse and questionable features are left off the plot — neither axis describes them.

Offline sync

Must-be

Nobody thanks you for it. Its absence loses the deal. Build it first, then stop.

n = 8A 0 · O 1 · M 7 · I 0satisfaction if present 13%dissatisfaction if absent 100%

4 people would have to answer differently to move this to One-dimensional.

Faster export

One-dimensional

More is better, linearly. Worth competing on, and worth measuring.

n = 8A 1 · O 6 · M 1 · I 0satisfaction if present 88%dissatisfaction if absent 88%

3 people would have to answer differently to move this to Must-be.

Dark mode

Attractive

Delights when present, forgiven when absent. The differentiator — but never before the must-haves.

n = 8A 7 · O 0 · M 0 · I 1satisfaction if present 88%dissatisfaction if absent 0%

4 people would have to answer differently to move this to Indifferent.

Custom themes

Attractive

Delights when present, forgiven when absent. The differentiator — but never before the must-haves.

n = 8A 4 · O 0 · M 0 · I 4satisfaction if present 50%dissatisfaction if absent 0%

⚠️ One person answering differently would make this Indifferent.

Paste into Notion, a doc, or your roadmap.

What this cannot know. Everyone answered about a feature described to them, not one they used, and nobody was asked to give anything up for it. Kano ranks how features feel; it does not price them, cost them, or tell you they are possible. And the categories move: today's delighter is next year's must-be, which is why a Kano chart older than a release cycle is a historical document.

Feature order settled, price still open? Van Westendorp gives the believable range and Gabor-Granger the demand curve.

Test the idea behind it

Kano sorts the roadmap of a product people already want. A 15-minute session ends with a written GO / WAIT / NO-GO and the reasoning behind it. Free tier, no card.

The two questions

Ask both about every feature, in this order, using the same five answers each time. The wording matters: “how would you feel” invites a reaction, while “how important is it” invites a rationalisation.

Functional

How would you feel if this feature were present?

Dysfunctional

How would you feel if this feature were not present?

  1. 1I like it that way
  2. 2I expect it to be that way
  3. 3I am neutral
  4. 4I can live with it that way
  5. 5I dislike it that way

The evaluation table

Every pair of answers lands in one cell. Read down for the functional answer, across for the dysfunctional one. This is the published matrix — it is not symmetric, and the two diagonal corners are contradictions rather than strong opinions.

If present ↓ / if absent →like itexpect it to beam neutralcan live with itdislike it
like itQAAAO
expect it to beRIIIM
am neutralRIIIM
can live with itRIIIM
dislike itRRRRQ

A attractive · O one-dimensional · M must-be · I indifferent · R reverse · Q questionable

The six categories

M

Must-be

Nobody thanks you for it. Its absence loses the deal.

Brakes on a car, a password reset that works. Presence buys nothing; absence is fatal. Build them, meet the bar, then stop — there is no reward for the best password reset in the category.

O

One-dimensional

More is better, in a straight line.

Speed, storage, battery life. Satisfaction rises with the amount you deliver and falls as it drops. These are what you compete on and what belongs on a comparison page.

A

Attractive

Delights when present, forgiven when absent.

Nobody asked for it, and its absence costs nothing — but its presence is what people tell their colleagues about. The differentiator, and the thing founders build far too early.

I

Indifferent

Moves nobody in either direction.

The most expensive category on the list, because it looks like work and reads like progress. Most roadmaps contain more of these than their authors believe.

R

Reverse

A real share of people want the opposite.

Usually a sign you surveyed two kinds of user at once — a feature the power user demands and the newcomer finds threatening. Segment before you decide.

Q

Questionable

The answers contradict each other.

Someone said they like the feature both present and absent. Almost always the question was misread. A few are normal; a sizeable share means rewrite the wording.

What Kano cannot tell you

It has no idea what anything costs. A must-be that takes two years and an attractive feature that takes a week come out of the survey looking like a priority and a luxury. Kano orders features by feeling, not by return.

Nobody gave anything up. Respondents were not asked to trade one feature against another or against a price. That is what conjoint analysis does, and it is the reason a Kano survey can return six features people all want.

People answered about a description. They have not used the feature. The gap between imagining a feature and living with it is exactly where indifferent results hide.

The categories move, and only in one direction. Delighters become expectations as competitors ship them. An analysis without a date on it is a liability.

What it does better than anything else of its weight is stop you building delighters on top of a missing must-be — the failure that produces a product people admire in a demo and do not buy.

Four ways Kano goes wrong

Building the delighters first

Attractive features are the fun ones, and they are visible in a demo. But a delighter on top of a missing must-be is a product people admire and do not buy.

✓ Instead: Must-be first, to the point of adequacy and no further. Then one-dimensional, where effort converts to satisfaction. Attractive last, and deliberately.

Asking “how important is this feature?”

A single importance scale collapses the whole point. Must-be features score low on importance — people take them for granted — right up until they are missing.

✓ Instead: Ask both questions. The information is in the pair, not in either answer alone.

Treating the category as permanent

Categories migrate in one direction: today's delighter becomes tomorrow's expectation. A camera on a phone was attractive in 2002 and must-be by 2008.

✓ Instead: Date the analysis and re-run it each cycle. A Kano chart older than a release is a historical document.

Averaging everyone together

Two segments with opposite needs produce a middle that describes neither. Reverse answers are the visible symptom; indifferent results are often the invisible one.

✓ Instead: Run the analysis per segment when you have enough answers, and read the reverse counts as a warning that you have more than one audience.

Before the features, the problem

A Kano survey assumes the product should exist. It sorts the roadmap of something people already want — and it will happily sort the roadmap of something nobody wants, producing a tidy chart either way.

Is the problem real, is there a market, will they pay are the three critical criteria in our Go/No-Go rubric. A 15-minute session works through all seven and ends with a written verdict and the reasoning behind it.

Test the idea behind the roadmap →

15 min · free tier, no card

Frequently asked questions

What is the Kano model?+
A method for sorting product features by how they affect satisfaction, published by Noriaki Kano and colleagues at the Tokyo University of Science in 1984. Each feature is put to customers as two questions — how would you feel if it were present, and how would you feel if it were absent — and the pair of answers places the feature into one of six categories: must-be, one-dimensional, attractive, indifferent, reverse, or questionable.
What are the two Kano questions?+
The functional question: "How would you feel if this feature were present?" The dysfunctional question: "How would you feel if this feature were not present?" Each is answered on the same five-point scale — I like it, I expect it, I am neutral, I can live with it, I dislike it. The combination of the two answers is what classifies the feature; neither answer means much alone.
What do "better" and "worse" coefficients mean?+
They were introduced by Berger and colleagues in 1993. Better — satisfaction if present — is (attractive + one-dimensional) divided by the total of attractive, one-dimensional, must-be and indifferent. Worse — dissatisfaction if absent — is the negative of (one-dimensional + must-be) over the same total. Reverse and questionable answers are excluded from the denominator because they describe neither direction.
How many people do I need for a Kano survey?+
Enough that the winning category is not an accident. A category chosen by a margin of one respondent is not a finding, and the calculator on this page says so: for every feature it reports how many people would have to answer differently to change the verdict. If that number is one, the result is undecided regardless of how many people you asked in total.
What does a "questionable" result mean?+
That someone said they like the feature both when it is present and when it is absent — or dislike it both ways. The answers contradict each other, and the usual cause is a misread question rather than a strange respondent. A few questionable answers are normal; more than roughly one in six means the wording needs rewriting before the data is worth analysing.
Is this calculator free, and does my data leave the browser?+
It is free and requires no account. Everything is computed in your browser — the answers you paste are never sent to us or stored anywhere.

Related guides