The Kano model is a framework for classifying product features based on their relationship to user satisfaction, developed by Noriaki Kano in the 1980s. Instead of treating all features as equally desirable, it forces you to distinguish between things users expect as baseline, things that scale satisfaction linearly, and things that genuinely surprise and delight.
The model sorts every feature into one of five buckets using a paired-question survey. For each feature, you ask how users feel if it's present, then how they feel if it's absent. The combination of answers determines the classification:
Must-be (basic): Users don't thank you for a working login page, but they'll leave if it's broken. These are table stakes. They only generate dissatisfaction when missing.
Performance (one-dimensional): More is better, linearly. Faster load times, more storage, better search accuracy. Users can articulate these needs clearly.
Attractive (delighters): The features users didn't know they wanted. They generate disproportionate satisfaction but zero dissatisfaction when absent — because nobody misses what they've never seen.
Indifferent: Users genuinely don't care. These show up more often than you'd expect, and they're a signal to stop investing.
Reverse: Features that actively annoy a segment of users. Autoplay videos are a classic example.
Here's where this gets practical for information architecture work. An e-commerce team was planning a card sort to restructure their support section. They had 47 potential topics to include — too many for a clean study. So they ran a Kano survey on those 47 topics first.
The results shifted their entire approach. "Track my order" and "Return an item" scored as must-be features — users expected to find these instantly. "Sustainability certifications" and "Company history" scored as indifferent. "Live chat availability indicator" scored as an attractive delighter.
They cut the card set down to 28 items by dropping the indifferent ones entirely. The must-be items went into the card sort with extra scrutiny — these needed to land in categories with high agreement rates because users would bounce if they couldn't find them. The delighters got included too, but the team treated their placement as exploratory rather than high-stakes.
The result: a tighter card sort that focused participant attention on the content that actually mattered. The full study ran in 8 minutes instead of 15, which cut their abandonment rate roughly in half.
Kano surveys take real effort to design well. The paired-question format (functional + dysfunctional) feels repetitive to respondents, and you'll see quality drop after about 20 feature pairs. If you have 50+ features to classify, run it in batches.
The model also assumes features are independent, which they rarely are in practice. A feature might delight users on its own but feel redundant alongside another feature. Kano won't catch that interaction.
And the category boundaries aren't always clean. You'll get features where 35% of respondents say "must-be" and 30% say "attractive." The model has a classification rule for this (majority wins), but close calls deserve qualitative follow-up rather than blind trust in the numbers.
What are the five categories in the Kano model? The Kano model classifies features into five categories. Must-be (basic) features cause dissatisfaction when missing but don't increase satisfaction when present. Performance (one-dimensional) features have a linear relationship — the better they work, the happier users are. Attractive (delight) features surprise users positively but cause no dissatisfaction when absent. Indifferent features have no impact on satisfaction either way. Reverse features actively cause dissatisfaction when present.
How does the Kano model connect to card sorting? Kano analysis helps you decide which features or content topics to include in your card sort. By running a Kano survey first, you can focus your card sort on the must-be and performance features that users actually need to find, rather than wasting card slots on indifferent or reverse features that don't belong in your navigation at all.
How many participants do you need for a Kano survey? A Kano survey typically requires 20-30 participants per user segment to produce stable category classifications. Unlike usability testing, Kano analysis is quantitative and needs enough responses to show clear patterns. If you're surveying multiple distinct user groups, plan for 20-30 per group.
Explore related concepts, comparisons, and guides