A card sorting tool people can actually take.
Open or closed sort, drag-and-drop for participants, a real share link. See which cards people group together and, for open sorts, the names they gave their own groups. Free, no signup -- try the pre-filled example below.
Edit the example, or start your own.
Pre-filled with content from an insurance broker portal as an open sort, so you can see the whole tool working before touching anything.
Cards
Already have a study?
Paste your study key and results (owner) key to view your results.
A sort gives you a hypothesis, not a spec.
Card sort results show you patterns of agreement, not a finished navigation. Where agreement is strong, trust it. Where it is split, that is real information too -- it usually means a card genuinely belongs in more than one place, which is worth designing for rather than forcing a single answer. This tool does not recruit participants for you -- you bring your own -- and a small sample should be read as indicative, not conclusive.
Common questions
What is card sorting?
Card sorting shows participants a set of content items -- one per "card" -- and asks them to group the ones that belong together. It is how you discover a navigation structure that matches how real people think, rather than one designed around your own team's internal logic, before you build anything.
Open sort or closed sort -- which do I need?
Open sort: participants create their own groups and name them, which is best when you do not have a structure yet and want to discover one from scratch. Closed sort: you provide the categories and participants place cards into them, which is best for checking whether an EXISTING or proposed structure makes sense before you commit to it.
How many cards should I use?
Somewhere around 30-60 cards is typical for a full information architecture exercise; this tool caps you at 60. More than that and participants tend to fatigue and sort carelessly toward the end, which quietly corrupts your results without looking like a problem.
What do the results actually show?
Which pairs of cards were grouped together most often, as a percentage of everyone who took the study -- the closer to 100%, the stronger the agreement that those two belong side by side. For an open sort you also see the group NAMES participants chose most, which is often the more useful finding: it tells you the language your actual users reach for, not the language your team uses internally.
Want help turning results into a real IA?
Tell us what you are working on and we will get in touch.
Part of a set
Once you have a structure, validate it with the Tree Testing Tool, or check a single screen with the First-Click Test Tool.
Beginner UX (AI) Design starts 21 September 2026.
Tools help you do the work. They do not teach you the judgement behind it. Beginner UX (AI) Design is taught live online by working UX professionals, built for career-changers with no design background. Get the brochure and see the full curriculum, the schedule and what the weeks actually involve.
Working in a team rather than learning on your own? We run in-house UX training built around your own product, or you can book a call to talk it through.