The study is the URL: a card sorting tool with no backend

Card sorting is one of the first research methods most designers reach for, and one of the first places they hit a paywall. The commercial tools are good, but if you just need to run a quick open sort with eight participants, creating an account, starting a trial, and inviting your participants through someone else's email pipeline is a lot of ceremony for what is, underneath, a small data model: cards, groups, and which ended up where.

So I built one into this site. It's free, there's no signup, and you can use it right now in the Lab.

The study is the URL

The defining decision was refusing a backend. A study, its title, instructions, cards, and optional categories, is serialized straight into the link you share. Send the URL and you've sent the study. There's no database because there's nothing to store, and no privacy policy to write because nothing leaves the participant's browser until they choose to export their results as CSV or JSON and send them back to you.

This constraint cut most of the product for free. No accounts, no dashboards, no participant management. What's left is the part that matters: a person, some cards, and their instinct about what goes together.

Touch before drag

Drag-and-drop is the obvious interaction for sorting, and it's in there. But drag-and-drop is a desktop habit, and half of your participants will open the link on a phone. So every card also works by tap-to-place: tap a card to pick it up, tap a group to put it down. Same model, no precision required, and no drag-and-drop library needed.

Open and closed sorts

If you define categories up front, participants sort into them, a closed sort. Leave the categories empty and participants create and name their own groups as they go, an open sort. That's the whole feature surface. The renaming of a group mid-sort is often the most interesting data you get.

What it doesn't do

There's no aggregation across participants, you get one CSV per person and merge them in a spreadsheet. For a fifty-participant quantitative study, use a commercial tool, that's what they're for. For the quick sorts that make up most real-world research, a link and a stack of CSVs is honestly all you need.

Tree testing is next on the bench. If the card sort saves you a subscription this month, pass it on to someone else who needs it.