We use cookies to help us provide you with the best experience, improve and tailor our services, and carry out our marketing activities. For more information, including how to manage your cookie settings, see our privacy notice.


Skip to content. | Skip to navigation

Community-made content which you can improve Case study from our community

How to use card sorting to structure your website

Card sorting is a great first step in designing or redesigning the structure of a website. By asking users to sort content into groups, you discover how they think and the language they use. This is often different from the perspective that you and your colleagues have. Building a structure based on users’ thinking is the best way to be successful.

There is plenty of information on card sorting on the web, but this guide looks specifically at overcoming the time and resource limitations that are so often a problem in the voluntary sector.

Things you'll need

  • Friendly website users
  • Sticky notes or index cards
  • A meeting room and refreshments

Decide what you want to achieve

Decide whether you’re looking looking at the structure for your whole website or just part of it. Ideally, you would card sort the whole site, as sections impact on each other. In practice, though, you may not have the time or authorisation to redesign the whole site.


Designing the tests – open or closed?

Open card sorting

In an 'open' card sort, you give users a pile of sticky notes or cards on which you've written the title of each page that you're looking at in the exercise, whether it's the whole website or just a section.

You ask them to group the cards according to which they think go together and then give each group a name that they think describes it best.

Closed card sorting

If you want to test a structure you've already designed (eg the current structure), do a 'closed' sort. You provide the categories, and the users sort their cards into these categories. It’s worth adding a category called 'I don’t know', where people can put cards that they get stuck on.

Closed sorts take less time, so this may be a factor in making your decision.


Designing the tests – remote or in person?

Card sorting is best done in person. It’s useful to hear users debating how to sort the cards, and good to be there to answer their questions. But getting together with your users is not always easy. If you invite them to your office you’ll need to pay for travel and refreshments, and if you go to them you need to find a place where there are plenty of them in one place.

You could offer an incentive, such as a voucher, free items from your organisation or special access to web content.

If you can’t see people in person then try a remote card sorting tool. Optimal Workshop has a free tool for up to 10 participants. Look out for charity discounts if you pay for a tool.

It’s best to brief people by phone before they do the card sort as it can seem a bit daunting. Pilot the test so you can warn them how long it might take.


Designing the tests – admin and recruiting

Now you need to find your helpers. Advertise on your website and make use of any existing newsletters or other communications. Directly ask anyone who has recently made a complaint about the website being hard to use. Also, see if colleagues are running any events that your users will be at: can you grab a slot in the agenda?

If your audience is the general public, you could advertise the card sort as a one-off volunteering opportunity. For an intranet, get line managers to encourage their staff to help.

Book a meeting room, arrange refreshments and – most importantly – create a few sets of cards and sticky notes. Also have a laptop or tablet on hand in case users ask you what a particular card (webpage) is actually about. Make sure everyone knows where to turn up and when.

Tip: If you've got your users for a day or a morning, you could plan to do a focus group or some usability testing too.


Running the tests

Start with an introduction and explanation of what you’re doing. You may want to demonstrate a small card sort. Put each participants at their ease by saying there are no wrong answers.

You can get participants to do the exercise either on their own or in small groups. However, keep in mind that group dynamics may affect the outcome if more than one person is doing the exercise at the same: some people are naturally more vocal when working in a group, so the result may not accurately represent everyone’s views.


Use this information in your website redesign

You should aim to test around 15 people, according to Jakob Nielsen and Boxes and Arrows.

When you've got as close to this figure as you can, analyse the results. Where are there similarities between different structures? Where are the differences? Hopefully you’ll see the basis for your new or updated design, and have some insight into where you might cross reference.

If you've done remote testing using an online tool, it will probably offer spreadsheets and other ways to analyse your data.

A tool from Donna Spencer at Maadmob is available in the 'Further information' section below.


What to do if you can’t find willing participants

None of these are ideal, but they may be worth a shot.

  • Set up a closed card sort at events like conferences and inductions, where people can come and sort as few or many cards as they like. People who are fed up with networking may appreciate doing some sorting – you could lures people over by using brightly coloured paper cups to represent the categories. Be on hand to explain the exercise.
  • Put a closed card sort in the reception or waiting room of your building –you might catch people who are hanging about with nothing to do.
  • If your user group is the general public, ask friends and family to do some card sorting for you. Get as diverse a mix as possible.
  • Do the sort as a survey – on paper or using a survey tool. Make a big grid with content (the 'cards') as rows and categories as columns. People can tick the place they would look for particular content.

Further information


Page last edited Jul 25, 2017 History

Help us to improve this page – give us feedback.