Card Sorting

This is a wonderful book on a much-needed topic. While card sorting is a basic tool of the trade, it’s previously received short-shrift in any practical publication. Donna’s done an amazing job explaining (in easy-to-understand terms) what every designer, architect, and researcher should know about the ins-and-outs of card sorting. (You might need to buy two copies, because I guarantee someone will borrow your first copy and never return it.)

Jared M. Spool, CEO and Founding Principal, User Interface Engineering

This book is a fresh, clear, practical explanation of the value of card-sorting, how to do it, and how to use the results. Spencer mixes step-by-step instructions and good examples with just enough theory. You’ll emerge from this book with new skills to create great user-centered information architectures–and smart responses to tricky questions from pesky stakeholders.

Tamara Adlin, Founding Partner, Fell Swoop, and co-author of The Persona Lifecycle: Keeping People in Mind Throughout Product Design

I wish we had this book when we first started doing card sorting. It’s a fantastic handbook that is full of very practical advice and examples from Donna’s extensive experience. We will be recommending it to all our customers.

Sam Ng, Creator of online card sorting tool OptimalSort

Donna has put together the definitive work on card sorting, a must have tool for all information architects. If you want to plan, run and analyse your own card sorts, this book has it all.

Andy Budd, User Experience Director, Clearleft

The ultimate guide to one of the under-appreciated user research methods in our toolbox. Whether you work on small web sites or in large corporate environments, this book is just the right size to give you everything you need to know to be a pro at card sorting.

Keith Instone, Information architecture lead, IBM.com user experience

This book is a godsend for those planning or using card-sorting methods. It is well and attractively organised, with a clear expository style and layout. Most important, it gives detailed and clear instructions for the planning, design, collection and analysis of this popular method of data-collection.

Anthony P.M. Coxon, Emeritus Professor, University of Wales, and author of Sorting Data: Collection and Analysis

• Chapter 1: All About Card Sorting
• Chapter 2: All About Organizing
• Chapter 3: Defining the Need
• Chapter 4: Choose the Method
• Chapter 5: Choose the Content
• Chapter 6: Choose the People
• Chapter 7: Make the Cards
• Chapter 8: Manage the Sort
• Chapter 9: Use Exploratory Analysis
• Chapter 10: Use Statistical Analysis
• Chapter 11: Use What You’ve Learned

These common questions about card sorting and their short answers are taken from Donna Spencer’s book Card Sorting: Designing Usable Categories. You can find longer answers to each in your copy of the book, either printed or digital version.

1. I wrote our content on cards/sticky notes and our team shuffled it around to create the IA. That’s a card sort, isn’t it?
   Not really. That’s just shuffling content ideas around the table (which is still useful, just not really a card sort). I think the essential element to something being a card sort is that it involves real users of your information.
   See Chapter 1 for more information on what a card sort involves. 
2. I need to test that my draft information architecture is okay. Should I do a closed card sort?
   A closed card sort is where you ask people to slot content into a set of categories that you give them. It is useful to learn about where they think content goes, but a closed card sort will not tell you whether they will be able to find it. If you need to make sure that people can find information in your IA, you should give them a set of tasks and ask where they would look.
   See page 149 for more information on how to test your information architecture 
3. My website is really big. How do I get the card sort to cover it all?
   This can be really tricky because you can’t just give people an enormous pile of cards. You can sort with topics instead of detailed content, focus on just part of the site at a time, or run a series of sorts to get good coverage.
   More tips for large sites are on page 70.
4. How many people should I involve so the answer is statistically significant?
   Statistical significance is really not important–you want insights and ideas rather than the one true answer. You should involve enough people so that you see enough similarities and differences to help with your design project.
   More tips on selecting people are in Chapter 6.
5. Should I let people put cards in more than one place?
   Participants often ask if they can put cards in more than one place, especially when there is not one clear home for a card. I always allow them to do so. It gives me useful information about content that may cross categories.
   See page 99 for more questions participants ask.
6. What do I do with all this data?
   Ah, that is the big question. Spend some time just looking for patterns and “interesting” things in the data. Then dig a bit deeper and look at similarities and differences. You may not get one perfect answer, but you’ll always learn interesting things for your project.
   Read about analysis in Chapters 9 and 10.
7. I don’t remember my university statistics. How do I analyze all this?
   If you don’t know how to do statistics, that’s okay. Don’t try! There are ways to analyze data without statistics–exploring it, looking for patterns, identifying similarities and differences. And you’ll learn more than if you plugged it into a statistics tool and got an answer. But make sure you don’t collect more information than you need, or this will be impossible to do.
   See Chapters 9 and 10 for information on how to analyze with and without statistics.

• Chapter 4 (PDF)

Over the past few years I have been slowly developing and refining a spreadsheet I use for analysis of card sorts. I have used it on many projects and find it invaluable for helping me manage the data and spot patterns.

I use it to analyse results from physical (i.e. not software) open card sorts. It could quite easily be used for closed card sorts as well, though I haven’t done that as I don’t do closed sorts.

I have refined it to the point where I’m happy with it, and have prepared instructions for it. The instructions and spreadsheet are completely free for you to use however you wish.

Files

• Card sort analysis spreadsheet instructions (PDF, 715KB)
• Empty card sort analysis spreadsheet template (ZIP, 765KB) (last updated 14 December 2009)
• Example spreadsheet with content (XLS, 2.0MB): based on papers from the IA Summit

If you use it, please let me know how it goes!

Notes

Some other random notes:

• I have done everything I know to make the PDF screen-reader friendly, but please let me know if it isn’t
• I use a PC and know it works on a PC. I have had mixed response from friends with Macs – for some it works, some it doesn’t. If it doesn’t, please let me know and I’ll explore further
• In the comments below is an answer to the question “What do I do with sub-categories”. I’ll include this in the instructions next time I revise them

Screenshots

Entering the card names:

Summary of all cards and standardised categories:

Analysis of categories:

Visualisation of results:

There’s something about cards. Cards can turn the drab and mundane into something strangely exciting. Recipe cards, for example, have a peculiar allure that cookbooks lack. A book full of tables of baseball statistics? Dull as dishwater. Put those same stats on the backs of trading cards, however, and now you’ve got something. You’ll see this same pattern unfold whenever card sorting is included in a user research session. Maybe the research participants have already been asked to poke at a prototype design, or they’ve simply answered a bunch of questions about their attitudes and preferences. In any case, their demeanor inevitably changes when the cards come out for sorting. “What’s this?” they seem to be saying to themselves. “This doesn’t look like what I expected.

This looks like it might actually be…fun!”

I don’t think this reaction can be attributed to their enthusiasm for the task they’re being presented with. There are a hundred ways you could ask someone to help you organize content for a website, but the 99 ways to accomplish this goal that don’t involve cards aren’t likely to provoke the same interested response.

Perhaps it’s because cards don’t seem like a serious tool to people. After all, for most of us, our first encounter with cards as children was not as something people did work with—it was as something people played with. And a stack of cards truly is an invitation to play: to hold them in our hands, to shuffle, deal, flip, match, stack, and sort.

That’s not all there is to it, of course. There is some deep satisfaction to be derived from simply finding a place for everything and putting it there, which surely explains part of the appeal for card sorting—and, indeed, the entire practice of information architecture. Like all the good card games we played as children (and many of us still play today), card sorting is about the tension between randomness and order, a tension the human mind finds infinitely engaging.

Exploring that tension is itself a kind of play. In this book, Donna Spencer lays out the rules of the game. Knowing those rules will help you get the most out of this deceptively simple technique. People often miss the subtleties involved in using card sorting effectively, but with Spencer as your guide, you can be sure of the best way to play your cards.

— Jesse James Garrett, author of The Elements of User Experience and president, Adaptive Path

Everyone says it, but it is actually true: While this book is mostly my effort, it has been hugely improved via help from many other people.

First, a general thanks to everyone who read chapters and provided great comments: Steve Baty, Dan Brown, Dustin Chambers, Ruth Ellison, Leo Frishberg, Patrick Kennedy, Jorge Larango, Sam Ng, Kristi Olsen, Gene Smith, Peter Van Dijck, Steven Weintraub, and Alex Wright.

I also want to say a special thank you to two particularly special reviewers. Leo Frishberg and Dan Willis both did something quite hard and told me how bad my early version was. And Dan rewrote my entire chapter structure into something that made sense. Thanks so much to both of you for being brave, honest, and constructive–your comments made a huge difference, and this book is much better because of it.

It was great to be able to involve my closest friends as well. Ruth Ellison took photos and Caronne Carruthers-Taylor, Nigel Carruthers-Taylor, and Andrew Boyd posed for them. Thanks for sharing your time and company to help me illustrate how card sorting works.

One of the things I really like about this book is the case studies and quotes. They show that I’m not the only person who thinks this way, and they help highlight the key points. Thank you to everyone who provided case studies and allowed me to print their words.

I also want to thank everyone who has participated in a card sort I’ve run, including everyone who has done the winery card sort as part of one of my information architecture workshops. You’ve all helped me learn how people think, and I design better products as a result.

And the last thank you is to all the nice folks who have used my analysis spreadsheet and then told me how good it is. It is rewarding to see that something I knocked together for myself was actually handy for you. One day I’ll hook up with a smart programmer and make a better tool to do this.

Card Sorting Blog View all Blog posts

• Card sorting analysis: a co-occurrence matrix,
  on June 5, 2012
• New articles and resources,
  on May 16, 2012
• Interviews,
  on December 15, 2009
• Card sorting doesn’t cut the custard,
  on August 20, 2009
• Inviting the guests to cook the food,
  on July 1, 2009