People often ask about what to put on their cards to get the best results from a card sort with OptimalSort. We have some guidelines we use for our own card sorting activities. So here they are:
Be consistent in the casing of the card labels
It is quite common to see participants assume that card labels that have all their words start with a capital letter must be a higher level heading for something. Cards should not suggest any pre-existing structure or classification.
For example, we changed “Company Structure and Organisational Values” to “Company structure and organisational values”
Avoid obvious patterns in the card labels
If participants see several cards that start with the same word then they consistently group the cards together without paying much attention to the meaning of the cards. It’s important to try and remove these type of clues.
For example, we used the terms “Wellington – where to buy a phone” and “Where to buy a phone in Nelson/Marlborough” instead of starting both cards with “Where to buy a phone in”
Use representative cards in a category, rather than an exhaustive list
It is not necessary to have every page in a category represented as a card. What we are looking for are cards that are representative of a given class of information.
For example, including all the items under an “HR” category would add extra load for the participants without gaining much additional information.
Exclude well known cards
There is usually little value in including cards that are linked on most pages of a website.
For example, “Site Map”, "Glossary"
Be consistent with the granularity
The purpose of a card sorting exercise is to make the participants group the cards. Any headings placed in the list will make people want to place other cards under them, instead of other groupings which may be more natural. This may just confuse and frustrate people.
Include enough cards of a given type for groups to be formed
It is important to ensure that the content has enough similarity to allow groups to be formed. If the content chosen is too varied or there are too few cards, participants will not be able to create natural groupings.
For example, two or more job cards, two media releases cards, two case study cards, etc.
The terms should be meaningful to the participants
Labels need to be short enough so that participants can quickly read the card, yet detailed enough that participants can understand what the content is.
For example, “Wild ideas” was changed to “Wild ideas competition” to ensure that participants understood that Wild Ideas was an actual competition.
See also Card Sorting: Pushing Users Beyond Terminology Matches
Do you have more tips for creating great cards? Please add them in the comments!