A good website is as easy as navigation in library – each book is exactly where it supposed to be. Good e-shop is organized as a logical business – the people in it will find the products you need … plus something extra appealing. To ensure that you always find the necessary information, information architecture take care of it.
What is information architecture?Information architecture is the skeleton site, defining its structure and contains all the sites that are on the web. More generally, it would be possible to apply the information architecture of how the individual information are distributed on the web.If the information architecture is designed poorly, it can cause major problems. To see if your architecture is in order, determined by using fast user testing. Once the testers begin asking questions, “What is this?” Or “Where do you get when you click here?” You know it’s wrong.
How to fix your defective architectureWhether you are working on a new site, or you want to look at the use of existing, follow one simple rule – listen to users. The best way to verify the information architecture is cardsorting.
Cardsorting – give users what they needUsers of card sorting cluster the cards that represent the different pages (categories) on the Web. Real-testers stores them into a logical structure that is their own and natural, and thus create a site hierarchy such that they would like to have it (or is deemed to be correct).Since cardsorting is about the verification of categorization and its concept, the card may not be named exactly as pages. Verify more than the actual site structure titles of each page. It is also important to ensure that each card can be categorized. Produce therefore only those cards that can be include at least in one category.
Cardsorting can be split like questions on the questionnaires:
- cardsorting open, where people unlimited cards classified into categories, which are also named;
- cardsorting closed, where people store cards into predetermined categories;
- cardsorting hybrid, where people like to set up the card categories, but they can create their own, if necessary.
When to use what type of cardsorting?Different types of cardsorting have different goals and it depends on the type of project and the state in which they are.Open cardsorting is particularly suitable as a creative exercise for new solutions. Users can create new tabs and then categorize them. Also, you will find, where users are looking for information on your site, how to structure a what web page titles users expect. Open cardsortig use as starting analysis of how people considering the topic and categorize information.Closed cardsorting focuses on categorizing predetermined cards. In this case, users will receive a set of cards that need to categorize those topics. The advantage of this method is that figuring out how people classify cards already predefined areas that may be important to you and you cannot move them. Closed cardsorting also helps to:
- testing a certain level of pre-existing architecture, which is problematic for users
- identify problem items of Architecture
Different people, different tastes
Individual testingThe simplest and most effective is cardsorting one user to one observer. There is no need of any specific room and can be implemented even in a coffee shop or at home on the terrace. It is important that the user at cardsorting describes what they do, and always explained the “Settings are here because …”.
Group testingAlternative to individual cardsorting is group testing in which user group categorises the card (or multiple groups at once). This method provides quick results and can be seen extensive connections in the mindset of people. The disadvantage may be the group itself and too dominant individuals. It is therefore necessary to sufficiently pay moderation testing.
How many users to test?Unlike a traditional user testing, in which five people is enough, Jacob Nielsen recommended in the case of cardsorting at least 15 testers. We recommend that when choosing to take into account the user to select and testing such people who represent real users. Cardsortig with users who are not your target audience may adversely affect the final categorization. For example, if you work on a project that is for older users (40 and older), you give test the architecture to the young testers (20), who have a completely different view on the subject.
Put the cards on the tableFor cardsorting with sufficient predictive value it is appropriate to use at least 30 cards, and, of course, depends on the extent of information architecture. Optimal Workshop website that provides an online platform for cardsorting recommended 30 to 60 cards for the best coverage of issues and gain the information.
Ideas for the content of the cards, you can look in several places:
- search on website
- brainstorming with colleagues and friends
- analysis of the current site
- analysis of competing websites
What you think is right for you, might not be right for the usersStop using the phrase “I think …”! Web architecture is the basic building block. If you already have the basics of web build on their impressions and unproven theories, you can sink the whole project. When creating any part of site, it is needed to view solutions from a user’s perspective.
Data sources that will also helpOn many operating sites reveal data from Google Analytics, which can show what pages are or are not popular. Another source of information may be the words that visitors search for on your site. If a search term is used too often, it is quite likely that your users cannot not find it, and thus points to the wrong Web architecture. The best can be deployed Hotjar that will show you just how people behave. So you can get different information such as data from Google Analytics.
Recipe for cardsortingCardsorting is very easy way to learn from the users all you want to know. It can be handle relatively quickly and also free or as part of a larger user testing.What we need to prepare for 1 successful cardsorting:
- One type properly selected on the basis of what information we want to get
- 15 testers from the target group
- 30 cards containing that represent information on the web
- 1 large enough table
- 90 minutes of your time to one tester or one group of testers
- Explain the tester, that he is going to test the site, not his/her.
- Explain the tester what will be the result of testing.
- Explain tester, what will his role during the test.
- Make sure the tester understands the contents of each card.
- Get into the test. Make sure that the tester commented, what it does and why it does it.
- Once the task is done ask the testers point of view – what was difficult, what was easy etc.
- Record the results.
- Evaluate the results and see a correlation between cardsorting.