Improving the Portlet Structure in SN7

ildikaildika RSS Feed

As the part of rebuilding the information architecture in SN7 we were highly motivated to redress the Portlet-structure to make its usage more handy

Since the purpose of this study was to improve navigation amongst Portlets, we decided to use the Card Sorting method, that facilitates the development process by involving users. This way we can map users’ factual needs, as the essential base of building IA successfully by understanding users, their backgrounds, their goals, in an early stadium of product development.

Good IA to good navigation

Navigation in between Portlets built upon the structure of Portlet systematization. Structural elements of navigation, are Portlets as nodes. Users will find them in a portlet-picker, displayed by tree-view. That offers the visible structure to navigate through to find a Portlet. Thus, the logical structure must be designed first, to enable easy navigation among Portlet-Categories. So, as the main points of determining the good architecture we took into consideration the following:

  • asking users to test, who updating websites the most often
  • providing a system, that supports earlier versions of Sense/Net as well
  • determining easy-to-maintain taxonomy

Card Sorting method

Because of the number and location of participants, we used a remote, computer-based method. This session requires participants to work independently. Though, we don’t have personal contact, we can get feedback about why they sort cards the way they do, and we can monitor the process of completing with the software, we developed. 80 users are invited to complete the test, so we expect cca 10% margin of error while evaluating results.

We applied Open Card Sort method, to learn about participants grouping conception, and enable them to create new Categories besides the predefined ones, in case they see the need. For learning the necessity of new Portlets, users are allowed to create new Cards in the test, as well.

Based on the nature of the research we determined the test-conditions by the following:

  • We ran Card Sorting with 83 cards (all Portlets we have in Sense/Net) and open to take up new ones, or exclude unnecessary ones, if deemed necessary
  • The order of Cards are randomized
  • Beside predefined Groups, users are allowed to take new Groups, as well
  • We show users in advance the estimate time, how long the test will take them

Cards to sort


In accordance with the aim of research, namely classifying the structure, how Portlets are stored, we had to include all built-in Sense/Net Portlets in the test, as Cards. Groups were established as Portlet-categories are formed in Sense/Net. As we also wanted to provide an opportinity to create new Card as new Portlet and new group as new Portlet-Category, we defined the test as Open Card Sort.

We expect that the result of the unmoderated test, with the high number of participants will be as accurate as possible. Due to the number of Portlets in our existing system, we had to envolve a relatively high number of cards. Regarding to this fact, we asked users to share their comments on cards, and groups, wherever they deem it appropriate. This way information can be obtained about the background of their decisions. Also they were asked, if an unnecessary, unused or uncleaerly configured item is found among Cards, do not place them into functional Groups. Hopefully, by the end of this test we will see, which Portlets are the the most useful for developers and endusers, and we also want to know which ones are not in use at all, by our users of any qualification.

Card sorting Groups

Excited? Take the test!

Stay tuned! In the next post, we’ll come back to you with what we learned and how we’ll proceed!

Have feedback on this post? Let @sensenet know on Twitter.

Need help or found a bug? Contact us.