Information architecture is the structure and description of data, instructions, and announcements that can be found within a portal, application, or any other digital system. We will discuss and show – based on our experience from designing more than 40 intranet platforms – what methods we use when designing a company intranet to help employees easily and intuitively find all the information and documents they need.
Intuitive information architecture, what does it mean?
Before I talk about methods, it is worth reminding why taking care of the right information architecture is so important. In simplest terms:
- Correct information architecture – an average user will not have problems with moving around the company’s resources, he or she will easily and quickly find what is looking for, the employee will understand the presented information instantly, and the layout will be perceived as logical. Thanks to this company’s staff will reach the information on their own, saving time for themselves and others. Otherwise, they’ll constantly ask the HR department or other colleagues for assistance.
- Faulty, misguided information architecture – for an average user it means disorientation, searching for information in various places before he or she clicks the right path, and it is not always possible! Finally – lost time and frustration.
The keyword here is “average” – it is easy to construct an inappropriate list of subpages or categories, guided only by our own intuition and what we – intranet creators and owners – think best describes the content. But as creators and curators of this platform, we are not representing the typical user and have much different knowledge and mental model. Therefore is it almost impossible to design good information architecture without proper user research.
Fortunately, the problem of ambiguous information architecture is so common that an appropriate toolset and methods have already been created to solve it.
In EDISONDA, based on effective and proven design and research methods, we have developed a process of creating and modifying the existing information structure consisting of several steps. It is worth carrying out such actions not only when implementing a new intranet but also when reviewing and optimising the currently used system where we may have great room for optimization. The results and obtained effects, in the form of faster, easier access to information and increased employee satisfaction, may surprise us.
Step 1: Analyse your existing intranet
You should start with what you already have. Inventorying your intranet involves making an inventory of all the types of content that is already published in it – this can be in the form of sub-page types. You can recreate the structure of your current intranet in a simple spreadsheet. The Inventory tool for managing content strategy has similar functionality.
Additionally, if the existing intranet has a built-in analytical tool, showing the most frequently clicked items or the most frequently undertaken tasks – it is worth using it. Very often we are surprised by how little intranet owners know about the traffic in their system.
The second option is to list all types of content that should be included in the new intranet, which you have managed to collect during research or interviews with users representing different stakeholders.
From our experience
Employees tend to pay attention primarily to functionality that is missing but should be on the intranet. This is important information for a manager, but equally, important information is… what is not missing. When examining the usability of current solutions, our researchers always pay attention to the elements that work well.
It is worth doing so that it is not overlooked in the new system. We all quickly get used to a good one and consider it a certain standard, a norm – that’s why it’s so easy to forget about it in designing a new intranet.
Step 2: Card Sorting
Card sorting is a UX research method to develop the optimal structure of any system. It is a quantitative method – therefore it requires the involvement of a larger number of users. Programs used to conduct it use complex statistical algorithms to cluster the information, nevertheless the exercise I describe below is very simple for participants. It does, however, require some preparations, which I describe below, but it is understandable and its execution does not pose any problems.
Cards represent sub-pages or content types. If we have data on the most common tasks on the current intranet – for example, from the analytics built-in there – we can select cards based on the three most common activities.
How does the exercise work?
Each participant of the exercise, who is also a user of the system, is given a set of cards. Then they are asked to group them according to a pattern that will make sense to them.
In the next step, the user has to name the groups of cards created in this way – in such a way as to best reflect their content.
Depending on the number and complexity of the cards, the exercise can be expanded to include tests or short tasks; the final scope depends on the scale of the project.
How to prepare and conduct the card sorting?
As I mentioned earlier, this is a quantitative research method. That means it requires a larger group of respondents. There aren’t many sources that give the optimal number of participants, but the absolute minimum would be around 30 people – that’s how many you need for the survey to be statistically valid and yield reliable results.
From our experience
Remember that it is always a good idea to invite more users to this exercise because not everyone will be able to complete it. Besides, while working with our clients we observe that the number of cards does matter. The more cards there are, the more difficult it is for participants to get acquainted with all of them, arrange them properly, etc. For this reason, it is best to use about 30 cards in a single study.
Using more, especially more than 50 cards, can overwhelm the participants and discourage them from accomplishing the study. Remember that they are employees, they still have a lot of work-related responsibilities on their minds and a busy schedule. In such a situation it is easy to fail, because too many participants may simply abandon the survey along the way and return to work. A poorly prepared survey means a lack of reliable results and wasted employees’ time at the same time.
Step 3: Tree testing
Once we have the cards sorted by the respondents, we can verify the results by information tree testing. That is why this method is sometimes called reverse card sorting.
How is the information tree tested?
The test uses a delineated intranet information architecture. It can be created on the basis of patterns repeated most often during card sorting.
Next, we ask the participants to imagine that they are to perform certain tasks. We can choose them based on the results of a top-task analysis to make sure that we don’t overlook important elements of employees’ daily routines.
In the intranet structure presented to them, participants are asked to identify elements that would allow them to perform specific tasks or find particular information. This provides additional confirmation as to whether, in fact, the content category names that seemed logical and sensible to participants in the previous study based on the card groups, will work when the starting point is reversed.
Why additional testing of the information tree?
In card sorting, the starting point was the content – and the user had to name it. In reverse card sorting, i.e., information tree testing, the starting point is the name – and the user determines what content lies beneath it. A combination of those two methods allows us to design and check if the structure and the naming conventions are all comprehensible.
This way, we find out for sure if our information architecture is correct. If, in most cases, the user will perform the task quickly and efficiently. Optimisation of processes in the company happens every day, and when our challenge is to manage a change – it is worth taking a moment to double-check new projects.
Proper communication is absolutely essential: using employees’ time in which they could perform their duties may meet with resistance, especially from their superiors. But it is worth remembering that the time spent on researching and testing systems will pay off in the future – not only through greater efficiency but also through employee satisfaction and comfort.