Information Architecture Research Case Study:

Improving a UN HR Portal’s Findability through IA Research

A large United Nations agency wanted to help its 2,000+ employees find HR-related content more easily. We led a card sort to generate a new information architecture. We then tested and improved the IA through a first-click test.


A large international agency within The United Nations System was redesigning its human resources intranet portal used by over 2,000 employees. A top goal of the redesign was to improve the findability of HR-related content and tasks.

Changes over time had led to an overly complex platform that required a large number of clicks to navigate to the right information. Many users abandoned their tasks and instead submitted an inquiry to the HR department.

The organization wanted to conduct user research to help create a more user-friendly information architecture and navigation that would allow staff to more easily find answers and complete tasks on the portal. They partnered with Marketade to lead this research.

We used two information architecture research methods for this project:

  • An open card sort to understand the audience’s mental models and to help generate a user-centered information architecture
  • A click test to validate/improve the new structure and to compare the performance of different design directions

With each method, we conducted both qualitative (moderated) and quantitative (unmoderated) testing.

Card Sort

In a card sort, we present users with a collection of cards that represent content or functionality; we ask users to group the cards in ways that make sense to them and then label the groups. After watching enough users go through the process, including some users thinking aloud as they go, we’re in a great position to generate intuitive IAs.

For this card sort study, we ran 6 moderated sessions and 50+ unmoderated sessions that represented a diverse pool of participants across various job titles, departments, and tenures at this UN agency.

Users were given 56 randomized cards representing HR portal content and functionality, e.g. “on-site childcare” and “work authorization.” We used the Optimal Sort online tool from Optimal Workshop to run the study.

We used the qualitative sessions to:

  • Pilot the study design prior to launching the unmoderated sessions; we adjusted the card labels and the number of cards based on what we learned.
  • Understand why users sorted content in certain ways.

Our primary method of card sort quantitative analysis was category standardization. We manually reviewed the 300+ categories created by our 50+ participants and identified grouping and labeling patterns. With the help of Optimal Sort’s analysis tools, we were able to standardize those categories and visualize the patterns.

We validated our analysis by reviewing a similarity matrix. This visualization indicates the frequency that cards were sorted together with darker shading.

3 Marketade researchers individually reviewed the data and drafted individual information architectures (IAs). We then collaborated to establish 2 new IA versions for testing.

Click Test

Click testing helps us to test an IA and gives additional data points on early-stage navigation design.

Users are presented with a series of realistic scenarios alongside a wireframe and asked to click the first place on the wireframe where they would look to complete that task.

For this study, we ran 6 moderated sessions and about 75 unmoderated sessions with a new set of participants from the UN agency.

We used both the qualitative and quantitative sessions to test 2 different versions of the IA.

Neither IA performed at the level that we considered to be satisfactory overall. Therefore, we undertook a task-by-task analysis in order to determine the most successful aspects of each IA individually. We then created a new IA incorporating the best of both versions.

We also identified the parts of the wireframe design that worked well and the parts that hurt findability. For each problem, we provided specific design recommendations.

A participant groups cards representing HR portal content during a card sort.


We delivered a 22-page report along with a 35-page deck that captured our research process, findings, and final recommendations.

We presented the deck to 10 stakeholders from the organization’s HR, IT, and design teams. We included video clips from qualitative research in the presentation.

After the project was complete, our primary stakeholder shared this feedback:

“Thanks so much for the excellent work and super compelling and thorough presentation!”

​More Case Studies


How a Card Sort Helped a Top Financial Firm Create an Intuitive IA

While challenging to do right, a card sort study is the best starting point for generating a user-centered information architecture. Here’s the 12-step process we used for a Fortune 500 content hub — and the challenges we overcame.

How the Smithsonian Improved Content Findability with Tree Testing

The findability of digital content is critical at the world’s largest museum, education, and research complex. Early in the redesign process of the Smithsonian Global website, we led IA research that pointed the way toward a more intuitive navigation.