Streamlining Project Documentation to Enhance Information Usability

My scrum team had been creating project documentation with very little guidance. This resulted in inconsistent page structure and content organization, making it difficult and time-consuming to find information.

EdTechInformation Architecture

Pages and categories

Company Overview
NWEA is a prominent educational assessment organization, acquired by Houghton Mifflin Harcourt in 2023, that serves over 13 million students globally. It is best known for MAP Growth, a computer-adaptive test that dynamically adjusts question difficulty based on student answers. Teachers use these assessments multiple times a year to pinpoint student knowledge, measure academic growth, and personalize classroom instruction.

Company
NWEA

Duration
3 Months

My Role
Visual Design, UX Research, Content Structure, Implementation

The Challenge

Improve project documentation in a clear, easy-to-find, and works for several different user groups.
Goals
  • Define information architecture
  • Define and document guidelines and best practices
  • Define page structure

Key Results

  • Increased accuracy on tree test
  • Reduced number of clicks to find a specific document

Discovery: Identifying User Groups and Analyzing the Current State

Define User Groups In collaboration with the Product Owner, I identified four primary user groups who required access to the project documentation at various stages:
  • Developers
  • Leadership
  • Product Owners
  • Psychometricians
Analysis I conducted a thorough analysis of the existing content to catalog and assess its structure. This process resulted in a total of 163 pages of content. Working closely with the Product Owner, we categorized similar content into 15 distinct groups, streamlining the documentation for easier access and organization
Content analysis
content affinity mapping
labels from card sorting reults

Gathering Insights: Survey and Card Sorting to Understand User Needs and Information Structure

I designed a survey and card sorting activity in UserZoom to gather insights from the four user groups we identified.

Survey
The survey focused on understanding what types of information users seek, how frequently they need it, at what points in their workflow, and what information might be missing.

Open Card Sorting
Building on the categories we had identified, we created an open card sorting activity in UserZoom, accompanied by a brief survey. The objective was to observe how users grouped and labeled the categories, providing valuable insight into their mental models.

Refining Information Architecture: Card Sorting Insights and Tree Testing for User Validation

Based on the card sorting results, I collaborated with the Product Owner to draft a proposed information architecture.

Tree Test
The proposed solution was further validated with users through a tree test study, with tasks tailored to the specific needs of each user group.

spreadsheet with proposed information architecture
tree test research results

Iterating for Success: Improving Information Architecture Through Tree Testing and User Feedback

The results of the initial tree test were below expectations, with a success rate of only 30%. After analyzing the feedback, the Product Owner and I made adjustments to the proposed navigation structure.

Tree Test
Using insights from the first test, I created a revised tree test to validate the changes. This time, the success rate improved to 60%. With this positive outcome, we felt confident in finalizing the information architecture and moving forward with implementation.

Enhancing Clarity and Navigation: Adding Context and Relevant Resources to Each Page

Context Ensure that each page includes a concise purpose or goal statement in the header, providing clear context and reassuring visitors that they are in the right place.

Navigation and Resources Incorporate child page navigation and include relevant resources on each page to enhance accessibility and guide users to related content.

Confluence: Team page structure

Results

Most people were able to find specific information or at least navigate to the correct area/level. We increased accuracy by 50% on most tasks. The updates to the information architecture included the following:
  • Defined and limited to 5 main categories
  • Limit navigation levels to 3
  • Created best practices and guidelines
  • Created a template for existing and new content