Fixing Fisheries

Testing and developing information architecture for the Canadian government.

Role

Information Architect, UX designer

Information Architect, UX designer

Year

2025

Timeline

4 months

Tools

Figma • Adobe • Proven4Users

Figma • Adobe • Proven4Users

Timeline

4 months

Tools

Figma • Adobe • Proven4Users

01 • THE BOTTOM LINE

Finding information shouldn’t feel like walking in a maze.

For this project, my team was tasked to redesign the information architecture of the Fisheries section of the Government of Canada's Fisheries and Oceans website. We used multiple research methods to test and improve the IA, such as card sorting and tree testing. Our final prototype consisted of over 5 pages of wireframes, showing an efficient way to organize the site's content.

SKILLS DEVELOPED

01

UX research

Conducted open card sorts and tree testing with over 20 users.

02

Wireframe Mockup

Helped create wireframes for over 10 pages.


03

Design

Worked in Figma + Adobe to build websites, posters, and welcome packages

02 • THE ORIGINAL SITE

Getting to know Fisheries

After exploring the site, we concluded that our target users are people with exploratory needs searching for information about licenses, policies, fishery management, and environmental influences.

Some users include:

Users looking to obtain commercial/recreational fishing licenses

Users looking for policies on fishing recreationally or commercially fishing

Users looking for policies on fishing recreationally or commercially fishing

We mapped the site and found duplicate content across multiple locations, pages that redirected users to the homepage after several clicks, and numerous broken or outdated links. These inconsistencies created unnecessary frustration and made it difficult for users to find information efficiently.

Levels:

1

2

3

4

5

Broken link

Problems

03 • OPEN CARD SORTING

Testing labels through an open card sort.

Users

7

Age

20-30

Across the study, participants created more than 30 unique category labels. We found that participants were more likely to group labels that contained similar or identical terminology. Otherwise, our card sort showed little to no correlation, except for a few labels that were paired.

SIMILARITY MATRIX

DENDROGRAM

Commercial Fisheries

+

Aboriginal Fisheries

Statistics

+

Marine charts, maps, and data

Paired together by

100% of users

Seal and sealing

+

Tuna Management

Paired together by

86% of users

Commercial Fisheries

+

Aboriginal Fisheries

Statistics

+

Marine charts, maps, and data

Paired together by

100% of users

Seal and sealing

+

Tuna Management

Paired together by

86% of users

04 • TREE TESTING

My team and I conducted three rounds of tree testing, with me leading the final test.

My team and I conducted three rounds of tree testing, with me leading the final test.

Users

15

Age

20-30

After developing and testing our initial labels, we wanted to understand whether the overall site structure made sense to users. We focused on testing areas where information overlapped, as these sections were most likely to confuse the user.

KEY TAKEAWAYS:

Overall, we decreased the average task time by 6 seconds and provided users with a better direct path to take.
Overall, we decreased the average task time by 6 seconds and provided users with a better direct path to take.

01

Participants naturally grouped content with similar terminology.

02

Several content labels were too broad or ambiguous.

03

Government and regulatory information was consistently categorized together.

TEST 1

83.3% correct

1 Direct, 3 Indirect

16.6% incorrect

0 Direct, 2 Indirect

33.6s

Avg. Time

Task: You are a tuna fisherman and would like to know about the stocks and management of tuna in Canada. Where would you find this?
Task: You are a tuna fisherman and would like to know about the stocks and management of tuna in Canada. Where would you find this?
83.3% correct

1 Direct, 3 Indirect

16.6% incorrect

0 Direct, 2 Indirect

33.6s

Avg. Time

TEST 3

83.3% correct

5 Direct, 0 Indirect

16.6% incorrect

1 Direct, 0 Indirect

27.7s

Avg. Time

83.3% correct

5 Direct, 0 Indirect

16.6% incorrect

1 Direct, 0 Indirect

27.7s

Avg. Time

04 • RECOMMENDATIONS

We recommend three things to improve the IA of the site.

  1. Have pages exist in multiple parts of the website.

Our tree testing revealed that users often took different pathways to complete the same task, suggesting that content is relevant to multiple areas of the site. Placing key pages in multiple appropriate locations lets users reach information more directly without backtracking.

SITEMAP REDESIGN

NOTE: Overall, we found that pages were inconsistently labelled, so going back to rename or relabel some pages would benefit users.

NOTE: Overall, we found that pages were inconsistently labelled, so going back to rename or relabel some pages would benefit users.

NOTE: Overall, we found that pages were inconsistently labelled, so going back to rename or relabel some pages would benefit users.

  1. Create an understandable and consistent organization of information

Since much of the website's content is location-, species-, or law-based, an alphabetical organization system would best support local navigation.
Since much of the website's content is location-, species-, or law-based, an alphabetical organization system would best support local navigation.

IDENTIFY A SPECIES WIREFRAME

MENU WIREFRAME

SEARCH BAR WIREFRAME

IDENTIFY A SPECIES WIREFRAME

MENU WIREFRAME

SEARCH BAR WIREFRAME

© 2026 Evie Chan