Research summary: Employment insurance maternity and parental benefits

August 23, 2019

This optimization project was undertaken by the Digital Transformation Office (DTO) in close collaboration with Service Canada’s Employment Insurance (EI) delivery and policy teams, and the Employment and Social Development Canada (ESDC) web team. Our goal was to make employment insurance maternity and parental benefits easier to find and understand for Canadians planning to expand their family.

During our discovery research, we looked at a lot of external websites. We wanted to know what Canadian parents were sharing about maternity leave, parental leave, and navigating the EI system. This helped us understand which aspects of accessing these benefits people find most complex.

We found that people struggled with:

Search analytics told us that searches using the keyword “leave” far exceed searches for “benefits.” This suggested that these concepts, although legally distinct, are not clearly differentiated. Leave is a right defined in the Canada Labour Code and in provincial employment standards acts. While on leave from work, EI benefits provide a sum of money to Canadians who have paid EI premiums. This sum represents a proportion of previous earnings.

When we spoke with EI service agents, we realized that many parents don't understand that they can't switch between benefit options. For example, parents may choose the extended option, planning to have more time at home. This provides 33% of earnings. Some parents later change their minds and want to switch to the standard option, which provides 55% of earnings over a shorter period. They don't understand that once they've chosen, the choice is final.

An extra challenge for this optimization effort was that we had to explain a new parental sharing benefit. This new “use it or lose it” benefit was set for launch in March of 2019. It is available only to partners who share EI parental benefits. It's tricky to understand the options available to each parent, and how one parent’s choices impact the other’s benefits.

From previous research, we knew that tasks that require people to do math are particularly challenging. Concepts and terms such as “hours of insurable employment” and “variable best weeks” are not self-explanatory. They make the task of figuring out a benefit amount more difficult.

Baseline testing and results

We worked with the EI team to define scenarios for testing. The 8 scenarios covered a range of circumstances (from high-risk pregnancy to adoption, etc.). We asked common questions about the benefits and the process for getting them. We tested with 22 participants between the ages of 25 and 39. All were employed full-time, and none had applied for any form of EI within the last 5 years.

Out of a total of 155 task trials across the 8 task scenarios, participants found the right page 46% of the time. Participants succeeded at the tasks only 28% of the time.

The design effort

After analyzing the video evidence from baseline testing, we developed a prototype in GitHub. We used the same service initiation template that we used in the EI caregiving benefits optimization project.

Through an intensive series of workshops with EI subject matter experts, we iterated the prototype. Key design features included:

We continued to iterate the prototype throughout testing. We observed each session closely to identify where people struggled, and we implemented fixes. This approach significantly improved the design and the testing results.

Validation testing results

Our target was to improve both findability and task success by a minimum of 20 percentage points, or to exceed 80% for both measures. In the validation round, we tested with 20 participants for a total of 167 task performance tests.

This chart shows the baseline measurement at the start of the project compared with the validation measurement on the prototype redesigned by project team

Task completion success rates – table
Task Baseline Validation
When to apply 33% 75%
Maternity/sickness 22% 80%
Insurable earnings 62% 89%
Max leave 5% 79%
Variable best weeks 32% 53%
Change parental benefit 24% 85%
Bonus included in earnings 0% 69%
When benefit payments end 44% 94%

What we did that helped people succeed

Matched expectations

We reorganized and regrouped content so that the content of each page was clearly related to the page heading.

Reduced complexity of content

To improve scannability, we added headings, and removed extraneous information. We hid technical details using the expand-collapse pattern. This kept pages from appearing overly complex. We guided people to EI-specific concepts and terms (such as “best weeks”), explained them clearly, and used them sparingly.

Saved them from doing math

We designed a simple estimator, allowing people to see the impact of their choices. The estimator requests minimal inputs to provide a detailed, useful answer that can help people make better-informed choices.

Used numbers with care

Both eligibility requirements and benefit entitlements are based on numbers of weeks. We took great care in presenting numbers to help people successfully differentiate between these.

Show the math

Before page for the number of benefit weeks.
People focused on the higher number thinking it was the total, and didn't even consider the benefits being used in combination.
After page for the number of benefit weeks.
Showing a formula people can use and relate to helps them do the math. This also explains the 5 or 8 weeks "sharing bonus" in a simple way.
Detailed description

The image first shows a section of the original page content with red highlighting around a statement about a maximum of 15 weeks of EI maternity benefits. There is red highlighting around a second statement about a maximum of 61 weeks.

Below this is a second image from the prototype version of the same content. There is green highlighting around a simple math equation. The equation shows 15 weeks maternity plus 61 weeks of extended parental equals 76 weeks total for Janelle.

Request the research

If you’d like to see the detailed research findings from this project, email us at

Let us know what you think

Tweet using the hashtag #Canadadotca.

Explore further