Giving job-seekers a better place to learn

An image of the UI for the empty state of the job tracker

Pathrise is an online mentorship program for job seekers in tech. Each fellow in the program is matched with a career mentor and an industry mentor who work together to help the fellow strengthen their job search.

Navigating the journey of applications, interviews, and offer negotiation isn't easy – and fellows had questions. But with each mentor managing an increasing caseload of fellows, Pathrise needed a way to help mentors support fellows at scale.

Goal

Increase engagement on the Pathrise knowledge base

Pathrise had an existing knowledge base, but it was underutilized. We hypothesized that improving the knowledge base would increase value for fellows while freeing up valuable time for mentors.

Why are fellows struggling with the existing knowledge base?

I turned to Fullstory, a digital experience analytics platform, to identify users' pain points within the existing knowledge base.

There was a prevalence of rage clicks in the sidebar, especially on section titles and on locked content.

Above: A confused user rage clicks on sidebar content.
Pain 1: Overwhelming
There were hundreds of resources and no way to search. Content could only be discovered by browsing the sidebar, but from that point onward, it just got more confusing.
Pain 2: Inconsistent
Sidebar navigation items ranged from program track ("Software Engineering") to curriculum ("Winter Curriculum") to job stage ("Profile") – all with no hierarchy.
Pain 3: Irrelevant
Software Engineering resources outnumbered the other track's resources by 20:1. Non-SWE users had to scroll to the very bottom to access their track-specific resources.
Pain 4: Uninformative
When new users clicked on "My Resources," they saw a blank page with no context. When they clicked on locked content, nothing happened. Many users bounced, assuming these pages were broken.

Ideation

How might we create a helpful, consistent, relevant experience?

Learning from others

The Pathrise knowledge base needed to be able to handle a variety of content, from workshop recordings to coding challenge solutions.

I examined a variety of comparative experiences: news aggregators, technical support sites, and online learning platforms. These familiar, vetted UI patterns heavily influenced my work.

An Apple news grid view.
Above: An Apple News grid view. Left alignment and white space around informational text makes this format highly scannable.

Card sorting software wasn't in the budget for our lean startup. So I invented a free workaround.

I used a combination of pre-made cards in Asana's Kanban view to mimic card sorting software. Our team was able to gain valuable feedback at zero cost.

The outcome? Users were confused by our existing navigation structure. They wanted to browse content by job search stage.

Above: A card sort participant names the categories he's created.

Multiple happy paths. Reusable components.

I reused components across multiple pages in the flow. Not only did this reduce technical complexity, it also established a familiar pattern for the user.

User flows showing the UI of the knowledge base and how they link to one another

At teh time, we had no documentation for interaction design guidelines. As a precursor to a full design system, I added an interaction library to articulate important state changes to the developers.

Pathrise Resources Interaction Library
Above: A few of the interactive components I introduced

Solution

Knowledge at your fingertips, on your terms.

Browse based on job search stage

Card sorting revealed that users grouped content primarily by job stage. They can now do this in multiple ways: from the home page, on the individual resource pages, and within the new mega menu.

Find what you want directly

If a fellow hears about a particular article in a workshop or mentorship session, it's now just a search away.

Filter by industry, resource type, & more

With multiple options for customizing via filters, the knowledge base is now a friendlier place for non-engineering fellows.

Whoa, that's actually a lot of tags. I recognized that I hadn't set constraints for engineers, or how to handle overflow. This was a great learning opportunity for me, something I'd improve if I did the project all over again.

Save your faves

Returning users shouldn't have to dig for the same resources twice. I aimed to reduce friction for repeat users by adding a bookmarking feature. Bookmarks are accessible from the mega menu.

Outcome

It's empowering to see people take learning into their own hands.

While we weren't able to accurately measure the reduction in overhead, we were impressed by engagement and feature retention mentrics. Analytics over the months post-launch showed that most users who visited the knowledge base returned to find another resource within 3 days.

75%
3-day feature retention  (users returned for subsequent visits)
56%
Weekly feature retention (users returned for subsequent visits)
83%
Conversion (visitor navigated to 1 or more resources)