Learning Paths
Overview
When I joined this project, it was clear that our curation ecosystem had become more complicated than anyone intended. Over the years, different teams had built Channels, Paths, Priorities, and Bookmarks to solve immediate needs. Each tool made sense on its own, but together they created a confusing experience for learners, admins, and even internal teams.
The data reflected that.
Completion rates were low. 75% of B2B learners never finished curated channels. 63% of users couldn't add content to a channel the first time they tried. And customers representing more than $89M ARR were consistently telling us the system wasn’t working as well as it should.
It was time to rebuild curation from the ground up, with a clearer model and a more intentional design.
Understanding the Problem
I'd been at the company for a few years and had a general idea of the problems, but now I could turn my full attention to them. I started by listening. I set up conversations (along with my awesome PM partner) with internal teams who worked with customers and this product area. That gave us a good baseline and lay of the land, and then we also interviewed customers themselves. I put together the discussion guide and led most of the conversations, gathering as many insights as we could into what was working well and what wasn't. A few key takeaways from those conversations:
Leaders didn't know which of our curation products was right for their use case, and a lot of them had overlapping features.
There wasn't a way for them to signal to their learners which content was most important.
Customers wanted to collaborate on curations, and we didn't support that very well.
Underneath it all was a deeper issue: there was no shared mental model for how curation was supposed to work.
A Surprising Insight + Incorporating AI
One of the most interesting findings came when we pulled the analytics apart.
Company-created curations consistently had extremely low completion rates. At first, it looked like a motivation or engagement problem. But after digging deeper and talking with several teams, we found the real reason: companies were adding way too much content.
Learning Science best practices suggest around 40 hours of learning per year.
Some companies were creating curations with hundreds of hours of content.
Learners had no chance of finishing. Leaders were left staring at analytics that looked terrible but didn’t explain the real issue. And these bloated curations made it almost impossible for anyone to understand what was important.
This gave us an opportunity to use Iris, our AI agent, in a more proactive and meaningful way. I designed nudges, suggestions, and guardrails that could recognize when curation best practices were being ignored, and gently help leaders build more realistic, more effective learning paths. It shifted the system from being a passive tool to an active partner. For example, if I curation only included videos, Iris can suggest a hands-on lab and short assessment to promote the "watch, practice, assess" learning model for better retention.
Setting a Direction
Before jumping into screen design, I wrote a design hypothesis to align the team:
If we do a great job building Curations, learners will more effectively achieve their learning goals by understanding company priorities, easily finding structured paths, and receiving clear guidance throughout their progress
This will lead to increased learner engagement, higher completion rates, and improved alignment between individual learning progress and company objectives.
And the strategy to achieve it:
1. Pluralsight learning paths will provide learners with high-quality, structured content from the start
2. Company learning will enable learners to align their priorities with company goals, ensuring relevant and targeted learning
3. Custom learning paths will empower learners to design personalized learning journeys tailored to their individual needs
Because this project impacted a large number of teams, it was crucial that we get alignment and buy-in early on. Starting with this hypothesis allowed us to share our vision and get feedback from internal stakeholders. We made some tweaks, but for the most part we were aligned and ready to move forward to low fidelity mockups, and then later to high-fidelity prototypes.
Simplifying the Ecosystem
I mapped out the various curation products and their features, as well as the roles/permissions to manage them. Seeing it all together helped the team understand why the ecosystem felt disjointed, and led to one of the most impactful decisions:
Consolidating these curation products into one would simplify the experience for customers and internal teams.
This dramatically simplified navigation and gave customers one clear place to build and manage learning. It also let us retire outdated tools like Hellfire and eliminate redundant UX patterns.
Not everyone was initially on board. The internal team who curated content for our world-class library had concerns about their quality content being on the same page as a learner's unstructured playlist. They had a good point: we needed a way to differentiate between Pluralsight-created content, Company-created content, and Learner-created content. We solved it by using different visual treatments for each type, and satisfied the content team's concerns.
Making Learning Feel Achievable
One clear theme from user interviews was that curated experiences felt overwhelming. When a learner opens up a learning path that starts with a 40-hour course, most of which isn't entirely relevant, it’s discouraging and engagement decreases. But there wasn't a way for leaders to add smaller bits of content.
To fix this, I discussed with engineering the possibility of allowing for curators to add smaller building blocks: clips, modules, assessments, and external links. We already organized our content this way, and luckily it wouldn't take much engineering time to implement. This made curated paths much more focused and free of unnecessary busywork.
Fixing the Creation Experience
The old experience forced users to leave the creation tool to go search globally, then come back and try to stitch things together. It was a frustrating loop that made a simple task feel complicated.
By bringing search directly into the creation flow, we created a faster, more intuitive experience. Users no longer had to bounce between tools. They could find, evaluate, and add content all in one place. We were able to leverage the existing Search APIs, so the experience was consistent.
Replacing Priorities With Something More Effective
Priorities was supposed to help B2B leaders communicate what mattered, but it never worked the way anyone hoped. Rather than trying to fix it, we introduced due dates as well as a prominent section at the top of the page that allowed customers to highlight the most important content to their learners.
Bringing Structure to Governance
Behind the UI was a complex set of roles and privileges. Ownership, editing rights, SME collaboration, publishing rules, and versioning all played a part.
To help the team align, I created simple “boxes and words” diagrams that stripped away visual distractions. This helped us make fast decisions about how the system should behave before getting into detailed design.
The Outcome
We replaced a scattered ecosystem with a single, scalable Learning Path system. We made content modular, streamlined creation, improved focus with Assignments, and gave leaders AI support to build healthier paths. We reduced complexity, retired legacy systems, and directly addressed customer pain tied to high-value accounts.
Most importantly, the experience now reflects how people actually learn: one step at a time, with clarity and intent.
How I Work
This project reflects the way I like to approach complex problems. I start by listening. I look for patterns. I help teams align on a shared mental model. I create design deliverables like wireframes and high-fidelity prototypes to drive alignment and create a visually-appealing product our customers will love. I strive to design systems that feel simple and supportive, even when the underlying logic is complex.
I love collaborating with others. On this project I worked closely with my PM partner, and also sought input from our design systems team, accessibility experts, and several internal teams.