Coursera

The Challenge

This project started small: the team was in the midst of implementing a design system, and I was tasked with taking a holistic look at the card interface that we were using to present courses.

After combing through existing qualitative research, we realized that the problem was larger than an inconsistent card interface: our learners were overwhelmed by our catalog and were having trouble choosing courses, leading to lower enrollments.

We decided to take a step back to look at this larger problem, rather than restrict ourselves to the card interface. The new project objective became to reimagine how learners find, understand, evaluate and compare courses that will help them achieve career success.

Clarifying the Problem

We realized right away that this needed to be a cross-team initiative, and a partnership between our Growth and Enterprise teams was born. Throughout the entire process, I worked very closely with a designer on the Growth team. We also included members of other teams — both design and cross-functional — in numerous workshops and share outs.

We started off with a thorough audit of the current experience, mapping out the user flows, looking at the quantitative data, and reviewing the large library of existing qualitative research reports. We also gathered input from across Coursera: speaking to our customer-facing teams, reviewing our customer community, and syncing with other product teams.

We pulled all of the relevant insights from the qualitative research library, then summarized, grouped, and prioritized them so that they could be used to inform future decisions.

At Coursera, we have a variety of audiences with quite different needs. For example, a learner on the consumer side may want to learn the skills they need to start a new career, a Coursera for Business learner might be more focused on completing the courses their manager has asked them to do so that they can do better at their current job, and a Coursera for Campus learner may want to find courses to supplement their existing curriculum.

Our end goal was to design a flexible discovery framework that would be able to adapt to our different product areas, meeting the needs of our varying audiences. To do this, we needed to understand the differences. We had a few different audience segmentation resources to pull from, and we gathered these together and created a matrix to summarize. This matrix was used in combination with personas to ground workshops.

	Re-skillers	Up-skillers
High Specificity (Late in learning journey)	"I know which skills I need for the job."	"I know which skills I need to advance, and I'm curious about a specific topic."
Low Specificity (Early in learning journey)	"I don't know which skills I need for the job. I try to identify them using job postings, trend articles, how-to reskill articles, Reddit, YouTube, and my network."	"I don't know which skills I need to advance. I try to identify them by speaking to my manager and my peers."

In a series of cross-team workshops, we used the prioritized insights and audience matrix to align on three key problem statements.

Problem #1	As a learner without specific content in mind, I find Coursera's content overwhelming. I don't know where to start and, once I figure out where to start, I can't dig deeper. This is preventing me from finding courses that meet my career needs.
Problem #2	As a learner with specific content in mind, I find it difficult to decide which course fits my preferences because there are so many different options. I don't know how I can narrow down my search to find what I need. This is preventing me from finding courses that meet my intended career outcome.
Problem #3	As a learner, I'm not sure what the information on the course cards means. I think some of it is relevant, but I wish there was more information to help me decide which course I should take. I'm finding it hard to compare courses against one another to find one that fits my goals and lifestyle needs.

We also did a thorough competitive analysis, looking at both direct competitors and products outside of our space with leading and innovative discovery experiences. In addition to helping us understand where we stand, we were able to pull insights from these experiences.

This spectrum plots these other products from powerful search experiences to compelling browse experiences.

Patterns were identified across the spectrum, and were used to inform ideation workshops.

Design

Now that we had a defined audience and a clear set of problems to solve, we set out to design a solution. In a cross-team workshop, we created a handful of hypotheses — how might we solve the identified problems? We grouped and prioritized these hypotheses, and synthesized them into three distinct concepts. For each concept, we wireframed out the various user flows that our learners would go through to discover relevant content.

Concept #1: Bookstore Owner

This concept focused on being helpful in the decision making process and anticipating career-switchers' needs and motivations.

Hypotheses:

1. Personalization: We believe that we can use learner data (provided both explicitly and implicitly), to provide learners with a more personalized experience that contains relevant content.
2. Guided search: We believe that helping learners refine their search with guided search will help them narrow down content.
3. Simplified course cards: We believe that simple course cards will be sufficient for learners to make a decision.

Concept #2: Networker

This concept focused on being intuitive and outcome-oriented to empower learners to find the best-fit content.

Hypotheses:

1. Supercharged search: We believe that combining easy-to-use, powerful filters with search will provide learners with an easy entry point, and will then help them narrow down content.
2. Searchable metadata: We believe that enabling learners to navigate through courses by clicking on the metadata will help them find courses more intuitively aligning with their interests and/or learning goals.
3. Links to similar content: We believe that providing learners with similar content (e.g. more with same filter) will prevent dead ends.

Concept #3: Curator

This concept focused on being simple and intentional to help learners feel less overwhelmed.

Hypotheses:

1. Progressive disclosure (collections & canned searches): We believe that showing learners collections and canned searches when they are starting to browse, rather than courses, will feel less overwhelming.
2. Progressive disclosure (structured search results): We believe that structuring search results into relevant sections will help learners narrow down content.
3. Following: We believe that allowing learners to “follow” topics/roles/skills/partners that they are interested in will help them find relevant content and get started.
4. Single/multi course clarification: We believe that we can use different styles for the course cards to help learners understand the differences between our content types.

We worked with a user research agency in India to test and validate these three concepts. These are some of the key insights that we gained from the research:

• ✅ Personalization: Learners were interested in providing their data if it would give them relevant recommendations.
• ❌ Guided search: Learners were confused by the guided search that we presented. They preferred something simpler.
• ❌ Simplified course cards: Learners wanted more information on the course cards, not less, to allow them to make a decision.
• ✅ Supercharged search: Learners saw a lot of value in more powerful filters. Once our more personalized recommendations set them down the right path, they were interested in applying granular filters in order to find a match that meets all of their requirements.
• ❌ Searchable metadata: Learners did not intuitively click on the metadata in order to navigate through content.
• ✅ Links to similar content: Learners appreciated the related content that was provided as they browsed.
• ✅ Progressive disclosure (collections & canned searches): Learners were much less overwhelmed when presented with collections and canned searches, rather than courses, when they were starting to browse. These provided a good starting point for their search.
• ❓ Progressive disclosure (structured search results): We didn’t get a clear read on this one from the research, but plan to revisit the idea later.
• ✅ Following: Learners expressed interest in “following” topics/roles/skills/partners that they are interested in so that they can stay up to date.
• ❓ Single/multi course clarification: The styles that we tested to help learners understand the differences between our content types did not help. However, we believe there’s value in looking into this further.

We used these insights to create one refined concept.

Showcase

Meanwhile, we also started the process of presenting this work to leadership, as well as the larger Coursera team, in order to get the support and resources we would need to see it come to life in the product.

To make the final concept as powerful as possible, we decided to apply visual design to the wireframes. During Coursera’s make-a-thon week, we conducted a design sprint. We used a 20 second gut test and a mood board to align on a visual design direction, then applied this design to the main user flows.

Ultimately, we identified eight user flows, across our various learner types and product areas, that told the full story of Coursera’s product. We created designs for each of these flows. Here's an example of one:

We did a series of roadshow presentations to share these final eight user flows with leadership and various teams across Coursera.

The Results

The roadshow presentations created the leadership buy-in that we needed, and a new product pod was created to bring the concept to life in the product.

Using the concept as a “North Star,” that pod has started building out specific aspects, one at a time, using a hypotheses-based approach to build and test the impact of each. One of the first tests — new search filters — led to increased engagement with the filters, making it easier for learners to find and compare courses relevant to their career goals.

The final concept was presented at Coursera’s annual conference, which generated a lot of customer excitement. We heard from many customers: “When can we have this?”

Role

Lead Designer

Company

Coursera

Next Project