Coleap: What I did to achieve a 70% completion rate

Role

Product Designer. Collaborated with PM, Engr, and Learner Experience. Success metric: increase completion rates (measured by % participation in activities).

Context

Coleap wants to make lifelong learning effective and more engaging. 85% of jobs that today’s workers will have by 2030 have not been invented. This is a massive continuous reskilling challenge, and traditional learning methods are either too slow to adapt or ineffective.

Problem

“We are not effective enough: learners do not participate in activities as they should.”

Research insights

• On-demand learning platforms: low completion rates; 5–15%.
• “I need a simple way to know and connect with relevant people.” (Coleap learner)
• Slack channels: in existence but disconnected from the learning platform. I used this to study conversation patterns.
• Relevance → Networking, Accountability, Fun.
• How can we leverage learners’ wants to drive completion?

Challenge 1: Defining relevance

Question:

How do we judge relevance? What parameters do we use to match peers?

What I did:

Monitored conversations and carried out surveys.

Insight:

“I like my peers”, “I am inspired by my peers”, “I have something in common with my peers”.

Solution:

Distill into a couple of parameters: similar availability, hourly commitment, goals and needs, job/company, projects/deliverables, and experience, with a ‘forced’ balance of diversity.

Challenge 2: Profile completion

The more information we can get into the user profile, the more effective for matching. How do we get learners to have close-to-complete profiles?

Option 1:

Stay with simple forms but explain why they are needed.

Option 2:

Make forms more interactive and engaging.

Solution:

Option 2 is chosen after an A/B test.

Challenge 3: Onboarding

“Although interesting, the profile completion makes the onboarding too long.”

What I did:

Track the most critical parameter: similar available times. Prioritise that and nudge users to complete their profiles later.

Challenge 4: Who does the matching?

Constraints:

< 3 months to ship, too short to build an effective matching algorithm.

Solution:

Rough matching + review & approval by admin.