How I took a dashboard for K2U learners from interviews and wireframes to a build that helped lift course completion by 24%.
K2U sells online Salesforce courses and exam preparation to individual learners. About 80% of the business is B2C, and most learners are in the US and the UK. During my time I owned a portfolio of 57 courses and launched about 45 of them myself, taking each from demand research through launch to maintenance.
Courses are self-paced. A learner buys, logs in alone, and works through the material around a job and a life. That makes the weeks after purchase the most fragile part of the whole product.
Selling a course is only half of the job. The learners who finish are the ones who pass their exam, come back for the next course, recommend us and, over time, are worth far more to the business. The learners who stall quietly drift away.
My goal was to make progress visible, so learners could see where they stood and what to do next. I framed it as a retention and lifetime-value problem, not only a completion problem.
Individual learners, mostly working professionals preparing for Salesforce certifications or building skills for a job. Premium learners also had a 24-hour response promise on their questions, and I managed the waiting queue for those questions, which showed me where learners got stuck.
I spoke with learners after they bought. I asked why they bought, and where they got stuck. What I heard shaped the priorities for the dashboard more than any assumption I had walking in.
I combined those conversations with what I saw in the premium learner questions queue, where repeat problems showed which parts of a course lost people.
I define the KPIs for every product at launch and track them afterwards. For this one the main number was course completion, with learner engagement and repeat purchase as the longer-term signals the dashboard was meant to support.
I turned the findings into wireframes and written requirements, made the case to management and got it approved, then led it through the build with the engineering team. I owned the design and requirements; engineering built it.
I owned the decisions on what the learner should see and why, settled the trade-offs as they came up, and signed off what was built against the requirements. My job was to keep the learner at the centre of every trade-off.
24%
Course completion rose by 24% after the dashboard launched, measured before and after launch.
The dashboard was one part of a wider effort to keep learners engaged. I also ran live bootcamps for learners, streamed on LinkedIn and YouTube. 450+ people attended, and about 17% bought a paid course within 24 hours. When one instructor did not show up, I rebuilt the session around a quiz with loyalty-point prizes and team-led sessions.
I then proposed the next step: a predictive retention and lifetime-value feature to spot learners at risk before they stall. Part of it was built before the acquisition.
Sandeep Tomar