The executive studies on a break, not at a desk

The problem

The company sold courses to management and senior leadership, and the goal for that cycle was not a screen: it was widening the channels through which students consumed the content, in order to lift conversion and LTV. The PM brought the problem, and he brought it framed as a business question, not a design request. That changes the work from day one: there was no broken screen to fix, there was a growth hypothesis to test.

The context was the opposite of what I would find later elsewhere. More than thirty developers, one PM and six designers, which means plenty of capacity to build and a real fight over priority. At that size, proposing the wrong thing is expensive, because the queue is long and whatever goes in pushes something else back.

The research

Here I had access to users, and I used it. The research was done with real students, combining interviews to understand the study routine, task completion tests to see where the current path stalled, and card sorting to find out how they themselves grouped the content. Information architecture was the axis of the work: before deciding on any interface, I needed to know what mental model the student already brought with them.

The data point that became the project showed up early: more than 60% of students consumed the course away from home, on a phone, a tablet or a laptop. The company was selling training to executives while picturing someone seated, with time set aside, and the real student was studying standing up, in transit, in the gap between two meetings. The channel we wanted to widen was already in their hand, and the product was the thing that was not there.

More than 60% of students watched classes away from home, on their phones. The course was sold to executives, and an executive studies on a break in the day, not sitting in front of a computer.

The recommendation

I recommended bringing access to class content inside the app, natively, instead of continuing to push the student into another environment. The research did not stop at the video screen: it pulled along the scope that made the class work in that context. Note taking during the class, because someone studying on the move does not have a notebook open beside them. Live classes, supporting materials and notifications, so returning to study did not depend on the student remembering to.

The whole feature was designed out of that, and not out of the internal wish list. It was also what pulled most of the company's design system into existence: the majority of components were born to make this delivery possible, not as a side project in standardisation. With thirty-odd developers building in parallel, the system was what guaranteed the same class looked like the same class anywhere in the app.

The result

Course completion rose by around 40% after the class started living inside the app, comparing the rate before and after launch. Engagement and retention rose by about 30%, measured by return to the platform and by class time actually watched. Those two numbers move together for a simple reason: cutting the friction between intending to study and the class starting is what brings someone back the next day.

The number that closed the case for the business was LTV, which went up by around 25%. That was exactly the metric that opened the project, and it is the reason this case exists: the question the PM brought at the start was never about classes on a phone, it was about how long the student stays with us. The class in the app was the route, not the destination.

What I took away

I take away two things, and both are about how work circulates, not about method. The first is aligning expectations with product and with the PM up front, not along the way. On a team that size, good research that arrives outside the moment of decision becomes a document, and a document does not move a roadmap.

The second is more uncomfortable and it changed how I work: understand the political position of the business before defending pure design. I already had the data on my side, but data does not walk on its own inside a company. It has to reach the right table, at the moment that table is deciding, in the format that table can use. Well argued, badly positioned design loses to a worse proposal that was better placed, and that is not the system being unfair, it is part of the job I was not doing.