Advance Learning

More than half of new students gave up at the sign-up form. Of the ones who got through, most never opened a course. The menu labels were vague, the dashboard was crowded, and nothing told a student what to do next.

We could not talk to students directly. The client was an embassy, the students were in another country, and the platform was in a language I do not speak. So the evidence came from the numbers, a short list of student quotes the client passed on, and the product itself.

One quote stuck: a 13-year-old said he quit when he saw too many choices on the sign-up page.

The form asked for everything up front. We cut it to what the school needed to create an account, and moved the rest to the settings page where a student can fill it in once they are inside. A seven-day trial replaced the wall, so a new student saw a course before they saw a payment screen.

The order mattered. Fixing the dashboard first would have polished a room most students never entered.

The dashboard went through two rounds. The first round cut the clutter. The second gave every state one clear action: continue the course you were on, or start the one you have not. Themes for each grade level came from the client as a requirement, and they earned their place, because a ten-year-old and a seventeen-year-old should not see the same school.

Wireframes were reviewed with the engineers before anything was drawn in high fidelity. That is where we caught what would not fit the build window.

The platform had no shared components. The same button looked different on every screen, and nobody could say how many versions there were. I ran a session with the engineers to list every element in use, sorted the list into atoms, molecules and organisms, then rebuilt the components from that inventory.

Halfway through, a retrospective surfaced a problem I had caused. Engineers were guessing at edge cases because my hand-offs did not cover them. From then on every flow shipped with a hand-off document: states, interactions, and what happens when something goes wrong.

The redesign launched in December 2022. The embassy approved it in January 2023 without a second round, which for a government client is the fast path.

I do not have post-launch numbers for sign-up completion. The client owned the analytics and did not share them after hand-over, so this study makes no claim about them.