“Imagine it's a Saturday morning and you want to go into Dehradun city. You need to rent a scooter from the app.”
Simple brief. Our first version needed 9–12 taps to get there — and testers hesitated at almost every one.
They tapped “Rent” and expected a list of bikes. They found the price late. After paying by card, several weren't sure anything had happened. A second scenario — renting for a busy friend — exposed the same gaps.
What students told us about renting a bike.
To find an available vehicle fast and book it in a few taps.
Vehicles not being available, safety and the weather.
Proximity matters: people expect a rental shop within walking distance.
UPI is the default: instant, final, familiar — so it should be the first payment option, not the fourth.
The core requirement: a simplified, low-tap booking sequence.
Where the experience breaks, from the first ad to the last review.
Pick the longest task. Test it three ways.
We mapped every task by number of taps. “Booking a rental” needed 9–12 — the longest, and the reason the app exists — so that's what we tested.
First a heuristic evaluation against Nielsen's 10, to predict where things would break. Then think-aloud sessions with students in an empty classroom: one moderator, one observer, phone audio and screen recording. Then post-task interviews.
The heuristics told us where to look. The think-aloud told us why it mattered.
Six clusters of friction.
Home felt like the wrong starting point. “Create profile” arrived unexpectedly. The back button behaved differently across sections.
Users tapped “Rent” expecting a list of vehicles — and didn't get one. The price showed up late in the process.
After entering card details, people weren't sure whether they had paid. Payment status and next steps were unclear.
Icons were small and ambiguous; the rating stars, filters and calendar slider were never explained.
Multiple pop-ups and redundant steps. Several testers asked, unprompted, for fewer taps.
Users couldn't tell whether filters had worked; the location filter and the popular-shops scroll were confusing.
Show the bikes. Show the price. Show the receipt.
Testers tapped “Rent” expecting a list of vehicles. What should that tap open?
91.7% of students prefer UPI. Where should it sit on the payment screen?
Testers weren't sure they'd paid. What closes the loop?
Ten students, the redesigned flow, two scores.
Nine of ten rated it above the 68 SUS benchmark; P7 (62.5) and P10 (TLX 34.1) are the sessions to watch next. Honest limit: we didn't run SUS on v1, so 84 shows the redesign is usable — not by how much it improved.