Too much platform.
Payments and deliveryOpen city-wide listingsA chat feed for everyone
Useful someday. Too much to build and test in two days.
THE BUILD LOG / NO TIME TRAVEL REQUIRED
From a messy reading-list problem to a working swap. Here’s what we kept, what we cut and what happened each hour.
Try the demo ↗PLAY THE WEEKEND BACK
HOUR / 48
We picked the problem: too many textbooks, too much scrolling.
Use the slider, arrow keys or step buttons.
H00 → H48 / EVERY STEP STAYS HERE
We picked the problem: too many textbooks, too much scrolling.
Asha wrote prompts about the last book students bought.
We asked students where their old books ended up.
Edition numbers mattered more than cover photos.
List a book, find a match, agree a campus handover.
Payments, postage and a public feed came off the board.
Asha drew the browse, match and handover screens.
Kit defined title, edition, course and condition.
The three screens got their first working navigation.
Noor wrote fictional listings for the test set.
The first book grid could be searched by course.
Edition details moved beside the book title.
We could browse, but not yet make a swap.
Kit wrote the rule: each person wants the other book.
We checked duplicates, missing editions and no match.
No match now became a useful answer, not a blank page.
Listings started feeding the match screen.
Noor gave the buttons concrete action labels.
Two sample listings produced one reciprocal match.
We removed a listing mid-match to test the failure.
The interface asked the tester to pick another book.
A public campus location replaced home addresses.
Asha checked focus order and button names.
We saved the build before the late-night edits.
One working loop. We agreed to rest before testing.
We reopened the notes and checked what still mattered.
We chose fixes over adding more features.
Noor watched one student try the task without help.
A tester picked the wrong edition from the grid.
Asha made edition information more prominent.
Kit added a plain explanation of reciprocal swaps.
The next tester could say why the books matched.
We tested the screens on a narrow viewport.
Book details and next steps got larger hit areas.
We changed meet anywhere to choose a public place.
Noor ran the revised task with the remaining testers.
We recorded task completion, not adoption promises.
Nine of twelve testers completed the sample swap.
No moderation, real accounts or live campus rollout.
New ideas went to the next-week list.
Noor made the problem, prototype and findings deck.
Asha prepared clean frames for the presentation.
Kit checked the demo route and resource bundle.
We practised the pitch and shortened the introduction.
Everyone explained their own contribution.
We checked empty matches and narrow screen layouts.
Code, notes and slides went into one bundle.
Three minutes, one working swap and honest limits.
We received the People’s pick and noted what comes next.
THE SCOPE SCISSORS
a smaller thing that works
Useful someday. Too much to build and test in two days.
A small, understandable loop we could put in front of someone.
A NOTE FOR NEXT TIME
We spent our first hours asking questions. That stopped us building a marketplace when the actual problem was finding the right edition.
FROM WEEKEND BUILD TO WORKING IDEA