Lesson 6.3Lesson 6.3
Prototype, test, and iterate
Closing the loop on the live brief
The hookYour combined concept is elegant — the service spine, the flexible main volume, Priya's study loft, the dignified double-duty seating. On paper it solves everything. That's exactly why you should trust it least. Every concept looks right on paper; the paper is silent, still, and flattering. The only way to know is to make it real enough to break, cheaply, and then try hard to break it.
Prototype: cheaply, in the right order
Climb the fidelity ladder from the bottom, matching each prototype to the question of the moment. Napkin and bubble (is the spine-and-flex concept sound?), then the question that matters most in 650 sq ft — does it all fit? — so the 1:50 plan with true-scale movable cut-outs (the two-cook kitchen, the flexible volume, the loft footprint). Then a foam massing model (does the loft make the room below feel sheltered or oppressive; is there headroom?), eyes lowered to its scale. Then fast AI renders to confirm the feel (steered toward the specific, used for desirability only). Then the daylight study (does dawn reach the grandmother's prayer corner?), and the good/better/best cost tiers so the family chooses within ₹8–10L. Build to learn, not keep, sequencing so the riskiest cheap question (does it fit?) is answered first.
Test: walk the family and the monsoon through it
Walkthroughs first — slowly, moment by moment. Walk Priya home to her loft; the grandmother to her dawn prayer corner; Meena and the grandmother both into the kitchen at once. Then stress-tests — the monsoon through the threshold, the power cut through a hot afternoon, ten years through the grandparents' knees. The testing finds a balanced set: wins to protect (dawn light reaches the prayer corner, the kitchen fits two cooks, Priya loves the loft — 'it's like a little treehouse that's mine') and flaws to fix (the loft ladder is unsafe for a nine-year-old at night and impossible for the grandparents; the entrance has no threshold, so monsoon water and wet shoes come straight in). Priya's reaction is desirability gold you couldn't predict from a plan; the monsoon flaw was invisible until you walked the condition through.
Iterate: the right distance, protecting the wins
Diagnose each flaw's depth. The monsoon threshold is a surface flaw — loop back to Prototype, add a proper threshold (level change, covered transition, shoe spot, drainage that slopes away), re-test. The unsafe loft ladder is deeper — a concept flaw, since no railing fully fixes the danger — so loop back to Ideate and reach into your reserve: instead of a ladder-reached loft, a half-level study nook raised a few steps, reached by a gentle safe stair that doubles as storage. It still gives Priya her elevated, treehouse-like space (protecting the win she loved), still reclaims the section, but is safe for the child and navigable for everyone. Re-test, and check it didn't break a neighbour: does the gentler half-level still leave headroom below and room for the flexible volume? It does. Version it: v0.1 → v0.2 (threshold added, light/kitchen held) → v0.3 (loft became safe half-level, Priya's joy and the section preserved).
Knowing when to stop
Re-run the success criteria: Priya has a study spot that feels hers (yes, and she loves it); grandparents have knee-friendly seating and a dawn prayer place (yes); Meena's kitchen works for two (yes); the home feels presentable (yes); within budget (yes, at the 'better' tier). Every box ticked, every stress-test passed. You're at the top of the curve, so you stop — you satisfice, resisting the pull to chase v0.4, v0.5 into over-design, because the design meets its purpose well and the family deserves their good-enough home now.
The test that matters most is the one with the real, overlooked person in the room — your imagined walkthroughs carry your bias; the moment Priya said 'a little treehouse that's mine' and the moment the grandparents couldn't manage the ladder were the real tests your imagination would have flattered past; get the actual vulnerable users in front of the prototype. The iteration that improved the design came from your reserve — the safe half-level wasn't a fresh invention under pressure but an option adjacent to ideas you'd already generated and kept, which is why you never throw the field away. And resist the two opposite end-game failures together — the anxious designer keeps hunting for more flaws past good-enough; the tired one declares victory before re-testing the fixes; the discipline is to fix what the test found, re-test to confirm, check against the success criteria, then stop with confidence.
1. Prototype to your riskiest question — name the cheapest prototype that answers your concept's biggest 'does this actually work?' and build it. Test against people and reality — walk your most vulnerable persona slowly, then one stress-test, listing wins to protect and flaws to fix. Iterate with discipline — diagnose each flaw's depth, loop the right distance, fix while protecting the wins, reach into your reserve for the deeper fix, re-test, and version it. Declare good enough — run your success criteria as a checklist and write your stopping declaration.
Check yourself
3 quick questions — pick an answer to see why.
Q1Why trust an elegant on-paper concept the least?
Q2In 650 sq ft, the riskiest cheap question to answer first is which?
Q3The unsafe loft ladder is a concept flaw, not a surface flaw. What does this mean for iteration?
Key terms
- Fidelity ladder
- The sequence of prototypes climbed from cheapest upward — napkin, scaled plan, foam massing, render, daylight study — each matched to the question of the moment.
- Flaw depth (loop distance)
- Diagnosing whether a flaw is a surface problem (loop to Prototype) or a deeper concept problem (loop to Ideate), so you iterate the right distance.
- Satisficing
- Stopping once the design meets its purpose well and passes every success criterion and stress-test, resisting the pull to over-design.
You've run the entire loop. The work is finished. But two things remain: assembling it into something you can hand the family, and the most important step for your growth — the one beginners always skip. What is it, and why does the deepest learning happen only after the project is done?