Lesson 5.5Lesson 5.5
When to stop
'Good enough' and the cost of over-design
The hookA designer has a strong design. v0.4 tested well — the family loves it, it survives the monsoon, the knees are looked after. But they keep going. v0.5, v0.6, v0.7, chasing a perfection always one version away. Weeks pass, the client grows anxious, the budget bleeds. And somewhere around v0.6 the design stopped getting better and started getting fiddly — over-worked, losing the clarity v0.4 had. Trying to make it perfect, the designer made it worse.
The loop that won't end
Every lesson taught you to improve, but the loop has to stop somewhere, and it doesn't stop on its own, because there is no perfect design — there's always another flaw, another refinement. 'When do I stop?' is one of the most important judgements a designer makes, and it cuts against the cultural celebration of the perfectionist who never settles. In design that instinct is dangerous, because past a certain point continuing to iterate doesn't improve the design — it harms it.
The curve of diminishing — then negative — returns
Early iterations are gold — v0.1 to v0.3 climb steeply, each loop fixing big real flaws; never cut this short. Then the plateau — diminishing returns, marginal improvements. Then, fatally, the curve turns down into over-design, where continued iteration makes the design fiddlier, over-complicated, losing the clarity and confidence it had at its peak. The skill is to stop at the top of the curve — at 'good enough' — not to keep climbing a curve that's descending. Perfectionism is the failure to see that the curve turns down.
Satisficing — 'good enough against the constraints'
Simon's insight (from Lesson 0.3): real designers don't optimise (there's no perfect house, and searching for it is a fool's errand) — they satisfice, finding a solution good enough against the constraints, and stop. This isn't settling for mediocrity; it's recognising 'perfect' isn't a coherent target, and the energy chasing it could be better spent — on the next project, or returned to the client as a finished home they can live in now. How do you know you've reached 'good enough'? You check against the success criteria from your brief (Lesson 2.4) — written at the start so that now you'd have an objective answer to 'is it done?' When the design meets its agreed success criteria and survives its stress-tests, it is, by definition, good enough. Without criteria, 'done' is a feeling that keeps moving the goalposts; with them, it's a checklist you can honestly complete.
The cost of over-design
It costs the design its clarity — over-iteration muddies the strong simple core idea (the parti) under fussy embellishment; often the bravest edit is removing, not adding, but the over-designer only adds. It costs time, money, and trust — endless iteration burns the design budget, delays the home, and erodes the client's confidence (decisiveness is professionalism; endless tinkering reads as its opposite). And it costs the better use of your effort — every hour polishing a good-enough design is an hour stolen from the next family who needs a home.
The deepest reason to stop is that the inhabitant, not the design, is the point — it's easy to fall in love with the design as an object of craft and keep perfecting it for its own sake, but once it serves Lakshmi's life well, additional perfection serves your craft-ego, not her; the family would rather have their good-enough home now than your perfect home next year. Stopping is an act of humility. 'Good enough' scales with the stakes — the right stopping point for a modest tight-budget 2BHK is earlier than for a flagship building; over-designing a simple project is as much a failure as under-designing a complex one. And stopping isn't abandoning the loop forever — you stop this cycle and deliver, but the deepest loop continues across projects and across your career; keep a private record of 'what I'd change next time' so each delivered, satisficed project raises the floor of the next. You stop iterating on the artefact; you never stop iterating on the self.
1. Run your success-criteria checklist — mark each met / not yet / partly. Separate flaws from alternatives — for every change you're tempted to keep making, ask whether it fixes something failing a success criterion or stress-test (a real flaw — fix it) or is just a different-but-not-better option (an alternative — let it go). Locate yourself on the curve — early climb, plateau, or past the peak. Write your stopping declaration — why this design is good enough, referencing the met criteria and passed stress-tests.
Check yourself
3 quick questions — pick an answer to see why.
Q1What happens if iteration continues past the peak of the returns curve?
Q2How do you objectively know a design is 'good enough'?
Q3What is satisficing, in Simon's sense, applied to stopping?
Key terms
- Satisficing
- Simon's principle that designers don't optimise toward a non-existent perfect solution but find one good enough against the constraints and stop.
- Over-design
- The descending part of the iteration curve where continued tinkering makes a design fiddlier, costing it clarity, time, and trust.
- Parti
- The strong, simple core idea of a design, which over-iteration muddies — often making the bravest edit a removal rather than an addition.
You've run the entire loop end to end. But a real project doesn't hand you the modes neatly separated — it hands you a messy brief, a real family, a real plot, a real budget, all at once. Are you ready to take a single live project from a stranger's first vague words all the way to a tested, good-enough design — and prove the method is now yours?