Decide · Lesson 1 of 8
Why a hard deadline is a design tool, and what the day actually looks like hour by hour.
4 min read
Before the next lesson
Twenty-four hours is not a stunt. It is a constraint chosen on purpose, because the thing that kills most side projects is not difficulty — it is duration. A project that lives for six weeks has six weeks' worth of chances to be abandoned. A project that lives for one day has one.
The deadline is a design tool. Every hour you spend is an hour you cannot spend on something else, which means the deadline makes decisions for you that you would otherwise agonise over for a fortnight. Should there be user accounts? Not today. Should it support teams? Not today. The clock answers questions that taste and ambition cannot.
Launched means all four of these, and nothing more:
Not "feature complete". Not "production ready" in the sense a large company means it. Live, useful once, and seen.
And one thing more, which is a condition rather than a milestone: from hour zero, everything is yours. The code sits in your repository, the app and the database run on infrastructure you hold the keys to, and the domain is registered in your name. A day is short enough that it is tempting to borrow someone else's platform to move faster. Don't. Lesson 4 covers the stack that keeps ownership and speed in the same hand.
The sprint is deliberately front-loaded with thinking and back-loaded with polish. Most people invert this — they start coding at hour zero and start deciding at hour eighteen, which is why they end the day with an impressive half of something.
| Hours | Phase | You are producing |
|---|---|---|
| 0–2 | Choose | One idea, three rejected |
| 2–4 | Cut | A single promise, and a kill list |
| 4–6 | Set up | A deployed, empty, live app |
| 6–14 | Build | The working spine, ugly |
| 14–18 | Polish | Something that looks deliberate |
| 18–21 | Sell | A page that explains itself |
| 21–24 | Ship | Pre-flight, launch, first ten people |
Eight hours of the day are unaccounted for, and that is intentional. You will sleep, or you will lose an hour to a dependency that hates you, or both. A schedule with no slack is a schedule that fails at its first surprise.
Protect the block. The sprint dies on interruptions far more often than on hard problems. Notifications off, phone in another room, the day marked busy.
One stack, chosen once. Lesson 4 names it; take it or take your own, but decide in the first six hours and stop shopping. Today is not the day to compare frameworks, try a database you have only read about, or restructure how you write CSS. Novelty is where the hours go.
Ugly is allowed until hour fourteen. Before that, styling is procrastination that feels like progress.
When you are behind, cut scope — never the deadline. This is the single rule that separates people who ship from people who have a folder of nearly-finished projects. The deadline is fixed. The product is the variable.
Pick your day and write the finish time on something you will see all day. Not "this weekend" — "live by 9pm on Sunday". Vague deadlines do not cut scope, and cutting scope is the whole method.