The critical path method has an enterprise-software reputation — Gantt charts, arrows everywhere, a scheduler in a trailer. Strip the ceremony and it answers the two questions every flipper asks weekly: which delays actually cost me money, and which don't? On a job where every day of hold has a dollar cost, that's not project-management theory. That's underwriting.
The idea in one paragraph
Some tasks wait on others: you can't hang drywall before the rough-ins pass inspection, can't pour before you form, can't paint before you mud. Chain those dependencies together and one sequence of tasks emerges that determines the finish date — the critical path. A one-day slip anywhere on that chain moves your closing by a day. Everything off the chain has float: room to slip before it matters.
Float is free delay — until it isn't
Float is the most useful number nobody computes by hand. If the tile can start any time in a nine-day window without touching the finish date, those nine days are free — take the cheaper tile setter who starts Thursday, nothing happens to your margin. Day ten starts billing you at your daily holding cost like any other slip.
This is why gut-feel scheduling burns money in both directions: rush fees paid to compress tasks that had float (bought nothing), and casual reschedules of tasks that had none (each one moved closing). Without the arithmetic, the two look identical.
Dependencies, minus the jargon
Nearly every relationship on a residential job is "finish-then-start": form then pour, rock then tape. Two refinements earn their keep:
- Lag — finish-then-wait-then-start. Pour, then five days of cure before you strip forms. The wait is part of the schedule even though nobody's on site.
- Start-together — the cure clock starts a day into the pour, not after it. Occasionally the honest shape of the work.
That's 95% of it. If a scheduling tool makes you think harder than that, the tool is the problem.
The ten-day rain test
Here's the question that separates a schedule from a poster: if it rains for ten workdays during framing, when do I close, and what does it cost? Answering it by hand means re-walking every dependency, remembering that weather shoves parallel trades too, and translating workdays across weekends — for one hypothetical you might discard. So nobody does it, and weather becomes a shrug.
Done right, the answer is three numbers: the new closing date, the holding-cost bill for the slip, and — the one nobody can eyeball — which tasks just became critical. A ten-day delay upstream can eat a downstream task's float entirely, turning a casual "we'll get to it" item into one that moves your close. Knowing that before committing to the delay is the whole game.
What this looks like day to day
Not a scheduler in a trailer. A schedule where the critical chain is marked, every task carries its float, and the finish date shows next to what a slip costs in holding dollars — so when the electrician calls to push a week, you look at one screen and answer with a number instead of a feeling. The budget and the schedule are the same discipline pointed at money and time; the flippers who compound returns run both.