The difference between a build that reaches roofing in twelve months and one that crawls for three years is rarely money alone, it is planning. Clients who map their cash flow to construction stages finish; clients who buy materials as money appears tend to stall at lintel level and never quite restart smoothly.
This guide shows how to plan a build-to-roof project end to end: defining scope, sequencing purchases against stages, choosing contractor engagement models, scheduling around the rainy season, and keeping a contingency that absorbs shocks without derailing the whole plan.
"Build to roof" is builder's shorthand, but your project needs its own definition in writing, because every later decision (money, contracts, inspections) hangs off it. The standard scope: foundation, blockwork, lintels, decking where designed, roof sheeting complete, and services first-fixed, the closed, weather-tight box. Explicitly excluded: rendering, screeding, all finishes, and usually external works (fencing, driveway, soakaway depending on the plot plan).
Write the exclusations down as deliberately as the inclusions, the "but I thought the quote included plaster" dispute is a scope sentence that was never written. The build-to-roof line is where the sequence's economics naturally pause: structure is permanent and patient; finishes are consumable and exposed. (The foundation-to-roof guide explains what each stage of this scope contains and inspects.)
The programme's costs arrive as peaks, not a slope: foundation (excavation, concrete, blocks to DPC) is the first real spike; decking (formwork, steel, a continuous pour, and its curing) is the biggest; roofing (timber, sheets, flashings) is the last major spike; between the peaks the spend flattens to labour-and-cement lines. Planning is mapping your income to those peaks in advance, so no peak meets an empty account.
The practical tool is a phase-funding table: each phase's estimated cost (run the build-to-roof estimator on your drawing for real numbers), the month it falls due under your chosen start season, and the funding source against it, savings tranche, contributions, loan drawdown. A phase funded by "we'll see" is a phase that introduces itself again later.
Two funding rules that separate finishers from stallers: fund each phase fully before it starts, a phase begun on partial money is borrowed trouble, because half-completed concrete is worse than unstarted concrete, and never fund a peak by borrowing from a later phase's materials; the sequence collects that loan with interest at the worst possible moment.
With money mapped, the purchasing calendar follows the sequence (the full method is in the materials estimation guide). The build-to-roof-specific version:
The discipline in one sentence: materials on the ground before each phase's crew, never three phases before. Storage is cost; delivery timing is savings.
Three honest models for a build-to-roof programme, each matching a different kind of owner:
Whichever model, keep the stage-gate payment structure: mobilization covers materials only, then payments follow the gates, foundation passed, DPC inspected, deck cured and struck, roof hosed and dry. Money that follows gates never has to be chased; money paid ahead of gates schedules its own supervision.
The rainy season (roughly April–October) is not a stop sign; it is a constraint set. The planning moves:
A build-to-roof plan without contingency is not a plan; it is a forecast with confidence issues. The honest structure: 10–15% of the structural programme, held as a separate line, released only by named events, over-dig at foundation, steel price movement before the decking order, a storm event, a gate that fails and re-opens a phase. Contingency is not "extra tiles money"; it is the programme's shock absorber, and spending it on scope upgrades (the parapet you "always wanted") converts the absorber into a spring.
The absorption rule: when a shock consumes contingency, the schedule absorbs the remainder, the plan flexes dates before it flexes stages. Never the reverse: a schedule that compresses stages (curing, gates) to protect dates is building the failure into the walls (see common construction mistakes).
The finished plan fits on one page, and its existence is the difference the intro promised: scope sentence, phase-funding table, purchasing calendar, engagement model with gate payments, start season with weather buffers, contingency with named release events, and the six inspection gates with dates to be filled as they pass. Clients who carry this page finish at roofing in planned time; clients who carry only enthusiasm carry it to lintel level, where it stalls, not from lack of money, but from lack of mapping.
A plan on paper enforces nothing; a plan attached to a weekly hour enforces everything. The owner's minimum viable involvement in a build-to-roof programme is one hour a week, structured: walk the site against the phase (is the work where the calendar says it is?), check the gate (did this week's inspection pass and get dated in the site book?), verify the delivery (did the week's materials arrive and reconcile with the store?), and make the next decision early (whatever the next phase needs from you, schedule confirmation, payment, approval, decide it this week, not on the day it blocks a crew).
That hour is the difference between the two clients in the intro: the one whose money "appears" buys materials against stalls, and the stall is discovered in the middle of a phase; the one who runs the weekly hour discovers the next phase's gap while the current phase is still moving. Cash flow shocks and material shortages both obey the same law, they are cheap seven days out and expensive today.
The contingency rule (Step 6) becomes concrete with the three shocks that actually arrive on build-to-roof programmes, and the correct response for each, because the wrong response to a shock is how a plan becomes a stall:
| Shock | The flexible response | The stall response (avoid) |
| A material price jump (cement, steel) | Re-sequence: pull forward phases whose materials are stable; use the contingency line for the affected order only | Stop buying entirely "until prices calm down", the crew ages on site and every phase behind it slips |
| A cash-flow gap between tranches | Downshift to a lower-burn phase (blockwork continues; the decking order waits for its funding) with a dated restart | Start the decking on partial money, a half-funded pour is the single most expensive mistake in the programme |
| A failed gate (rework found) | Fix immediately, re-dated in the site book, schedule absorbs it via the buffer | Build past it, the defect compounds under everything after, and the schedule is lying from that day forward |