March 26, 2026

The budget that matters most is the first one

Projects rarely blow up in construction. They blow up two years earlier, in a spreadsheet, when someone writes down a number that was never real.

The first budget is the one that gets a project approved. It goes to the board, into the pro forma, into the lender’s model. Every decision afterward is measured against it. If it was thin — priced off the wrong benchmark, missing escalation, light on soft costs — the project spends its whole life trying to catch a number it never had a chance of hitting.

We’ve built and reviewed first budgets for towers, industrial facilities, retail expansions and master-planned communities across Canada. The failures follow patterns, and the patterns are avoidable.

Where first budgets go wrong

Borrowed benchmarks. The most common failure is pricing a project off a different project — a different building type, a different city, a different year. A per-square-foot number from a suburban low-rise doesn’t survive contact with a downtown high-rise site, and a 2022 number doesn’t survive 2026. Benchmarks are a starting point for questions, not a substitute for measurement.

Escalation applied to the wrong date. Construction is bought over years, not on the day the budget is written. Escalation has to carry costs to the mid-point of construction — sometimes further for long-procurement packages. A budget escalated “to today” on a project that tenders in eighteen months is quietly short by whatever the market does in between.

Contingency by convention. Five percent because the pro forma had room for five percent is not risk analysis. Early budgets carry the most unknowns and should carry the most contingency; we’ve written more on sizing it in how much contingency is enough.

Soft costs treated as an afterthought. Consultants, permits, development charges, insurance, financing costs, testing, commissioning, FF&E — on many projects these run 20–30% on top of construction. First budgets routinely carry a thin allowance that gets consumed before design development ends.

Site conditions assumed clean. Nobody paid for the geotechnical investigation yet, so the budget assumes the ground cooperates. The ground is under no such obligation.

What a defensible first budget contains

A budget you can take to a board has five properties.

A measured basis. Quantities taken from the drawings that actually exist — even concept drawings support real measurement — with the gaps filled by explicit, written assumptions rather than silent ones.

Current, local pricing. Rates from projects genuinely bought in this market, this year, at this scale. This is where independent estimating earns its fee: the pricing has no position to defend.

Escalation tied to the real schedule. Not the hoped-for schedule. If the realistic path to shovels is two years, the budget says so.

Contingency in named buckets. Design, construction, escalation and owner contingency, each sized to what’s actually unknown, each with a plan to burn off as the unknowns resolve.

A written exclusions list. Every budget excludes something. The dangerous ones exclude things silently. A one-page list of what’s out — and who’s carrying it instead — is the cheapest insurance a pro forma can buy.

Pressure-testing a number you’ve been handed

If the first budget arrived from elsewhere — a broker’s feasibility, a contractor’s early number, a consultant’s cost plan — five questions expose most weaknesses:

• What drawing set and date was this measured from?
• What escalation rate was used, and to what date does it carry?
• What are the five largest allowances, and what are they based on?
• What’s excluded — and can I see the list?
• Who benefits if this number is accepted?

That last question isn’t cynical, it’s structural. Every number in a development has an author, and every author has a context. A construction consultant with no stake in the project proceeding is the one party whose number can afford to be unpopular.

The uncomfortable advice

If the honest number doesn’t work, the time to find out is before you’ve bought the land, signed the leases, or told your board. A project that dies in feasibility costs you a consulting fee. A project that dies in construction costs you everything attached to it.

We’d rather give you a number you don’t like than a number that isn’t true. Owners are sometimes surprised how often the honest number, found early, doesn’t kill the project — it reshapes it into one that works. Smaller floorplate, phased delivery, a different structural system, a different quality tier in the right places. Those options exist in feasibility. They evaporate at tender.

Frequently asked questions

How accurate should a first budget be?

Expect a genuine early budget to carry a stated accuracy range — commonly ±15–20% at concept, tightening as design matures. Be suspicious of early numbers presented without a range; precision at concept stage is a costume.

Who should build the first budget?

Someone with current cost data and no stake in the answer. Contractors bring real pricing but also a business-development lens; consultants bring rigour but may defend the design. An independent build or review — even of someone else’s number — is cheap against what it protects.

How often should the budget be re-tested?

At every design milestone — concept, schematic, design development, and before tender. Budgets don’t fail suddenly; they drift. Milestone checks catch the drift while the drawings are still cheap to change.

If your project’s first number is being written right now — or was written by someone with a position — that’s exactly the moment to talk to us.