Why construction budgets fail before the first bid arrives

An owner sends a set of documents to bid with a budget of twelve million dollars. Bids come back at fifteen. The instinct is to conclude that the market is hot, the bidders are opportunistic, or the design is gold plated. Occasionally one of those is true. Usually the budget was wrong from the beginning, in ways that were visible months earlier.

Six failures account for most of the gap.

Cost per square foot is a useful sanity check and a terrible budget. On Southern California projects, particularly hillside and infill sites, earthwork is frequently the largest single line item and the one with the widest range. Remedial grading depth, rock, groundwater, import and export volume, and haul distance can swing a site development budget by millions on a project where the building cost is essentially fixed.

A benchmark derived from completed projects averages all of that away. Two projects with identical building programs on different sites can differ by thirty percent in total cost, entirely because of what is underground.

What to do instead. Read the geotechnical report as a cost document. Quantify remedial grading, import and export, and haul separately, with a stated range, before adopting any total.

Every set of drawings implies work that it does not show. Temporary shoring. Dewatering. Traffic control staging. Utility relocations that follow from the alignment. Erosion control that satisfies the permit. Testing and special inspection. Protection of adjacent property. None of these have a symbol on a sheet, and all of them cost money.

A budget that prices only what is drawn will be low by exactly the amount of what is implied, and a bidder will price it correctly.

What to do instead. Estimate from a scope narrative, not from a drawing set. Write down what the project requires, then confirm each item is either drawn or priced.

Conditions of approval are construction obligations. Widen the frontage, underground the utility, construct the drainage facility, provide the traffic mitigation. They are agreed to during entitlement, frequently by people whose job is to secure the approval rather than to price it, and they arrive later as a construction cost with a timing trigger.

The same is true of permit conditions, water quality requirements, and agency standard details that differ from what the design assumed.

What to do instead. Price every condition of approval as it is drafted, while it can still be negotiated. See land development and entitlement.

Ten percent is the most common construction contingency in use and one of the least defensible. Appropriate contingency is a function of how well the scope is defined, not of tradition. A conceptual budget on an unentitled site with no geotechnical work may warrant twenty five percent or more. A bid quality estimate against complete construction documents in a known jurisdiction may warrant five.

Applying the same number to both means one of them is badly wrong.

What to do instead. Size contingency to the actual level of design definition and state in writing what it is covering.

A budget prepared today for a project that bids in fourteen months and builds over twenty four is not a current cost estimate. It is a forecast, and it needs to carry escalation across the actual construction window, weighted to when the spend occurs.

Owners routinely escalate to the bid date and stop, which understates the cost of everything purchased after month one of construction.

What to do instead. Escalate to the midpoint of construction, not to bid, and state the assumed rate so it can be revisited.

A single estimate produced at schematic design cannot track a design that develops for another year. Scope creeps in small increments that each seem immaterial, and nobody reconciles the total until bid day.

What to do instead. Estimate at three milestones: schematic design, design development, and bid documents, with a written reconciliation between each showing exactly what changed and why. This is the single most effective cost control practice available to an owner, and it is also the least used.

Each failure is a form of the same thing: a number produced without stating what it depends on. An estimate that documents its scope, its exclusions, its assumptions, and its basis can be checked, challenged, and updated. An estimate that is just a total cannot.

That is why every estimate we produce is delivered with the assumptions attached. See cost estimating and preconstruction services.