This is a practitioner guide to how construction budgets are developed, refined, and controlled on complex residential projects in Los Angeles. It covers the full progression from a rough feasibility range to a Guaranteed Maximum Price - including the methodology at each stage, the math behind budget variance, and the systems that keep a project on track during construction. Whether you're an owner evaluating a construction number, an architect working through design-phase cost decisions, or a financial advisor assessing a capital commitment, the goal is the same: after reading this, you should understand how construction money actually works.

Last updated: February 2026

Why Construction Budgets Are Wrong So Often

Budget overruns on custom residential projects are not occasional failures. They are the norm under the conventional process. Industry data consistently shows that 30-50% of custom residential projects exceed their original budget, and the overruns are not small - they typically range from 20% to 40% of the original number, sometimes more. On a $7M custom home in Los Angeles, that translates to $1.4M-$2.8M in additional cost the owner did not plan for and did not authorize.

The typical pattern plays out like this: the owner gets a number early in the process - maybe from the architect, maybe from a friend who recently built, maybe from a real estate agent who knows what homes sell for in the neighborhood. "This should cost about $5M to build." That number settles into the owner's financial planning. It informs the construction loan application. It sets the emotional expectation. The owner then proceeds through 12-18 months of architectural design - schematic design, design development, construction documents - investing $150,000-$300,000 in design fees along the way. When the completed drawings finally go out to bid, the number comes back at $6.5M-$7M. The 30-40% gap triggers a crisis. Redesign. Scope cuts. Value engineering sessions that feel more like triage than optimization. Strained relationships between owner, architect, and contractor. Sometimes project abandonment.

The conventional explanation for this is that contractors underbid to win work, then make it up in change orders. That happens. But it's not the structural problem. The structural problem is that budgets are developed with the wrong methodology at the wrong time by people who don't have the right data.

The Wrong Person Is Providing the Budget Number

On most custom residential projects, the first construction budget number comes from the architect. This makes intuitive sense - the architect is the first construction professional the owner hires, and the owner naturally asks, "What will this cost to build?" The architect provides a conceptual budget based on the best information available to them: cost per square foot from similar projects they've designed.

This is appropriate at the conceptual stage. It's the best information available from the architect's expertise, and it gives the owner a financial frame of reference before design begins. But it's a parametric estimate - a top-down calculation based on aggregate data - and its accuracy range is ±25% at best. That means a "$900 per square foot" estimate on a 6,000-square-foot home could legitimately range from $4.05M to $6.75M. The $2.7M spread is not error. It's the inherent precision of the methodology.

The problem is not that the architect provides this number. The problem is that the owner treats a ±25% parametric estimate as a ±5% commitment. The architect doesn't have the estimating infrastructure, current subcontractor pricing, or trade-specific cost data to provide anything more precise. They don't solicit competitive bids from framing contractors. They don't price caisson foundations against grade beam alternatives. They don't know the current market rate for structural steel fabrication and erection. That's not their job. Blaming the architect for a budget overrun caused by the absence of construction input is blaming the wrong person for the wrong problem. For a detailed discussion of what architects are and aren't responsible for regarding cost, see The Architect's Role in Complex Residential Construction.

The Budget Doesn't Evolve as Design Progresses

On a project without a construction manager engaged during design, the conceptual number from schematic design follows the project through Design Development and into Construction Documents - unchanged. Nobody is refining the estimate as design decisions add cost. The architect specifies a European steel window system; nobody quantifies the $180,000 cost difference from a domestic alternative. The structural engineer designs a caisson foundation with 40 piles to 30 feet; nobody prices the caissons against the budget allowance for a spread footing system. The civil engineer specifies 4,500 cubic yards of earth export; nobody calculates the $225,000 in hauling cost and the three weeks of truck traffic. The landscape architect designs a pool with an infinity edge cantilevered over a hillside; nobody prices the structural support against the flat-grade pool in the original budget.

Each of these design decisions is reasonable in isolation. The architect and engineers are solving design problems within their expertise. But every decision has a cost consequence, and without someone tracking those consequences against the budget in real time, they accumulate silently. By the time someone with construction pricing capability sees the project - during bidding - the cumulative cost of untracked design decisions has blown the budget. The gap isn't the result of one bad decision. It's the result of dozens of individually reasonable decisions that nobody was tracking financially.

Nobody Bids the Work Until It's Too Late

The most reliable construction cost data comes from competitive subcontractor bids on trade-specific scope packages. Under the conventional design-bid-build process, this doesn't happen until construction documents are complete and the project goes out to bid. At that point, 12-18 months of design work is committed. The architect has invested hundreds of hours developing the design. The owner has spent $150,000-$300,000 in design fees. The structural engineer has designed the foundation. The MEP engineers have laid out the mechanical systems. Everyone is emotionally and financially invested in the design as drawn.

Discovering a 35% budget gap at this stage is catastrophic. The options are all bad: redesign (which costs months and additional design fees), scope cuts (which means giving up features the owner wanted), value engineering (which may compromise design intent), or proceeding over budget (which strains or breaks the financial plan). Under a CMAR model, preliminary subcontractor pricing begins during Design Development and becomes competitive during Construction Documents - so the budget reflects real market pricing long before the owner faces a GMP decision.

The Estimate Methodology Doesn't Match the Stage

Different stages of design require different estimating methodologies, and each methodology has an inherent accuracy range. Using the wrong methodology at the wrong stage produces a number that is either inappropriately precise (creating false confidence) or inappropriately vague (failing to identify cost problems early enough to address them). A cost-per-square-foot estimate at the feasibility stage is the right tool. A cost-per-square-foot estimate at the Construction Documents stage - when detailed drawings exist and every system is specified - is negligent. The design has progressed to a level of specificity that demands trade-specific pricing, and relying on parametric data at that stage means the owner has a ±25% number when they could have a ±5% number.

Understanding which methodology belongs at which stage is fundamental to budget management. The next section walks through this progression in detail.

Pre-Feasibility vs. Feasibility - The First Budget Numbers

Most owners conflate pre-feasibility and feasibility - or skip both entirely and jump directly into design with no construction cost input at all. These are two different stages with two different purposes, and understanding the distinction prevents the two most expensive early-stage mistakes: investing in a feasibility study for a project that was never financially viable, and investing in architectural design for a project whose real cost won't be known until bidding.

Pre-Feasibility: "Is This Project in the Right Universe?"

Pre-feasibility is a screening exercise. Before any detailed analysis, the owner needs to know whether the project they're contemplating is in the financial neighborhood they can afford. This is a 1-2 hour exercise - not a formal engagement - that produces a rough order of magnitude range based on four inputs.

Parametric cost modeling - cost per square foot for the project type, adjusted for current market conditions. A new-construction hillside home in Bel Air is a different cost profile than a flat-lot remodel in Santa Monica. Current parametric ranges for complex residential work in Los Angeles vary widely by project type - see What Does Construction Cost in LA for detailed cost-per-square-foot data across project categories.

Comparable project data - what similar projects actually cost to build, not what they sold for. The sale price of a home includes land value, developer profit, and market premiums that have nothing to do with construction cost. A home that sells for $2,000 per square foot may have cost $800-$1,100 per square foot to build, depending on the land value and the seller's profit margin.

Site complexity factors - hillside premiums for sloped sites, fire zone premiums for PGRAZ and VHFHSZ areas, access constraint premiums for narrow streets or difficult terrain, and utility infrastructure factors for sites requiring transformer upgrades or main extensions.

Known cost drivers - PGRAZ requirements for fire rebuilds, utility upgrades identified from preliminary research, and foundation complexity indicators from available geology data or neighboring project experience.

Pre-feasibility produces a range - not a number. "A 5,000-square-foot hillside home in Bel Air with a full basement, based on current market conditions, is likely to cost between $5.5M and $8.5M in construction alone, depending on site conditions, foundation requirements, and finish level." That $3M spread is not imprecision. It's honesty. At this stage, the site hasn't been evaluated in depth, the geology isn't known beyond available published data, the design doesn't exist, and the regulatory requirements haven't been mapped. A range this wide is the appropriate level of confidence for the information that exists.

Pre-feasibility tells the owner whether to proceed to feasibility - whether the project is even worth the $15,000-$25,000 investment in a formal analysis. If the owner's all-in construction budget is $4M and the pre-feasibility range is $5.5M-$8.5M, the conversation changes before anyone spends serious money. Maybe the project concept needs to shrink. Maybe a different site makes the numbers work. Maybe the owner decides $7M is acceptable and proceeds with adjusted expectations. All of those are good outcomes - far better than discovering the gap after $200,000 in design fees.

Feasibility: "What Does This Project Actually Look Like Financially?"

The feasibility analysis is a formal engagement - typically 3-4 weeks of structured work - that evaluates the specific site, the specific regulatory requirements, and the specific project concept. The feasibility report is the document that converts a pre-feasibility range into an informed ROM (Rough Order of Magnitude) estimate. It produces four deliverables.

A ROM estimate with ±20-25% accuracy. This is more reliable than the pre-feasibility range because it's based on site-specific data rather than category-level assumptions. The estimate reflects actual soil conditions (or available geotechnical information from published data and neighboring projects), actual regulatory overlays identified from ZIMAS and agency research, actual utility infrastructure assessed through preliminary provider contact, actual access constraints observed during a detailed site visit, and actual grading implications evaluated from topographic survey data.

A risk register identifying factors that could move the number higher. These are the known unknowns - conditions that exist but haven't been fully quantified. Geotechnical conditions that require a full investigation to determine. Utility capacity that may require a transformer upgrade. Permitting timelines that could extend carry costs. Retaining wall requirements that depend on final building placement. Each risk is identified, its potential cost impact is estimated, and the action required to resolve it is specified.

A preliminary timeline. Not a construction schedule - the design doesn't exist yet - but a realistic assessment of the total project duration from design through completion, including the permitting timeline (which in Los Angeles often takes 6-12 months or longer for complex projects) and the carry cost implications of that duration.

A proceed/proceed with conditions/do not proceed recommendation. Not every project should proceed. Some sites have conditions that make them financially impractical. Some project concepts exceed the owner's budget by a margin that scope reductions can't close. The feasibility report gives the owner an informed basis for that decision before they've committed $150,000+ to architectural design.

The feasibility estimate narrows the pre-feasibility range from approximately $3M to approximately $1.5M-$2M. "This project is estimated at $6.2M-$7.8M, with the primary variables being foundation depth (dependent on the full geotechnical investigation) and the LADWP transformer timeline (which affects carry costs during construction)."

The Budget Evolution - What Happens at Each Design Stage

This is the core of how construction budgets develop. The progression from ±25% at feasibility to ±5% at construction documents to a fixed GMP doesn't happen automatically. It requires a construction professional doing specific estimating work at each design milestone - different methodology at each stage, matched to the design information that exists. Without this work, the budget number stays at ±25% regardless of how far the design progresses, and the owner doesn't know it.

Feasibility Stage (±20-25%)

Design information available: Program requirements (number of bedrooms, bathrooms, target square footage), a site identified and initially evaluated, conceptual massing only. No architectural drawings exist. The owner may have reference images or a written description of what they want, but nothing has been designed.

Estimating methodology: Parametric cost modeling - cost per square foot for the project type, adjusted for site complexity, finish level, and current market conditions. ROM pricing of major site work items (foundation system, grading and earthwork, retaining walls, utility connections) based on comparable project data and the preliminary site assessment from the feasibility analysis.

What drives the ±20-25% range: At this stage, the biggest unknowns are foundation cost (dependent on geotechnical conditions not yet fully investigated), site work scope (dependent on the building footprint not yet designed), structural system (steel frame vs. wood frame vs. concrete - a decision that can swing the structural budget by 30-50%), and finish level (dependent on selections not yet made, which can vary the interior budget by $100-$200 per square foot or more). The ±20-25% range reflects these unknowns. It's not imprecision - it's appropriate acknowledgment of what isn't yet known.

Deliverable: ROM budget with cost ranges by major category, identified cost drivers, and risk factors that could move the number. This is the financial framework that informs the architect's design brief - the document that tells the architect, "Design within this financial envelope."

Schematic Design Stage (±15%)

Design information available: Floor plans, building sections, and elevations at schematic level. The building footprint and massing are defined. The structural concept is identified (wood frame, steel frame, or concrete). The basic spatial organization is established - room sizes, ceiling heights, major circulation paths, exterior wall configurations.

Estimating methodology: Assembly-based estimating - the building is broken into major assemblies and each assembly is priced using unit costs appropriate to the assembly type. This is significantly more granular than parametric modeling but less granular than trade-specific pricing. The major assemblies typically include: foundation system, structural frame, exterior enclosure, roofing, MEP systems, and interior buildout (estimated per square foot by area type).

What narrows the range: The building footprint is now defined, so the foundation area is known rather than assumed. The structural concept is identified, so the structural cost can be estimated within a narrower band. The massing determines the exterior wall area, roof area, and enclosed volume - all of which are measurable quantities that drive cost. The range narrows from ±20-25% to approximately ±15% because measured quantities are replacing assumed quantities for the largest building systems.

What the CM does: Constructability review of the schematic design to identify cost-intensive elements before they're developed further in Design Development. If the architect has designed a 12-foot cantilever over a hillside, the CM identifies the structural cost premium before the structural engineer designs the support system. If the floor plan creates a complex roof geometry with multiple valleys and hips, the CM flags the roofing cost premium before the design is further developed.

The CM also compares the SD estimate to the feasibility budget to flag variances early. If the SD estimate exceeds the feasibility target, the conversation happens now - when design changes cost nothing but the architect's time - rather than later when they cost months of redesign and re-engineering. This is where the architect-CM collaboration produces the most financial value: the architect understands the design implications of changes, the CM understands the cost implications, and together they present options to the owner with both perspectives quantified.

Deliverable: SD budget organized by building system with unit cost assumptions, allowances for items not yet detailed (interior finishes, landscaping, specialty systems), and a reconciliation to the feasibility target showing where the estimate has moved and why.

Design Development Stage (±10%)

Design information available: Detailed floor plans with wall types and dimensions. Wall sections showing construction assemblies. MEP layouts showing duct routing, plumbing risers, and electrical panel locations. Structural drawings showing member sizes and connection details. Window and door schedules with sizes and types. Material specifications beginning to solidify. Fixture selections underway.

Estimating methodology: Quantity takeoffs for major systems - actual quantities of concrete, structural steel, framing lumber, roofing, drywall, insulation, etc., measured from the DD drawings. These are not assumptions or parametric estimates. They are measured quantities derived from the actual design. In addition, the CM solicits preliminary subcontractor pricing on the largest and most variable trade packages: structural steel, foundation and shoring, mechanical systems, plumbing, and electrical. At the DD stage, these aren't competitive bids - they're budget pricing from pre-qualified subcontractors who review the DD documents and provide ROM pricing based on the design.

What narrows the range: Real quantities replace estimated quantities across the major building systems. The foundation design is now specific enough to price with reasonable confidence. The structural steel tonnage is estimated from member sizes shown on the structural drawings. The MEP systems are defined enough for subcontractors to provide budget pricing. Window and door schedules allow the CM to price the specified systems rather than carrying allowances.

What the CM does: This is the most active estimating phase of the entire project. The CM is measuring quantities from the drawings, soliciting preliminary pricing from key subcontractors, refining allowances as selections are made, and tracking the budget against the target in real time. If the budget is trending over the target, the CM presents value engineering options to the architect with full cost-impact analysis.

These aren't generic "cut costs" suggestions. They're specific, quantified alternatives: "Option A saves $140,000 by substituting domestic steel windows for the specified European system; here's the design impact and the product comparison. Option B saves $85,000 by switching from a standing-seam zinc roof to standing-seam steel with a zinc-finish coating; the visual difference is minimal and the warranty is comparable." The architect evaluates the design impact, the owner evaluates the financial impact, and the team makes an informed decision together.

Deliverable: DD budget with line-item detail by trade, reconciliation to the SD budget showing where and why costs have moved, VE options with cost and design impact analysis, updated allowances for items not yet specified, and an updated project timeline reflecting current design decisions.

Construction Documents Stage (±5%)

Design information available: Complete construction drawings and specifications. All systems are detailed - structural, mechanical, electrical, plumbing, fire protection. All materials are specified. Window, door, hardware, plumbing fixture, and electrical fixture schedules are complete. The permit set is ready for submission.

Estimating methodology: Competitive subcontractor bidding on trade-specific scope packages developed by the CM from the construction documents. This is where real numbers replace estimates. A typical custom residential project has 20-30 trade packages: demolition, earthwork and grading, shoring and underpinning, structural concrete, structural steel, rough framing, roofing, waterproofing, windows and doors, exterior cladding, lath and plaster, drywall, painting, finish carpentry and millwork, cabinetry, tile and stone, flooring, plumbing, HVAC, electrical, fire sprinklers, elevators, pool and spa, landscaping and hardscape, low-voltage systems, and general labor.

For each trade package, the CM develops a scope document that describes exactly what's included: the work shown on specific drawing sheets, the specifications that apply, the inclusions and exclusions, the schedule requirements, the insurance and bonding requirements, and the coordination responsibilities. This scope document ensures that every subcontractor bidding on a trade package is pricing the same work - an essential requirement for meaningful bid comparison. The CM solicits a minimum of three competitive bids per trade from pre-qualified subcontractors. The bids are then leveled (see the next section) to ensure apples-to-apples comparison.

What narrows the range: Everything. The design is complete. The quantities are precise. The subcontractor pricing is competitive and current. The construction schedule is defined, so duration-dependent costs are calculated rather than estimated. The only remaining uncertainty is conditions that can't be determined until construction begins - concealed conditions, subsurface conditions, weather impacts - and these are addressed through contingency reserves with specific protocols.

Deliverable: CD budget ready for GMP conversion, with all trade packages competitively bid and leveled, subcontractor recommendations for each trade, a complete schedule of values, and a draft GMP breakdown.

GMP Establishment (Fixed)

The Guaranteed Maximum Price is assembled from the competitive CD-stage pricing, plus the components described in detail in the GMP Structure section below. The GMP is a contractual commitment - the owner's cost ceiling. Cost overruns beyond the GMP are the CM's responsibility, not the owner's. Cost savings below the GMP are shared between owner and CM per the contract terms.

The GMP is not a number the CM picks. It's the mathematical product of months of progressive estimating work, validated by competitive market pricing from 60-90 subcontractor bids across 20-30 trade packages. By the time the GMP is presented to the owner, they have been tracking the budget at every design milestone. There are no surprises, because the GMP confirms what the progressive estimating work has already shown.

Bid Leveling - Why the Lowest Bid Isn't the Lowest Cost

Bid leveling is one of the most valuable and least understood aspects of pre-construction cost management. The concept is straightforward: when multiple subcontractors bid on the same trade package, their bids are almost never directly comparable. Each bidder interprets the scope slightly differently, includes or excludes different items, carries different assumptions about coordination and logistics, and may or may not include items that the owner will need regardless of who performs the work. Bid leveling is the process of normalizing those differences so the owner is comparing equivalent scopes, not just raw numbers.

Why Bids Are Never Apples-to-Apples

Even when every bidder receives the identical bid package - the same drawings, the same specifications, the same scope document - their bids will differ in ways that go beyond price. One structural steel fabricator includes erection in their bid; another provides material and fabrication only, excluding field installation. One plumber includes fixture installation; another provides rough-in only and excludes finish trim. One electrical contractor includes the main switchgear; another assumes it's in the owner's direct-purchase budget.

These differences aren't mistakes. They reflect each subcontractor's business model, their reading of the scope document, and their assumptions about what other trades will handle. A framing contractor who regularly works on projects where the general contractor provides crane service will exclude crane costs from their bid. A framing contractor who typically self-performs crane operations will include them. Both bids are legitimate. But comparing them without adjusting for the crane cost would penalize the contractor who included it and reward the one who excluded it - until the crane cost shows up as an unbudgeted expense during construction.

Example 1: Steel Doors and Windows

A project specifies a European steel window and door system throughout the home - 42 units total, including fixed windows, operable casements, French doors, and a large pivot entry door. Three qualified installers submit bids:

At first glance, Bidder B appears to be the clear winner - $125,000 less than Bidder A and $125,000 less than Bidder C. But the leveling analysis reveals the following:

Bidder B exclusions that must be covered: Flashing and waterproofing at 42 openings ($38,000) and second-trip final adjustment ($16,000). Steel windows require adjustment after the surrounding construction is complete; this is not optional. Total adjustment to Bidder B: $595,000 + $54,000 = $649,000.

Bidder A exclusion: Hoisting above first floor ($22,000). This project has windows on three levels; hoisting is required for approximately 60% of the units. Total adjustment to Bidder A: $680,000 + $22,000 = $702,000.

Bidder C: Full scope included. No adjustments needed. Leveled price: $720,000.

Bidder B is still the recommended selection after leveling, but the real gap between B and A is $53,000, not $85,000. And if the owner had simply accepted the lowest raw bid without leveling, the $54,000 in excluded scope would have appeared as change orders during construction - charges that feel like overruns but are actually costs that were always part of the project, just not captured in the original bid.

Example 2: Foundation and Shoring

A hillside project requires a caisson foundation with grade beams, plus temporary shoring to protect the adjacent downhill property during excavation. Three foundation contractors bid the work:

Bidder D adjustments: Dewatering (geotech report identifies a perched water table at 28 feet, required for at least 12 of the 18 caissons): $65,000. Spoils export (approximately 400 cubic yards at current trucking rates): $32,000. Adjusted total: $1,180,000 + $97,000 = $1,277,000.

Bidder E adjustments: The additional depth assumption (40 feet vs. 35 feet) adds approximately $120,000. The geotech report indicates competent bearing material at 32-38 feet. The CM's assessment: 35-foot depth assumption is reasonable for budgeting; Bidder E is conservative. Adjusted for equivalent depth: $1,420,000 - $120,000 = $1,300,000.

Bidder F adjustments: Shoring (required by the grading permit and essential for hillside neighbor protection): $185,000. Dewatering: $65,000. Excavation spoils export (additional 600 cubic yards): $48,000. Adjusted total: $1,280,000 + $298,000 = $1,578,000.

The lowest raw bid (Bidder D) remains the lowest leveled bid, but the apparent $240,000 gap between D and E shrinks to $23,000 after leveling - a difference small enough that Bidder E's more conservative depth assumption might actually represent better value. And Bidder F, who appeared to be $100,000 less than Bidder E at raw pricing, is actually $278,000 more expensive when the excluded scope is accounted for.

Example 3: Mechanical (HVAC)

A 7,500-square-foot home with a complex multi-zone HVAC system including a dedicated wine room cooling system and a pool house split system. Three mechanical contractors bid:

Bidder G adjustments: Controls integration with home automation system: $28,000. Duct insulation in unconditioned spaces (required by Title 24): $18,000. Smart thermostat upgrade per specification: $8,500. Startup and commissioning: $6,000. Adjusted total: $385,000 + $60,500 = $445,500.

Bidder H adjustments: Pool house HVAC system (excluded; work is in the scope): $42,000. Startup and commissioning: $6,000. Adjusted total: $448,000 + $48,000 = $496,000.

Bidder I adjustments: Controls integration: $28,000. Smart thermostat upgrade: $8,500. Adjusted total: $410,000 + $36,500 = $446,500.

The $63,000 apparent gap between Bidder G and Bidder H at raw pricing shrinks to $50,500 after leveling. More importantly, Bidder I - who appeared to be $25,000 more expensive than Bidder G - is functionally equivalent in price after leveling and includes startup and commissioning, which is a meaningful quality differentiator for complex HVAC systems.

GMP Structure - How the Guaranteed Maximum Price Is Assembled

The Guaranteed Maximum Price is not a single number. It's an assembly of components, each of which is visible to the owner, each of which has a defined purpose, and each of which is tracked independently during construction. Understanding the components is essential to evaluating whether a GMP is well-constructed.

GMP Component Breakdown

On a representative $8M GMP for a complex hillside home, the assembly might look like this:

Trade Costs

The sum of all leveled subcontractor bids - the actual cost of the construction work. Concrete, steel, framing, roofing, waterproofing, MEP, drywall, finishes, landscaping, and every other trade required to build the project. On a typical custom residential project, trade costs represent 65-72% of the GMP. Every trade cost in the GMP traces back to a specific leveled bid from a specific subcontractor. The owner can see the bid, the leveling analysis, and the scope that's included. There are no lump-sum allowances hiding unpriced scope. If a scope item isn't priced, it's identified as an allowance with a stated basis for the allowance amount.

General Conditions

The cost of running the project - the infrastructure and management resources required to coordinate, supervise, and support the construction work. General conditions include the on-site superintendent (full-time for the project duration), project management support, temporary facilities (construction fencing, temporary power distribution, sanitation facilities, material storage, weather protection), safety equipment and compliance, project controls (scheduling software, document management systems, progress photography), builder's risk insurance, and general equipment.

On hillside projects with complex logistics - limited access, crane requirements, material staging constraints, extended durations - general conditions run higher (8-12% of trade cost) than flat-lot projects (6-8%) because the logistics complexity requires more management resources and the construction duration is typically longer. General conditions are presented as a detailed line-item budget, not a lump-sum number. The owner can see exactly what's included. This transparency is a defining feature of the CMAR delivery model.

CM Fee

The construction manager's compensation for managing the project, carrying the GMP risk, providing the pre-construction services that developed the budget, and coordinating the entire construction operation. The CM fee is a visible, negotiated line item in the GMP. It is not hidden in subcontractor markups, not buried in overhead charges, and not added to material costs.

The typical range for CM fee on complex residential projects in Los Angeles is 8-14% of cost of work (trade costs plus general conditions). The variation reflects project complexity, project duration, risk profile (hillside projects with geotechnical risk carry higher risk premiums than flat-lot projects), and the scope of pre-construction services provided. For a comparison of total management costs under CMAR vs. cost-plus with an owner's representative, see CM at Risk vs. Owner's Representative.

Construction Contingency

A defined reserve for unforeseen conditions that arise during construction - conditions that were not visible in the construction documents and could not reasonably have been anticipated during design. Examples include subsurface conditions not captured by geotechnical investigation, concealed conditions in existing structures, and code-compliance issues discovered during construction that require design modifications.

Construction contingency is not a slush fund. It is not used for scope additions, owner upgrades, or design changes. It has a specific purpose (unforeseen conditions) and specific access protocols: documented description of the unforeseen condition, photographic and/or third-party documentation, subcontractor pricing with backup, and CM/owner joint approval before the work proceeds. Typical construction contingency ranges from 3-7% of cost of work, depending on risk profile. New construction on a well-investigated site: 3-5%. Hillside construction with complex geology: 5-7%. Renovation with unknown concealed conditions: 7-10%.

Owner Contingency

A separate reserve, entirely owner-controlled, for owner-directed changes during construction. On an 18-24 month custom residential project, the owner will make changes. This is not a problem to be prevented - it's a reality to be managed. The kitchen island gets longer after the owner visits a friend's house. The master bathroom adds radiant floor heating after the owner stays at a hotel with heated floors. The landscape plan evolves after the owner decides they want a fire pit they hadn't originally planned.

Owner contingency acknowledges this reality rather than pretending changes won't happen. It provides a defined reserve so the owner can make changes without exceeding the GMP. Typical range: 5-10% of cost of work.

How the GMP Protects the Owner

If total project costs exceed the GMP, the CM absorbs the overage. The owner pays no more than the GMP amount (adjusted for any owner-approved change orders that formally amend the GMP). This is the contractual risk transfer that makes the GMP meaningful - the CM has a financial incentive to manage costs because overruns come out of their margin.

If total project costs come in below the GMP, the savings are shared between the owner and the CM per the contract terms. Typical sharing ratios range from 50/50 to 75/25 owner-favorable. This creates a second incentive alignment: the CM is motivated to find efficiencies and negotiate favorable pricing because they participate in the savings. The CMAR delivery model aligns the CM's financial interest with the owner's financial interest at both ends of the cost spectrum.

Cost Control During Construction - The Systems That Keep It on Track

The GMP sets the ceiling. The cost control systems keep the project under it. These are not aspirations or best practices - they are operating procedures that run on every project, every month, from groundbreaking to final closeout.

Open-Book Accounting

Every cost on the project passes through at actual cost. The owner sees every subcontractor invoice, every material purchase order, every equipment rental, every general conditions charge. Nothing is marked up. The CM's compensation is the fee - a defined line item in the GMP - not a margin on subcontractor costs. If the framing contractor invoices $48,000 for the month's work, the owner sees a $48,000 charge. Not $48,000 plus 15%.

This is fundamentally different from a cost-plus arrangement where the contractor marks up every cost that passes through the project (typically 15-20%), or a lump-sum contract where the owner never sees the actual costs at all. Under open-book CMAR, the owner has complete financial visibility into what the project costs and where the money goes.

Earned Value Measurement

Traditional cost tracking answers one question: "How much have we spent?" Earned value measurement (EVM) answers a more useful question: "How much work have we completed relative to how much we've spent?"

The distinction matters. A project that has spent 45% of the budget might be in excellent shape (if 50% of the work is complete) or in serious trouble (if only 35% of the work is complete). Simple budget-to-actual comparison can't tell you which scenario you're in. EVM can.

EVM tracks three metrics: the planned value (the budgeted cost of the work scheduled to be complete by now), the earned value (the budgeted cost of the work actually completed), and the actual cost (what has actually been spent on the work completed). From these three numbers, two critical ratios emerge.

The Cost Performance Index (CPI) divides earned value by actual cost. A CPI of 1.0 means the project is on budget. Above 1.0 means under budget. Below 1.0 means over budget. A CPI of 0.92, for example, means the project is getting 92 cents of value for every dollar spent - an 8% cost overrun rate that, if it continues, would produce an 8% overrun at completion.

The Schedule Performance Index (SPI) divides earned value by planned value. A SPI of 1.0 means on schedule. Above 1.0 means ahead. Below 1.0 means behind. The SPI provides early warning when the project is falling behind schedule - which matters for cost because schedule delays drive general conditions overruns, carry cost increases, and potential escalation exposure.

The power of EVM is early detection. A cost problem that's identified at 20% project completion can be corrected. The same problem identified at 80% completion can only be absorbed.

Monthly Cost Reporting

The owner receives a detailed monthly report - typically 12-20 pages - that provides a comprehensive financial picture of the project. The report includes budget vs. actual for every line item in the GMP, showing what was budgeted, what has been billed to date, what is committed but not yet billed, and the projected cost at completion for each line.

The report tracks committed cost vs. total GMP - showing how much of the GMP is locked up in executed subcontracts, approved change orders, and general conditions commitments. Contingency tracking is reported in detail: the starting balance, each draw with a description and justification, the current balance, and the projected sufficiency of the remaining contingency. The change order log shows every change from initiation through pricing through approval or rejection. Cash flow projections for the coming 90 days allow the owner to manage construction loan draws. Variance analysis flags any line item trending over budget with an explanation and a corrective action plan.

This is the deliverables framework in practice - structured, documented financial reporting that keeps the owner informed and enables proactive decision-making.

Contingency Management

Both construction contingency and owner contingency are tracked as separate reserves with separate approval protocols and separate reporting.

For construction contingency, every draw follows a documented process. The unforeseen condition is identified and documented with photographs, inspection reports, or geotechnical findings. The CM describes the condition, explains why it could not reasonably have been anticipated, and presents the scope of work required to address it. Subcontractor pricing is obtained with full backup. The CM and owner jointly review and approve the draw before the work proceeds.

This discipline serves two purposes. First, it ensures that contingency is used only for its intended purpose - genuinely unforeseen conditions - and not for scope additions, optimization decisions, or cost overruns on budgeted work. Second, it creates a documented record of every contingency draw, which is essential for the savings-sharing calculation at project closeout.

Formal Change Control

All scope modifications - whether owner-initiated, architect-directed, or contractor-identified - follow a documented change control process. This process prevents the "surprise at final billing" that plagues projects without formal change management.

Step 1: Potential Change Notice (PCN). Within 24 hours of identifying a potential change, the CM issues a PCN to the owner and architect. The PCN describes the condition or request, identifies the likely cost and schedule impact (at a preliminary level), and flags it for formal pricing.

Step 2: Impact analysis. The CM develops a full impact analysis: scope definition (exactly what work is required), cost impact (subcontractor pricing with backup), schedule impact (will this change affect the completion date?), and any design coordination required.

Step 3: Change Order Proposal (COP). The CM presents a formal COP with complete pricing, schedule analysis, and a recommendation. If the change is a correction of a document error, it's priced at cost with no additional fee. If it's an owner-directed addition, it includes appropriate markups per the contract.

Step 4: Owner authorization. The owner reviews the COP and either approves, rejects, or requests modification. No work proceeds until written authorization is received.

Step 5: Execution and GMP amendment. Approved changes are executed through formal change orders that amend the GMP. The GMP increases (or decreases, for scope reductions) by the approved amount, maintaining the contractual cost ceiling at the updated scope.

This process means the owner is never surprised by a cost they didn't authorize. Every dollar above the original GMP has a corresponding change order with the owner's signature on it.

Why Budget Accuracy Requires Someone Who Can Bid the Work

The progression from a ±25% range to a ±5% estimate to a fixed GMP requires specific capabilities at each stage. Understanding what those capabilities are - and who has them - explains why the construction manager is the only team member who can deliver budget accuracy.

Parametric modeling at the feasibility stage requires access to current comparable project cost data - not sale prices, not published indices, but actual construction costs from completed projects of similar type and complexity in the same market. A construction manager who has completed dozens of complex residential projects in Los Angeles has this data from direct experience.

Assembly-based estimating at schematic design requires knowledge of building systems and current unit costs for each assembly type. What does a caisson foundation cost per lineal foot in the current market? What is the current cost per pound for structural steel fabrication and erection? These unit costs change with market conditions and staying current requires active engagement with the trade contractor market.

Quantity takeoffs at design development require the ability to read and measure construction drawings - to calculate the volume of concrete in a foundation, the tonnage of steel in a structural frame, the square footage of drywall on every wall and ceiling. This is trained estimating work that requires understanding of construction drawings and building systems.

Subcontractor bidding at the CD stage requires active trade relationships - the ability to call on 3-5 qualified subcontractors for each of 20-30 trade packages and receive competitive pricing within the bidding window. These relationships take years to build. A construction manager who has worked with a structural steel fabricator on five previous projects gets a responsive, accurately scoped bid. A first-time caller gets a budget number, if they get a response at all.

Bid leveling requires the construction knowledge to identify what each subcontractor has excluded, what's required by code, and what coordination is necessary between trades. Knowing that a window installer's exclusion of flashing means the waterproofing contractor needs to add scope requires understanding of how the building envelope is assembled.

No other professional on the project team has all of these capabilities. The architect has design knowledge and may have general cost awareness, but does not have estimating infrastructure, current subcontractor pricing, or the trade relationships required for competitive bidding. The owner's representative may have project management experience, but typically does not have the trade relationships, estimating capability, or the technical depth to develop scope packages and level bids. This isn't a criticism of any of these professionals. It's a description of scope boundaries.

The Real Cost Impact - What Budget Variance Means on an $8M Project

Budget overruns on construction projects are not linear. A 20% overrun does not cost 20% more. It costs significantly more, because the overrun triggers cascading secondary costs that compound the original gap. Understanding the true cost of budget variance - not just the direct overrun but the full financial impact - is essential for evaluating the investment in getting the budget right.

Scenario 1: 10% Overrun ($800,000 Direct)

The project was budgeted at $8M. At 60% completion, the CM or contractor identifies that the actual cost trajectory will exceed the budget by approximately 10%. Even a managed 10% overrun has secondary costs:

• Schedule delay from redesign and re-procurement: 2-3 months. Cost: $80,000-$120,000 in construction loan interest ($8M loan at 7-8% annual rate)
• Architect redesign fees: $40,000-$80,000 depending on scope of modifications
• Re-permitting costs: $20,000-$40,000 if changes require revised plan check
• Subcontractor re-mobilization and delay claims: $30,000-$60,000 for trades on hold

Total impact of a "10% overrun": approximately $970,000-$1,100,000 - effectively 12-14% of the original budget.

Scenario 2: 20% Overrun ($1,600,000 Direct)

This is the scenario that breaks projects. A 20% overrun is typically discovered later - at bidding or early construction - when the owner's options are more constrained and the redesign effort is more extensive.

• Schedule delay: 4-6 months for significant redesign, re-engineering, and re-bidding. Interest cost: $190,000-$280,000
• Architect redesign fees: $100,000-$150,000 for substantial design modifications
• Structural and MEP re-engineering: $40,000-$80,000 if the redesign affects structural or mechanical systems
• Re-permitting: $40,000-$60,000 for revised plan check and potentially new discretionary approvals
• Subcontractor impacts: $60,000-$120,000 in re-mobilization, delay claims, and re-bidding costs
• Emotional and relationship cost: Not quantifiable, but real. A 20% overrun typically damages the owner-architect relationship and introduces stress that affects decision-making for the remainder of the project

Total impact of a "20% overrun": approximately $2,030,000-$2,290,000 - effectively 25-29% of the original budget.

Scenario 3: 30% Overrun ($2,400,000 Direct)

This is the crisis scenario. A 30% overrun on an $8M project means the owner needs $10.4M in construction funding when they planned for $8M. The construction loan may not cover it. The owner's liquidity may not absorb it. Some owners abandon the project at this point, losing their entire design investment - potentially $200,000-$400,000 in architectural and engineering fees.

• Schedule delay: 6-9 months or more for major redesign. Interest cost: $280,000-$420,000
• Architect and engineer redesign fees: $150,000-$250,000
• Re-permitting with potential discretionary review: $60,000-$100,000
• Subcontractor impacts and potential litigation exposure: $100,000-$200,000
• Construction loan restructuring costs: $40,000-$80,000 in origination fees for increased loan amount

Total impact of a "30% overrun": approximately $3,030,000-$3,450,000 - effectively 38-43% of the original budget.

The Alternative: The Cost of Getting It Right

Now compare those overrun costs to the cost of progressive budget development. A construction manager engaged from feasibility through GMP development provides estimating, bid management, and design-phase collaboration at each stage. The pre-construction investment is typically structured as a pre-construction fee that is credited toward the construction fee if the owner proceeds to construction.

On an $8M project, the pre-construction phase represents an investment of $80,000-$150,000 in CM time (much of which is credited to the construction phase). Compare that to the $1M-$3.4M total impact of a 10-30% budget overrun. The return on investment in proper budget development is not 2x or 5x. It's 10x to 25x.

How This Connects to the Rest of the Process

Budget development doesn't exist in isolation. It's one component of a broader process that manages cost, schedule, quality, and risk across the full project lifecycle. Here's how the budget connects to each element of that process.

The feasibility analysis produces the first informed budget number - the ROM estimate that establishes the financial framework before design begins. Without a feasibility analysis, the owner enters design with no construction cost input, and the first real budget number arrives 12-18 months later at bidding.

Pre-construction collaboration with the architect keeps design aligned with budget through schematic design, design development, and construction documents. The architect designs. The CM prices. Together, they present options to the owner with both design and cost impact quantified. This collaboration prevents the scenario where design decisions accumulate untracked costs that surface as a budget gap at bidding.

The GMP structure within the CMAR delivery model converts the budget into a contractual commitment. The progression from feasibility estimate to GMP is the progression from data point to guarantee.

The incentive alignment of CMAR vs. other delivery models determines whether the budget is an honest number or an adversarial starting point. Under cost-plus, the contractor has no incentive to control costs because their fee is a percentage of spending. Under lump-sum, the contractor's incentive is to minimize what they deliver. Under CMAR, the GMP aligns interests: the CM absorbs overruns and shares savings.

Cost per square foot data provides the parametric benchmarks used at the feasibility stage. Understanding what drives cost per square foot variation (hillside vs. flat lot, foundation complexity, finish level) is essential for evaluating early-stage budget numbers.

The project timeline is inseparable from the budget. Every month the project takes to build costs the owner money in general conditions, construction loan interest, rental or housing costs, and exposure to material price escalation. A 24-month project carries twice the time-dependent costs of a 12-month project. Schedule efficiency is cost control.

Frequently Asked Questions

The information on this page is based on 24 years of construction management experience on complex residential projects in Los Angeles. Cost ranges, percentages, and timelines reflect current market conditions for the greater Los Angeles area and may vary based on project-specific conditions, site complexity, regulatory requirements, and market fluctuations. This content is educational and does not constitute a cost estimate, financial guarantee, or professional advice for any specific project. Consult qualified professionals for project-specific guidance.