Chapter 12 — Estimating Fundamentals: Quantity Takeoffs, Unit Costs, and Building a Bid from Plans and Specs

The Hook: A Red Pencil and 148 Footings

Tomás Reyes, Kestrel's chief estimator, has a laptop with two monitors, a license for every takeoff package on the market, and a model of the Northgate Outpatient Pavilion sitting on the network. On the Tuesday of Dani Okonkwo's third week, he ignored all of it.

He walked to the plot table, rolled out a printed E-size sheet — S-101, Foundation Plan — and set three things on top of it: an architect's scale, a paper takeoff sheet with green columns, and a red pencil.

"Footings," he said. "All of them. Volume in cubic yards. Take your time."

Dani looked at the drawing, then at the pencil, then back at Tomás. "We have Bluebeam."

"We do. Digital takeoff is faster. This is how you learn to see." He tapped the sheet twice with one finger. "Come find me when you have a number."

It took Dani four hours and forty minutes. They did it properly — or thought they did. They found the footing schedule on S-501, built a table by mark, computed the volume of each type, counted the marks on the plan with a red tick beside each one so nothing got counted twice, and extended the whole thing. They caught the column pedestals sitting on top of the pads, which is the trap everybody warns you about, and priced those separately. They double-checked the count. Then they walked into Tomás's office and put the sheet on his desk.

1,129 cubic yards.

Tomás read it for maybe fifteen seconds. He did not reach for a calculator.

"The number in our estimate is 1,240," he said. "You're nine percent low."

Dani's face did the thing that everyone's face does the first time. Tomás waved it off before it could get anywhere.

"Nine percent is a good first takeoff. I was twenty-two percent low on my first one and I was missing an entire wing. Sit down. There are exactly three things wrong with this sheet and every one of them is a thing you will do again unless we name it now."

He pulled the drawing over and put his own pencil on it.

"One. You took off what was on this sheet. There are nine footings that are not on this sheet. Six under the entry canopy and three combined footings at the loading dock, and they live on S-104, the enlarged site-entry plan, because the canopy is a separate structural package. Your count is 139. The schedule says 148. The schedule told you your count was wrong and you didn't listen to it.

"Two." He slid the schedule across. "F-6. You carried three feet six. It's four feet. You read across from the F-5 row. Six footings, fourteen by fourteen, six extra inches of concrete each. That is a line-skip on a schedule and it is the single most common arithmetic mistake in this entire office. Every estimator in here has done it. The fix is a straightedge, and I am not joking.

"Three, and this is the one I actually care about." He tapped the bottom of Dani's sheet, where the total sat naked with nothing under it. "There is no waste line. You gave me a geometry number. Nobody pours geometry. You pour into a hole in the ground that a track hoe dug, and the hole is bigger than the drawing, and the last truck goes back with three-quarters of a yard in the drum. What you handed me is the volume of the footings. What I need is the volume of the concrete I have to buy."

He wrote three numbers in the margin in red.

   missed footings + their pedestals ...........  65.8 CY
   F-6 thickness (line-skip) ...................  21.8 CY
   waste / over-excavation allowance ...........  23.6 CY
                                                 --------
                                                 111.2 CY

"Call it a hundred and eleven yards. Loaded — concrete, forms, rebar, the extra dig — that is a shade under forty-three thousand dollars." He handed the sheet back. "At the fee we carry on this job, Kestrel has to build one and a third million dollars of work to earn forty-three thousand dollars. You just gave it away on one line item, on one sheet, in one trade, on a job with about three hundred more sheets in the set."

He picked up the red pencil and held it out.

"Again. And this time write down every assumption you make, next to the number it belongs to."

That is the whole chapter, really. Everything that follows is the long version of those three misses and the sentence at the end.

🏃 Fast Track: If you have run takeoffs before, skim §12.3 and §12.4 for the Northgate worked quantities and the unit-conversion traps, then go straight to §12.7 (building a unit cost from a burdened labor rate) and §12.9 (markup versus margin — check yourself on that one even if you think you know it). Do the 📋 Try it drill; it is the chapter in miniature.

🔬 Deep Dive: Appendix A has the conversion tables and geometry formulas. Appendix C has the CSI division checklist, representative productivity ranges, and waste factors in one place. Then Chapter 13 assembles everything here into a complete priced bid, and Chapter 20 goes deep on where productivity rates actually come from.

Why this chapter matters, in one sentence: the takeoff and the unit cost are the two places where a construction company's entire year is decided, because every dollar in the estimate — every markup, every contingency, every fee — is a percentage of a number that somebody measured off a drawing.

12.1 The Anatomy of an Estimate

Before you measure anything, you need to know what you are building toward. An estimate is not a list of quantities. It is a structure, and the structure has three layers.

Direct costs are the costs of the physical work: the labor, material, and equipment that go into the building, plus the subcontracts that cover the scopes you are not self-performing. If you could walk the finished building and point at it, it came from a direct cost.

Indirect costs are the costs of running the job that do not become part of the building. Your project manager's salary. The trailer. The dumpsters. The temporary power bill. The safety supplies. The final cleaning. Nobody points at a dumpster in the finished lobby, but the lobby does not happen without it. These split into two families that people mix up constantly:

• General requirements are the Division 01 scope items the specifications require — temporary facilities and controls, testing and inspection, submittals, project cleanup, waste management, mockups, closeout documentation. They are specified, which means they are contractual.
• General conditions are your project team and its infrastructure — the PM, the superintendent, the field engineer, the project vehicles, the trailer and its utilities, the IT. They are yours, and they cost money for every day the job is open. On Northgate the agreed extended-general- conditions rate is $5,150 per calendar day, which is the number that turns every schedule conversation into a cost conversation.

Markups sit on top: escalation, insurance, bond, contingency, overhead, profit or fee, and taxes where they apply to the contract rather than to the purchase.

Here is the whole structure as a waterfall. Read it from the bottom up, because that is the order in which you build it.

                          ┌───────────────────────────────────┐
                          │        BID  /  GMP  /  PRICE      │
                          └───────────────────────────────────┘
                    ┌─────────────────────────────────────────────┐
                    │  Payment & performance bond premium         │
                    ├─────────────────────────────────────────────┤
                    │  Overhead + profit  (or CM fee)             │
                    ├─────────────────────────────────────────────┤
                    │  Contingency  (owned, drawn down, tracked)  │
                    ├─────────────────────────────────────────────┤
                    │  Escalation  (time between bid and buy)     │
                    ├─────────────────────────────────────────────┤
                    │  Insurance: GL, builder's risk, excess      │
              ┌─────┴─────────────────────────────────────────────┴─────┐
              │  INDIRECT COST                                          │
              │    General conditions — staff, trailer, vehicles, IT    │
              │    General requirements — Div 01, temps, cleanup, tests │
        ┌─────┴─────────────────────────────────────────────────────────┴─────┐
        │  DIRECT COST OF WORK                                                │
        │    Labor (fully burdened)   Material (+ waste, freight, tax)        │
        │    Construction equipment   Subcontracts   Allowances               │
        └─────────────────────────────────────────────────────────────────────┘
                    ▲
                    │  Every dollar above this line is a percentage of a
                    │  number that somebody measured off a drawing.

📊 Diagram (described). The waterfall above is deliberately drawn as a stack of decks resting on a foundation, not as a pie chart. A pie chart implies the pieces are independent. They are not. Widen the bottom slab — add 111 cubic yards of concrete — and everything above it moves: more labor hours, more equipment days, a bigger insurance base, a bigger bond base, a bigger fee. That structural dependence is why an error in the takeoff is never just the cost of the missed material.

Northgate's actual guaranteed maximum price is built exactly this way, and you have seen it before:

Look at what that structure tells you. Kestrel's fee is $1,804,800 on a $47,500,000 contract. That is a 4.0% markup on the subtotal and a 3.80% margin on the GMP — and we will come back to that difference in §12.9, because it is the arithmetic error that costs the industry more money than any other.

Now note the leverage. If the direct cost of work were understated by 1% — $400,000 — the contingency does not grow to cover it, because contingency was priced as a percentage of a number that was already wrong. The fee does not grow. The GMP does not grow, because it is guaranteed. That $400,000 comes out of one place. It comes out of the fee, which is $1,804,800. A 1% takeoff error consumes 22% of the profit on the job.

🔄 Check your understanding. Kestrel's project engineer proposes moving the cost of the temporary construction elevator from general conditions into the direct cost of work, arguing "it's equipment, and equipment is a direct cost." Does it matter which bucket it sits in?

12.2 🚪 An Estimate Is Not a Prediction

🚪 Threshold concept.

An estimate is not a prediction of what the job will cost. It is a priced bundle of assumptions, quantities, and risks — and every single one of them belongs to somebody.

The estimator's real product is not a number. It is a documented set of decisions about what is included, what is assumed, and who owns each unknown. The number is a byproduct.

This is a gateway. On one side of it you think estimating is a measuring exercise with a calculator at the end. On the other side you understand it as a risk-allocation exercise with a measuring exercise inside it. Almost everything estimators do that looks strange from the outside makes sense once you cross.

Before you cross this threshold, you think:

• The estimate is a forecast. The job either "comes in" or it doesn't, and a good estimator is one whose forecasts are accurate.
• Accuracy means the total is close to the actual cost.
• Exclusions are fine print. The real content is the number.
• An assumption is something you write down if you have time at the end.
• If you missed something, you were careless.

After you cross it, you think:

• The estimate is a proposal about who owns what. Every quantity you took off is a statement: "I will build this much, at this productivity, under these conditions, and if any of that is wrong, here is who pays."
• Accuracy means every assumption in the estimate is stated, testable, and owned. An estimate that is $200,000 high but whose every basis is documented is worth more than one that is $2,000 low and unexplained, because you can manage the first one.
• Exclusions are the most important sentences in the document. They are the boundary of the promise.
• An assumption written next to its number is a control. An assumption in your head is a liability.
• If you missed something, the process was missing a check. Find the check.

Tomás says it more compactly, and he says it a lot:

"A number isn't a prediction. It's a story about risk, and somebody has to own every chapter of it."

🔍 Why this works. Here is the mechanism, not the slogan. A construction estimate is produced weeks or months before the work happens, from documents that are incomplete, by people who will not do the work, for conditions nobody has seen yet. Under those conditions, no number can be reliably correct. What can be reliable is the statement of what the number assumed. So the profession optimizes for the thing that can actually be made reliable. A basis-of-estimate document converts an uncertain forecast into a set of certain, checkable claims — and a checkable claim can be assigned, priced, insured, negotiated, or transferred by contract. An unstated assumption cannot be any of those things. It can only be discovered, usually by an owner's representative, usually at the worst possible moment. That is why the paperwork is the product.

💡 Aha moment. This is also why two competent estimators can produce numbers 6% apart on the same drawings and both be right. They are not disagreeing about arithmetic. They are disagreeing about who owns the unknowns — and one of them may have quietly decided that the owner does. Whether that decision is stated on the bid form is the difference between a professional and a problem. We come back to this in §12.10.

🔄 Check your understanding. Two bidders on the Willow Street Community Center submit numbers $180,000 apart. Bidder A's estimate carries a stated 3% contingency and an exclusion for rock excavation. Bidder B's carries no contingency and no exclusions. Whose number is more accurate?

12.3 The Discipline of the Quantity Takeoff

A quantity takeoff is the systematic measurement of every item of work in the contract documents, in the units in which that work is bought and installed. That is the entire definition. The discipline is in the word systematic.

12.3.1 The seven rules

These are Tomás's, taped to the wall above the plot table. They are not stylistic preferences. Each one exists because somebody lost money.

1. Take off in the order the work gets built. Site, foundations, structure, envelope, interiors, MEP, finishes. When you follow the construction sequence, gaps announce themselves — you notice that nothing sits between the top of the foundation wall and the underside of the slab, because you just built the wall in your head and you have nowhere to put the slab.
2. Use consistent units, and write the unit next to every number. Not "620." Six hundred twenty what? A quantity without a unit is not a quantity.
3. Never round down. Round up, or carry the decimal. You will be wrong about something; be wrong in the direction that does not stop the pour.
4. Measure once, check twice — and make the second check a different method. Re-measuring the same way reproduces the same error. Count the footings, then extend the schedule, then compare the two. That is what would have caught Dani's nine missing footings in four seconds.
5. Keep every quantity traceable to the sheet and detail it came from. Not "footings — 1,240 CY." "Footings — 1,240 CY — S-101, S-104, schedule S-501/2." When the drawings are revised, and they will be, you need to know which quantities to re-take.
6. Take off gross, then deduct — and record both. Take the whole wall, then deduct the openings. If you take off only net, you cannot check yourself and you cannot re-price when a door moves.
7. Document every assumption in the same document as the number. Not in a separate file. Not in an email. On the takeoff sheet, in the row next to the number it modifies.

⚖️ What the contract says. Rules 5 and 7 are not just good practice; they are the difference between a change order and an argument. When the architect issues a bulletin that changes footing sizes, your entitlement to additional compensation depends on your ability to show what you priced before the change. A takeoff traceable to a specific sheet at a specific revision is evidence. A spreadsheet with a total and no sources is a recollection, and recollections lose. This is the same principle you will meet in Chapter 31 when we price changes and in Chapter 33 when we try to prove them.

12.3.2 Units of measure, and where they bite

Three conversions you will do a hundred times, worked:

Square feet of slab to cubic yards. Formula: CY = SF × (thickness in inches ÷ 12) ÷ 27. Northgate slab on grade: 33,000 SF at 5 inches.

   33,000 SF × (5 ÷ 12) ft = 33,000 × 0.41667 = 13,750 CF
   13,750 CF ÷ 27 CF/CY   = 509.3 CY  →  carried as 510 CY

That is the canonical 510 CY on the Northgate quantity sheet, and it is the single most-repeated calculation in commercial estimating. Memorize the shape of it: area, times thickness in feet, divided by twenty-seven.

Square feet to square yards. SY = SF ÷ 9. A 33,000 SF asphalt lot is 3,667 SY, not 11,000 SY. At $28/SY that error is $205,324 of phantom paving — and estimators make it every year, because "yard" sounds like "three feet."

Short tons and metric tonnes. Northgate's steel is 985 short tons = 1,970,000 lb = 893.6 metric tonnes. If a fabricator quotes €/tonne and you extend it against 985, you are 10.2% off before you start.

🧩 Productive struggle. Before you read the next section, try this with a pencil for three minutes.

You have two concrete elements. Both contain exactly 37 cubic yards of concrete.

• Element A: a foundation wall, 100 feet long, 12 inches thick, 10 feet tall.
• Element B: a footing pad, 20 feet by 15 feet, 3 feet 4 inches thick.

Check the volumes if you like (they are both 1,000 CF). Now: which one costs more to build, and by roughly what multiple? Do not look ahead. Write down your reasoning before you write down a number.

12.4 The Northgate Takeoff, Worked End to End

Let us actually do it. Everything below is the real Northgate quantity set, and every number is extended so you can follow the arithmetic.

12.4.1 Earthwork — and the three-measure problem

Northgate's canonical mass earthwork is 44,000 CY cut, 12,000 CY fill, 32,000 CY net export. Those three numbers are all in bank measure — material as it sits in the ground, undisturbed. That is the convention on a mass diagram, and 44,000 − 12,000 = 32,000 tells you the sheet is internally consistent.

But you cannot buy, haul, or place bank measure. Dirt exists in three states:

Swell takes you from bank to loose. Shrinkage takes you from bank to compacted. Both come from the geotechnical report and the material testing, not from a book:

These ranges are directional planning values. Use the project geotechnical report and, on a large earthwork job, a test strip. Do not treat a published range as a number.

Northgate's soils are common earth. Using 25% swell and 12% shrinkage:

Export haul volume:

   32,000 BCY × 1.25 = 40,000 LCY to be hauled off site

At a hauled-and-dumped rate of $18.50 per loose cubic yard, including tipping:

   40,000 LCY × $18.50 = $740,000
   Equivalent bank-measure rate: $740,000 ÷ 32,000 BCY = $23.13/BCY

💰 Money check. Suppose you take the 32,000 CY off the grading plan — correctly — and extend it against the hauler's $18.50 rate without converting, because the hauler said "eighteen-fifty a yard" and you wrote down "a yard."

   Wrong:   32,000 × $18.50 = $592,000
   Right:   40,000 × $18.50 = $740,000
   Gap:                       $148,000

$148,000 on one line, from one missing multiplication. On Northgate that is 8.2% of Kestrel's entire fee. Notice also that the arithmetic is not hard and nobody involved is stupid. The failure is a units failure — which is why Rule 2 says write the unit next to every number.

Now the trap on the fill side. The grading plan calls for embankment. Read carefully: does it call for 12,000 CY of compacted embankment, or does the mass diagram show 12,000 BCY of cut being reused as fill? On Northgate the earthwork line says the latter — 12,000 BCY of the cut stays on site. What does that actually produce?

   12,000 BCY × (1 − 0.12) = 10,560 CCY of compacted embankment

If the grading plan needs 12,000 CCY in place, you are 1,440 CCY short, and you have to import borrow:

   Borrow required (bank equivalent): 1,440 ÷ 0.88 = 1,636 BCY
   1,636 CY × $16.50/CY delivered and placed = $26,994  →  about $27,000

⚠️ Safety alert. Earthwork quantities and excavation safety are the same calculation, and estimators who forget that price an operation that is both wrong and lethal. OSHA's excavation standard, 29 CFR 1926 Subpart P, requires a protective system for trenches five feet deep or more in most soils; for Type C soil the maximum allowable slope is 1½ horizontal to 1 vertical. Price a 6-foot-deep, 2-foot-wide utility trench as a vertical cut and you have priced an illegal excavation. Price it sloped and the geometry changes completely:

   Vertical (illegal at this depth without a system):
     2 ft wide × 6 ft deep = 12 CF/LF = 0.44 CY/LF

   Sloped 1½:1 in Type C soil:
     top width = 2 + 2(1.5 × 6) = 20 ft
     average width = (2 + 20) ÷ 2 = 11 ft
     11 ft × 6 ft = 66 CF/LF = 2.44 CY/LF

That is 5.5 times the excavation and backfill. Over 1,850 LF of underground sanitary at $9.50/CY to excavate and backfill, the difference is $42,883 versus $7,733 — about $35,150. Which is also exactly why, on a tight urban-edge site like Northgate's north property line, you price a trench box instead: it is cheaper than sloping and it is the only option when you cannot widen the trench. Either way, you price something. A trench with no protective system in the estimate is a trench with no protective system in the field, and Subpart P violations are among the most consistently fatal exposures in the industry. We build the trench-box cost into a real unit price in §12.7.

12.4.2 Concrete — the footings, worked

Here is the corrected footing takeoff Tomás was holding. This is the second pass, with the nine canopy and loading-dock footings from S-104 included, the F-6 thickness read from the correct row, and the volumes carried to two decimals so the extension does not drift.

Pads (per the footing schedule, S-501/2):

Two extensions in full, so the pattern is unmistakable.

F-4, a 10-foot-square pad 3 feet thick:

   Volume each = 10.0 ft × 10.0 ft × 3.0 ft = 300.0 CF
   300.0 CF ÷ 27 CF/CY = 11.11 CY each
   11.11 CY × 26 footings = 288.89 CY

F-6, the 14-foot-square pad — the one Dani read wrong:

   Correct:   14.0 × 14.0 × 4.0 = 784.0 CF ÷ 27 = 29.04 CY each × 6 = 174.22 CY
   As read:   14.0 × 14.0 × 3.5 = 686.0 CF ÷ 27 = 25.41 CY each × 6 = 152.44 CY
   Difference:                                                          21.78 CY

Six inches of thickness on six footings. Twenty-two cubic yards.

Pedestals (per the pedestal schedule and typical detail 4/S-501):

Rolling it up:

That is where the canonical 1,240 CY comes from. Note the last two rows carefully. The estimate does not carry the geometry; it carries the geometry plus a named, quantified allowance. And note that the allowance is applied to the quantity, not to the unit price — because if you put waste in both places you will double-count it, which is a real and common error we will name again in §12.5.

12.4.3 Formwork — where beginners lose the most money

Now go back to your 🧩 Productive struggle answer.

Element A — a 100-foot wall, 12 inches thick, 10 feet tall — contains 100 × 1.0 × 10 = 1,000 CF = 37.04 CY. Its formwork is both faces of the wall:

   100 LF × 10 ft × 2 faces = 2,000 SFCA

Element B — a pad 20 × 15 × 3'-4" — contains 20 × 15 × 3.333 = 1,000 CF = 37.04 CY. Its formwork is the perimeter only:

   perimeter = 2 × (20 + 15) = 70 LF
   70 LF × 3.333 ft = 233 SFCA

Same concrete. 2,000 SFCA versus 233 SFCA — an 8.6-to-1 difference. At a formwork unit cost of $6.28 per SFCA (which we build from scratch in the 📋 Try it drill), that is $12,560 of forms on the wall versus $1,463 on the pad. The concrete itself, at $242.34/CY, costs $8,976 either way. On the wall, the formwork costs forty percent more than the concrete.

💡 Aha moment. Formwork quantity has almost nothing to do with concrete quantity. It is driven by surface geometry — perimeter, height, number of faces, number of reuses. This is the single biggest reason novice concrete estimates go wrong, and it is why concrete is never estimated as "$X per cubic yard, all in." A cubic yard of slab, a cubic yard of footing, a cubic yard of foundation wall, and a cubic yard of column are four completely different products that happen to share a material.

Northgate foundation formwork, taken off properly:

One extension shown: F-4 formwork. A 10-by-10 pad has a 40-LF perimeter; at 3 feet deep that is 40 × 3 = 120 SFCA each; 120 × 26 = 3,120 SFCA.

Compare the two totals: 2,370 CY of foundation concrete (1,240 footings + 620 walls and grade beams + 510 slab on grade) requires 38,946 SFCA of formwork. That is 16.4 SFCA per cubic yard on average — and the average is useless, because the slab needs almost none and the walls need 46.

🔄 Check your understanding. A value-engineering proposal thickens the foundation walls from 12 inches to 16 inches so a wall can be shortened. Concrete volume goes up 33%. What happens to the formwork cost?

12.4.4 Reinforcing steel — from cubic yards to pounds, carefully

Early in design, before the placing drawings exist, you convert concrete volume to reinforcing weight using a pounds-per-cubic-yard factor. These factors are legitimate for conceptual and budget work and dangerous everywhere else.

261,560 lb ÷ 2,000 = 130.8 tons, or 2,615.6 CWT. At $1.05 per pound furnished, fabricated, delivered, and installed, that is $274,638.

The slab-on-grade bar was taken off directly rather than by factor, because you can:

   #4 bar @ 18" o.c. each way over 33,000 SF:
     LF of bar = (33,000 ÷ 1.5) × 2 directions = 44,000 LF
     add 8% for laps and starter bars      = 47,520 LF
     #4 bar weighs 0.668 lb/LF
     47,520 × 0.668 = 31,743 lb  →  31,700 lb

The caution, stated as strongly as it deserves: the moment a reinforcing schedule and placing drawings exist, the lb/CY factor is no longer an acceptable basis. It is a stand-in for information you do not have. When you have the information, use it — bar by bar, mark by mark, with laps and bends. Detailers and rebar fabricators do exactly this, which is why the fabricator's quantity and your factor-based quantity will differ, and why you must reconcile them rather than picking the one you like.

💰 Money check. A 15% error in a lb/CY factor on Northgate's foundations:

   261,560 lb × 15% = 39,234 lb
   39,234 lb × $1.05/lb = $41,196

$41,200 from choosing 110 lb/CY instead of 126. That is why conceptual factors get replaced, and why the basis-of-estimate note next to that number must read "lb/CY factor — replace when placing drawings issue."

🏗️ From the field. Watch the rate, not just the quantity. Cottonwood Creek Bridge — the DOT job over in the next county — carries reinforcing steel at $1.42 per pound. Northgate's foundation rebar is $1.05. Same commodity, 35% apart. Why? Cottonwood's bar is epoxy-coated, heavily bent, placed over water under traffic control, and inspected by a state inspector who checks every splice. It is not the same work. If you ever find yourself borrowing a unit price from a different project type because it was handy, this is the failure mode. Unit prices are not prices for materials; they are prices for situations.

12.4.5 Structure — 985 tons and 99,000 square feet of deck

Structural steel is bought by the ton and it is one of the slipperiest units in the business, because "a ton of steel" can mean at least four different things:

Northgate's canonical 985 tons erected. At a furnished-and-erected price of $4,850 per ton for a four-story, moderately complex steel frame:

   985 tons × $4,850/ton = $4,777,250
   Per building square foot: $4,777,250 ÷ 132,000 SF = $36.19/SF

That $/SF figure is not the estimate — it is the check. You take off a quantity, extend it, then convert the result into a unit you have historical data for and see whether the answer looks like reality. This is the conceptual-estimating skill from Chapter 11 used in reverse, as a sanity test on detailed work. Do it on every major package.

Composite metal deck: 99,000 SF, with 3¼-inch lightweight topping. At $3.35/SF furnished and installed, the deck is 99,000 × $3.35 = $331,650. The topping concrete is a separate takeoff and a separate trap:

   Average concrete thickness over a 3" composite deck with 3¼" cover
   is roughly 4¾ inches — NOT 3¼ inches, because the flutes fill.
   Get this from the deck manufacturer's published section-property
   tables for the specific profile; it varies with deck depth and rib geometry.

   99,000 SF × (4.75 ÷ 12) ft ÷ 27 = 99,000 × 0.3958 ÷ 27 = 1,451 CY

Use 3¼ inches instead of 4¾ and you take off 993 CY instead of 1,451 — you are 458 cubic yards short, which at $198/CY of lightweight concrete is $90,684 of material you did not buy, before you count the labor and pumping. That is a shape-of-the-deck problem, and the only fix is reading the manufacturer's table.

12.5 Waste, Shrinkage, and Loss — and the Double-Count Trap

Every material quantity has two versions: the quantity that ends up in the building and the quantity you have to buy. The gap between them is waste, and it is a real, recurring, budgetable cost, not a rounding error.

These are not universal constants. They vary by crew, by layout, by the stock sizes available in your market, by weather, and by how tidy your site is. The company that tracks its own waste from its own cost reports — see Chapter 28 — replaces every number in this table within about three years, and its estimates get better for it.

Working a waste factor is trivial arithmetic. Applying it correctly is not. Two rules:

Rule one: pick a place and put it there. Either inflate the quantity or inflate the unit price — never both. Kestrel's convention is quantity, because a quantity you can see is a quantity you can audit. Watch what happens when you forget:

   Footings, net:                    1,216 CY
   Quantity inflated 2%:             1,240 CY
   Unit price with 2% waste built in: $181.46/CY  (vs. $177.90 clean)
   Extended:  1,240 × $181.46 = $225,010
   Correct:   1,240 × $177.90 = $220,596
   Phantom cost:                       $4,414

On concrete alone, on one job, from a bookkeeping ambiguity. Now scale it across forty divisions.

Rule two: waste is a material factor, not a labor factor. Your crew does not install the waste. They install the neat quantity. If you apply a waste factor to your labor hours you are budgeting crew time to install concrete that went home in the drum. (The exception: material that is handled twice — cut-off drywall that has to be carried to a dumpster — is a real labor cost, but you capture it in the productivity rate, not in a waste factor.)

The deck-deflection example, because it is the one that bites hardest. Northgate's elevated slabs compute to 1,451 CY of topping. Composite deck deflects under the weight of wet concrete, and the concrete finds the low spot, so you place a thicker average slab than the geometry says:

   1,451 CY × 5% deflection allowance = 72.6 CY additional
   72.6 CY × $198/CY lightweight = $14,375

Fifteen thousand dollars is unpleasant. Being seventy cubic yards short at 11:40 a.m. with a pump on the deck, a finishing crew standing there, and a batch plant forty minutes away is much worse, because now you are buying a cold joint you did not design.

🔄 Check your understanding. Northgate's gypsum board is 412,000 SF. Your drywall subcontractor's estimator says they carry 8% waste; a competing bidder says 12%. On material alone at $0.42/SF, how much is that disagreement worth — and what does it actually tell you about the two bidders?

   412,000 × 1.08 = 444,960 SF
   412,000 × 1.12 = 461,440 SF
   Difference: 16,480 SF × $0.42 = $6,922

12.6 Digital Takeoff: What It Changes and What It Absolutely Does Not

Nobody at Kestrel takes off a 132,000 SF hospital with a red pencil, and Tomás would be the first to say so. On-screen takeoff has been standard for two decades and model-based extraction is now routine on projects that have a usable model.

What digital takeoff genuinely changes:

Bluebeam Revu is the common on-screen measurement and markup tool; On-Screen Takeoff is a long- established dedicated takeoff package; Assemble and similar tools do model-based quantity extraction; and most modeling platforms will produce element schedules directly. Kestrel uses more than one, because no single tool is best at both flat sheets and models. Which specific tool your company uses matters far less than the discipline you bring to it. (See Appendix H for the broader tool landscape.)

What digital takeoff does not change — any of it:

• The assumptions. Software does not know that the partition schedule shows walls to deck in the imaging suite and to six inches above ceiling everywhere else. You know that, or nobody does.
• Whether the item is on the sheet. Dani's nine missing footings would have been missed just as cleanly in Bluebeam, because Bluebeam was not open to S-104 either. Digital measurement makes you faster at measuring the drawing you are looking at.
• The scope boundary. Software does not know whether the topping slab is in the concrete package or the steel package. That is a decision, and decisions have owners.
• Whether the units are right. A tool will happily report 11,000 SY.

And it adds one genuinely new risk. Quantities extracted from a model whose level of development (LOD) does not support them are precisely wrong, which is far more dangerous than being roughly wrong. A model built for spatial coordination represents a partition as a generic wall of the correct thickness in the correct place. It may say nothing true about stud gauge, stud spacing, whether the wall runs to deck, how many layers of board it carries, or whether it is rated. Extract "412,000 SF of gypsum board" from that model and you will get a number with three decimal places and no meaning.

Grace Lindqvist, Kestrel's VDC manager, puts a single line at the top of every quantity extraction she issues: "Extracted at LOD 300 — geometry and location reliable, materials and assemblies not." That one sentence is the difference between a useful tool and an expensive mistake. We take this apart properly in Chapter 35.

💡 Aha moment. Watch for the precision illusion. A tool that reports "18,614.37 LF of partition" feels more trustworthy than a hand count of "18,600 LF," and it is not — it is the same measurement of the same drawing, displayed with false precision. Decimals are a property of the display, not of the knowledge. Tomás rounds every extracted quantity to the precision he can defend before it goes in the estimate, specifically so that nobody downstream mistakes resolution for accuracy.

12.7 From Quantity to Cost: Building a Unit Cost From First Principles

This is the heart of the trade. A quantity is not money. To turn one into the other you need a unit cost, and the reason you build it rather than look it up is that once you can build one, you can price anything — including work nobody has a published rate for.

The formula, in words and symbols:

                  crew cost per hour
   unit cost  =  --------------------  +  material per unit  +  equipment per unit
                 production per hour

Three terms, and they behave completely differently. Learn the difference and you will never be confused again:

• Labor is time-based. It costs money per hour whether anything gets installed or not.
• Equipment is time-based. Same. A rented excavator bills for Tuesday whether it dug or idled.
• Material is quantity-based. It costs money per unit installed, and the rate does not care how long you took.

🔍 Why this works. This is the mechanism behind the whole formula. Two of your three cost families are billed in time and one is billed in quantity — but the estimate has to be denominated in quantity, because that is what the drawings give you. The production rate is the exchange rate between the two. Divide a time-based cost by a production rate and it becomes a quantity-based cost. That single division is what converts "a crew costs $257.46 an hour" into "formwork costs $4.43 a square foot." Everything else in estimating is bookkeeping around that one conversion. It also tells you, immediately and without further analysis, which term moves when conditions change: anything that slows the crew inflates the labor and equipment terms and leaves material alone. You will use that in §12.8.

12.7.1 The fully burdened labor rate, built line by line

The wage is not the cost. The burdened rate is the cost, and the gap is enormous. Here is a carpenter on Kestrel's open-shop private work, built from a $34.00 base wage:

   Burden factor = $50.47 ÷ $34.00 = 1.484   →   48.4% burden

Six things in that table deserve a sentence each, because each one is a place people go wrong.

The taxable wage base is $35.87, not $42.08. Bona fide contributions to a benefit plan are generally not taxable wages; vacation and holiday pay generally is. So payroll taxes here are computed on $34.00 + $1.87 = $35.87. Get this backwards and you over-burden every hour in the estimate. Tax treatment of fringes varies and changes — confirm the current treatment with your controller, not with a textbook.

FUTA has a low annual wage cap and SUTA rates vary enormously by state and by your own experience rating. The 2.9% blended figure here is a Kestrel-specific, year-specific number derived from last year's actual payroll. Yours will be different. This is a place where "look up the rate" is the wrong instinct and "ask your accounting department for last year's effective rate" is the right one.

Workers' compensation is priced by class code, and the spread is brutal. A carpenter, a roofer, an ironworker, and an office employee are four different rates for the same company on the same day — sometimes differing by a factor of five or more. Manual rates are set by state rating bureaus and vary by jurisdiction; your experience modification rate (EMR) then multiplies them. Kestrel's EMR of 0.82 reflects a good loss history, and it is worth (1.00 − 0.82) × 9.80% × payroll on every hour the company works — which is the most direct financial argument for a safety program you will ever see. We make that argument properly in Chapter 24.

The same worker costs different amounts in different crews. Watch this in §12.7.4: a $26.00 laborer burdens to $39.51 in a framing crew and $40.74 in a trench crew, because the excavation class code carries a higher comp rate. Same person, same wage, different work, different cost. If your estimating system carries one burdened laborer rate for the whole job, it is wrong somewhere.

Small tools and consumables — blades, bits, screws, string, chalk, gloves, layout paint — are real and are usually carried as a percentage of labor because tracking them per unit is not worth the effort. 2–4% is the common range.

Union work is a different build-up. On public work with prevailing wage, the fringe package is defined by the wage determination, the contribution rates are specified rather than negotiated, and certified payroll makes the reporting non-optional. The structure of the calculation is identical; the inputs come from a different place. Prevailing-wage requirements, thresholds, and reporting vary by jurisdiction and by funding source — the federal Davis-Bacon framework and state "little Davis-Bacon" statutes are not the same and do not always apply to the same work. Check the solicitation, not your memory. Chapter 20 covers this in full.

12.7.2 Material cost, built properly

   material per unit = (unit price + waste) + freight + sales tax

Northgate's footing concrete:

Three notes with money attached.

Freight is a line, not an assumption. Ready-mix is quoted delivered. Steel studs, board, and manufactured products often are not, and "FOB factory" means the truck is yours. On a 132,000 SF building, freight on interior materials alone runs into six figures.

Sales tax on construction materials varies more than almost anything else in this book. Some jurisdictions tax the contractor as the consumer of materials; some tax the improvement; some exempt public work; some exempt materials incorporated into manufacturing facilities; some apply a use tax when materials cross a line. Rates change. Confirm the treatment for your jurisdiction and your project type before you carry a number, and put the answer in your basis of estimate.

A quoted material price has a shelf life. A ready-mix quote might hold thirty days; a structural steel quote in a volatile market might hold ten; a curtain-wall quote might be firm only until a stated purchase date. Every quote in the estimate should be logged with its validity date, and every scope with a buy date past that validity needs an escalation line. Forgetting this is a top-five estimating error and we price it in §12.10.

⚖️ What the contract says. Watch the difference between an allowance and a priced scope. An allowance is a stated dollar amount carried in the contract sum for work that is not yet defined — "$95,000 for signage," say. Under the standard AIA general conditions framework, allowances typically cover the material and delivery cost, with the contractor's handling, labor, overhead, and profit carried separately unless the documents say otherwise, and the contract sum is adjusted by the difference between the allowance and the actual cost. Two consequences for the estimator: read what the allowance is stated to include, and carry the labor and installation yourself if the documents put it on you. An estimator who carries a $95,000 allowance and nothing else has priced material for a scope whose installation is still their problem.

12.7.3 Equipment cost

Equipment splits into owned and rented, and the distinction changes the arithmetic more than people expect:

Two rules that save money. First, price mobilization separately and explicitly. Moving a crawler crane onto a tight urban site is a priced operation with its own crew, permits, and traffic plan; it is not "included." Second, price idle time honestly. If the excavator is needed on Monday and Friday of the same week, you are paying for the week. Equipment utilization is a schedule question, which is why equipment estimating and scheduling are done together — Chapter 21 does the own- versus-rent analysis properly, and Chapter 14 supplies the durations that drive it.

12.7.4 The formula, worked three times

Example 1 — one cubic yard of footing concrete, placed and finished.

(Formwork and reinforcing are priced separately. This is placement and material only.)

Crew — Jamal Foster's footing placement crew:

Production: 62 CY per 8-hour crew-day for scattered spread footings with a mix of chute and pump placement = 7.75 CY per crew-hour.

   Labor      = $371.99 ÷ 7.75 CY/hr                       = $48.00 / CY
   Material   = (from §12.7.2)                              = $177.90 / CY
   Equipment  = blended pump allowance $9.50 + buggies,
                vibrators, forklift allocation $6.94        = $16.44 / CY
   ------------------------------------------------------------------
   UNIT COST                                                = $242.34 / CY

Extended against the takeoff: 1,240 CY × $242.34 = $300,502.

What it means for the job: three hundred thousand dollars of footing concrete, of which only 20% is labor. On this scope you protect the number by controlling material waste and short loads, not by pushing the crew.

Example 2 — one square foot of metal-stud partition framing.

Kestrel subcontracts drywall, so why build this? Because you must have an independent number for every major subcontracted trade — §12.9 explains why. This is the number Tomás checks the drywall bids against.

Production: 1,050 SF of wall per 8-hour crew-day for 3⅝-inch studs at 16 inches on center, 10'-6" walls, one-hour rated, in a fit-out with many short walls and openings = 131.25 SF/hr.

Material, per SF of wall (framing only, no board):

   Studs:  0.75 LF of stud per SF of wall × $1.18/LF        = $0.885
   Track:  2 LF per LF of wall ÷ 10.5 ft × $1.06/LF         = $0.202
   Fasteners, clips, deflection track, backing, bracing     = $0.340
   Waste on studs and track, 8%                             = $0.087
   Subtotal                                                  = $1.514
   Sales tax @ 7.25%                                         = $0.110
   Material per SF (freight included in delivered price)     = $1.620

   Labor      = $307.93 ÷ 131.25 SF/hr                      = $2.346 / SF
   Material                                                  = $1.620 / SF
   Equipment  = scissor lift, laser, screw guns
                $185/day ÷ 1,050 SF                          = $0.176 / SF
   ------------------------------------------------------------------
   UNIT COST                                                 = $4.14 / SF

Northgate's partitions: 18,600 LF × 10.5 ft average height = 195,300 SF of wall. 195,300 × $4.14 = $808,542.

Example 3 — one linear foot of 4-inch underground sanitary pipe.

Watch the workers' comp class code change the same laborer's cost.

Production: 160 LF per 8-hour crew-day for 4-inch SDR-35 PVC at 4 to 6 feet of cover inside the building pad, using a trench box = 20 LF/hr.

   Material:
     4" SDR-35 PVC with gaskets                              = $4.35 / LF
     Fittings allowance (wyes, bends, cleanouts, prorated)   = $2.10 / LF
     Bedding and pipe-zone stone, 0.119 CY/LF @ $31.00/CY    = $3.68 / LF
     Waste on pipe, 3%                                       = $0.13 / LF
     Sales tax @ 7.25% on taxable items                      = $0.48 / LF
                                                    Material = $10.74 / LF

   Equipment:
     Mini excavator, rental + fuel/maintenance   $480 / day
     Trench box rental                           $145 / day
     Plate compactor / jumping jack               $48 / day
                                          Total  $673 / day ÷ 160 LF = $4.21 / LF

   Labor      = $254.29 ÷ 20 LF/hr                          = $12.71 / LF
   ------------------------------------------------------------------
   UNIT COST                                                 = $27.66 / LF

Extended over 1,850 LF of 4-inch underground sanitary: 1,850 × $27.66 = $51,171.

Now read that unit cost as a set of claims, because that is what it is. It assumes a trench box (so it assumes you are not sloping, which assumes you have room). It assumes 4 to 6 feet of cover (so it assumes no deep runs). It assumes 160 LF a day (so it assumes clear access, no existing utilities to hand-dig around, and no rock). And it contains no spoil haul-off, because the assumption is that trench spoil goes into the mass-excavation export already priced in §12.4.1. Every one of those claims belongs on the basis-of-estimate page. Three of them are wrong on some jobs. On this job you know which, because you wrote them down.

📋 Try it — build a formwork unit cost from scratch.

You are pricing 14,200 SFCA of 10-foot foundation wall formwork on a Kestrel job. Everything you need is below. Do not look at the answer until you have your own numbers.

The crew: 1 foreman at $46.00/hr base, 3 carpenters at $34.00/hr base, 1 laborer at $26.00/hr base. Eight-hour day.

The fringe and burden schedule:

Production rate: 465 SFCA per crew-day (form, set, plumb, brace, strip, clean, and oil).

Material: panel amortization $0.62/SFCA, ties and accessories $0.41/SFCA, form release and consumables $0.15/SFCA.

Equipment: one rough-terrain forklift charged to this crew at $310 per day.

Find: (a) the fully burdened crew cost per hour; (b) the crew cost per day; (c) the unit cost per SFCA; (d) the total for 14,200 SFCA; (e) the unit cost and the total if production drops 15%.

   Health and welfare                                 $4.85
   Retirement       4.0% × $46.00                  =  $1.84
   Vacation accrual 5.5% × $46.00                  =  $2.53
   Wage + fringes                                     $55.22
   Taxable base = $46.00 + $2.53                   =  $48.53
   Payroll taxes  10.55% × $48.53                  =  $5.12
   Workers' comp   8.04% × $48.53                  =  $3.90
   General liab.   1.90% × $48.53                  =  $0.92
   Small tools     2.50% × $55.22                  =  $1.38
   BURDENED                                           $66.54

   H&W $4.85 + retirement $1.36 + vacation $1.87
   Wage + fringes                                     $42.08
   Taxable base = $34.00 + $1.87                   =  $35.87
   Payroll taxes  10.55% × $35.87                  =  $3.78
   Workers' comp   8.04% × $35.87                  =  $2.88
   General liab.   1.90% × $35.87                  =  $0.68
   Small tools     2.50% × $42.08                  =  $1.05
   BURDENED                                           $50.47

   H&W $4.85 + retirement $1.04 + vacation $1.43
   Wage + fringes                                     $33.32
   Taxable base = $26.00 + $1.43                   =  $27.43
   Payroll taxes  10.55% × $27.43                  =  $2.89
   Workers' comp   7.10% × $27.43                  =  $1.95
   General liab.   1.90% × $27.43                  =  $0.52
   Small tools     2.50% × $33.32                  =  $0.83
   BURDENED                                           $39.51

   Crew:  1 × $66.54  =  $66.54
          3 × $50.47  = $151.41
          1 × $39.51  =  $39.51
          ---------------------
          CREW COST PER HOUR = $257.46

   Labor      = $2,059.68 ÷ 465 SFCA         = $4.4295
   Material   = $0.62 + $0.41 + $0.15        = $1.1800
   Equipment  = $310 ÷ 465 SFCA              = $0.6667
   ----------------------------------------------------
   UNIT COST                                 = $6.2762  →  $6.28 / SFCA

   14,200 × $6.2762 = $89,122
   (at the rounded $6.28: $89,176 — a $54 difference. Extend at full
    precision and round the total, not the rate.)

   Labor      = $2,059.68 ÷ 395.25           = $5.2110
   Material   = unchanged                     = $1.1800
   Equipment  = $310 ÷ 395.25                = $0.7843
   ----------------------------------------------------
   NEW UNIT COST                             = $7.1753  →  $7.18 / SFCA
   Total: 14,200 × $7.1753 = $101,889
   Increase: $101,889 − $89,122 = $12,767

12.8 Productivity Rates: Where They Come From and What Moves Them

The production rate is the most consequential number in the unit cost and the least defensible one. Change 465 SFCA per day to 395 and you moved $12,767. Nobody audits it. Everybody assumes it.

12.8.1 Three sources, ranked

1. Your own historical data. This beats everything. If Kestrel formed 41,000 SFCA of foundation wall across four jobs last year and the cost reports say it took 705 crew-days, that is 58.2 SFCA per crew-hour with your crews, your form system, your supervision, in your market. No published figure can compete with that, and it is why Chapter 28 matters to estimators more than to almost anyone else. The cost report is what generates next year's estimate. A company that codes its costs sloppily is not just failing at cost control; it is destroying its own estimating database, and it will not notice for three years.

2. Published data, used honestly. RSMeans and comparable published cost data are real, useful, and widely used. They are compiled from broad surveys, adjusted by location factors, and organized by CSI division. What they give you is a defensible starting point and a sanity check. What they cannot give you is your crew, your site, your foreman, or your form system. Use published data to price scopes you rarely touch, to check a number that surprises you, and to build a first-pass conceptual estimate. Do not use it as your primary basis on work you self-perform, and never present a published figure as though it were company data. Always note in the basis of estimate which source and which adjustments you applied.

3. Crew experience. Ask Jamal Foster. He has formed more foundation wall than the database has. His number will be right about the shape of the work — "that wall's got eleven corners and two step-downs, you're not getting 465 out of it" — and it will be optimistic about the rate, because foremen remember good days. Use it to adjust, not to originate.

🏗️ From the field. I want to tell you about a number I got wrong, because it is the single most expensive mistake I have made as an estimator.

Early on, I priced a corridor renovation in an occupied hospital. I had beautiful historical data for metal-stud framing — 1,180 SF per crew-day — from a distribution warehouse we had just finished. Clean floor, no ceilings, sixteen-foot straight runs, material staged twenty feet away, crew of six working without interruption for eight hours.

The hospital corridors were eight feet wide, occupied, with a fire watch, with infection-control barriers, with material carried in on carts through a service elevator we shared with dietary, with work windows that closed at 3 p.m. for shift change, and with a nurse manager who could and did stop us for a patient transport.

We got 540 SF a day. Not 1,180. Fifty-four percent of my rate.

I did not have a bad crew. I had a bad assumption, and I had not written it down, so nobody caught it. If I had put one line on the takeoff sheet — "framing rate 1,180 SF/CD, basis: warehouse, unobstructed, full-shift access" — Margo would have read it in nine seconds and asked me what I thought a hospital corridor was. That one missing sentence cost about $310,000.

The rate was not the mistake. The undocumented rate was the mistake.

12.8.2 What moves a production rate

Read these as ranges and as planning aids, not as entitlements. Magnitudes vary widely across studies, trades, and projects; several are contested in the literature; and the effects are not independent — stacking trades and running sixty-hour weeks is not simply the product of the two factors. Industry bodies including AACE International and the Lean Construction Institute have published extensively on inefficiency measurement, and the honest summary is that the direction of these effects is well established and the precise magnitude on any given job is not. Use them to price a decision, and use measured data from your own job to prove one.

Worked example — the 10-hour, 6-day schedule. This is the calculation behind every acceleration decision in the book, including the Northgate steel recovery in Chapter 14.

Baseline, from the 📋 Try it drill: crew of 5, 465 SFCA per 8-hour day = 58.125 SFCA/crew-hr, crew cost $257.46/hr.

Go to 6 × 10 = 60 hours a week, sustained. Take the mid-range factor: 0.84.

   Adjusted rate = 58.125 × 0.84 = 48.825 SFCA/crew-hr

   Baseline week:  58.125 × 40 hr = 2,325.0 SFCA
   OT week:        48.825 × 60 hr = 2,929.5 SFCA
   Additional output:                  604.5 SFCA   (+26.0%)

Now the cost. Overtime premium applies to the wage, not to the hourly fringe contributions — health and welfare is paid per hour worked, not per dollar earned — but the premium is taxable wages, so payroll taxes, comp, and GL ride on it:

   Crew base wages: $46.00 + (3 × $34.00) + $26.00 = $174.00/hr

   Overtime premium per OT hour (0.5 × base, plus burdens on the premium):
     Foreman     0.5 × $46 = $23.00 × 1.2049  = $27.71
     Carpenters  0.5 × $34 = $17.00 × 1.2049  = $20.48 each × 3 = $61.44
     Laborer     0.5 × $26 = $13.00 × 1.1955  = $15.54
     Premium per OT hour                       = $104.69

   Weekly labor cost:
     60 hr × $257.46 (straight-time all-in)    = $15,447.60
     20 OT hr × $104.69 premium                =  $2,093.80
                                          Total = $17,541.40

   Baseline week: 40 hr × $257.46              = $10,298.40

   Labor unit cost, baseline: $10,298.40 ÷ 2,325.0   = $4.43 / SFCA
   Labor unit cost, 60-hr:    $17,541.40 ÷ 2,929.5   = $5.99 / SFCA
                                                       +35.2%

💰 Money check. You spent $7,243 more per week ($17,541 − $10,298, a 70% increase in weekly labor cost) to gain 604.5 SFCA of output (a 26% increase).

   $7,243 ÷ 604.5 SFCA = $11.98 per additional SFCA

Nearly triple the baseline unit cost of $4.43. That is what acceleration costs, and it is why the Northgate steel-recovery decision comes out as close to a wash on pure arithmetic. Slipping substantial completion costs Kestrel $10,650 per calendar day — $5,150 of extended general conditions plus $5,500 of liquidated damages — so absorbing the 23-day steel delay costs $244,950, and accelerating to recover 17 of those days costs $168,000. The gap is about thirteen thousand dollars, which is noise on a $47.5M job; the decision gets made on something else entirely (Meridian's interim clinic lease expiring October 1, Year 2). Sustained overtime is not a productivity strategy. It is a time purchase, and you should know its price before you buy it. Chapter 29 works the full acceleration decision.

⚠️ Safety alert. There is a second cost on that table and it does not appear in any of the arithmetic above. Sustained sixty-hour weeks raise incident rates. Fatigue degrades judgment, housekeeping, and the willingness to stop and re-inspect something. On Northgate, the acceleration after the steel delay produced trade stacking, a rework event on deck-edge detailing, and a spike in near-misses in weeks 34 through 36 — including the scaffold near-miss on the north elevation that Bea Salgado investigated. The third finding in that investigation was "a crew running behind after the steel acceleration, with an unwritten 'make it up' pressure." An estimate that carries an optimistic production rate is a schedule-pressure generator, and schedule pressure is a hazard exactly like an unguarded edge. Price the rate you can actually achieve safely.

🔄 Check your understanding. Your historical database says a crew hangs 1,050 SF of partition a day. The Northgate imaging suite is on the second floor, adjacent to the occupied clinic, and will be framed while the mechanical contractor is running duct mains in the same corridor. What rate do you carry, and — more importantly — what do you write down?

12.9 Rolling Up to a Bid: Indirects, Markups, and the Margin Trap

You have direct costs. Now build them into a price.

12.9.1 The estimate summary

Here is a complete rollup for a mid-size lump-sum bid, structured the way Kestrel's summary sheet is structured. Follow the order; the order is not arbitrary, because several lines are percentages of the lines above them.

Four notes on that sheet.

Rounding down $1,808 is a decision, not an accident. Someone made it, on purpose, and it is a gift to the owner of $1,808. Round consciously. Never let a number get rounded by a spreadsheet you did not check.

Contingency is not padding. It is a priced, owned, drawn-down reserve against identified risks — the concept from Chapter 6. A contingency with no risk register behind it is either fat or a lie. When you carry 2%, you should be able to say what the 2% is for.

Escalation is applied to the scope that will actually be bought late, not to the whole job. If $2.9M of your material and subcontract scope will be purchased twelve to eighteen months after the bid, that is what escalates. Applying an escalation percentage to the total is lazy and usually wrong in both directions.

Sales tax on the contract is a separate question from sales tax on materials. Some jurisdictions impose a gross-receipts or contractor's excise tax on the contract amount itself. Where that applies, it is a line on this summary, computed on the price, and it must be included in the price — which makes it circular in exactly the way the bond premium is. Confirm the treatment where you are building.

12.9.2 The bond premium, worked — including the circularity

Payment and performance bonds are usually priced on a sliding scale applied to the final contract amount. A representative structure — rates vary by surety, by the contractor's financial strength, by contract size, and over time; get a rate letter from your agent — looks like this:

Worked on the Willow Street Community Center's $6,800,000 contract:

   First   $500,000 × $25.00/M  = $12,500
   Next  $2,000,000 × $15.00/M  = $30,000
   Next  $2,500,000 × $12.00/M  = $30,000
   Next  $1,800,000 × $10.00/M  = $18,000
   ------------------------------------------
   BOND PREMIUM                  = $90,500
   Effective rate: $90,500 ÷ $6,800,000 = 1.331%

And here is the wrinkle nobody tells you about. The premium is a percentage of the contract price, and the premium is in the contract price. That is circular. If S is your subtotal before bond and r is the effective rate:

   Price = S + (r × Price)      →      Price = S ÷ (1 − r)

On a subtotal of $6,710,000 at r = 1.33%:

   Naive:    bond = $6,710,000 × 0.0133 = $89,243   →  price $6,799,243
   Correct:  price = $6,710,000 ÷ 0.9867 = $6,800,446
             bond  = $6,800,446 − $6,710,000 = $90,446
   You under-bonded yourself by $1,203.

Small on Willow Street. Scale the same error to Northgate's $47.5M and it is roughly $8,400 — and unlike most estimating errors, this one is pure arithmetic and completely avoidable.

12.9.3 Markup is not margin, and the difference is money

This is the most common arithmetic mistake in the construction industry. Not a subtle one. A common one, made by experienced people, every week.

• Markup is a percentage added to cost. Price = Cost × (1 + markup)
• Margin is the profit as a percentage of price. Margin = (Price − Cost) ÷ Price

They are never the same number, and the gap widens as the percentage rises.

Work it both directions on a $1,000,000 cost.

Direction 1 — you mark up 10% and want to know your margin:

   Price  = $1,000,000 × 1.10 = $1,100,000
   Profit = $100,000
   Margin = $100,000 ÷ $1,100,000 = 9.09%

You wanted 10%. You got 9.09%.

Direction 2 — you want a 10% margin and need to know the markup:

   Price  = Cost ÷ (1 − margin) = $1,000,000 ÷ 0.90 = $1,111,111
   Profit = $111,111
   Markup = $111,111 ÷ $1,000,000 = 11.11%

The two conversion formulas — write these on the inside of your hard hat:

   markup = margin ÷ (1 − margin)
   margin = markup ÷ (1 + markup)

💰 Money check. Northgate's direct cost of work is $40,000,000. The difference between marking up 10% and marking up the 11.11% required to actually earn a 10% margin:

   $40,000,000 × 11.11% = $4,444,000
   $40,000,000 × 10.00% = $4,000,000
   Difference:             $444,000

$444,000. That is nearly a quarter of Kestrel's entire fee on the job, and it is the difference between two numbers that sound identical when spoken aloud in a bid room at 1:40 p.m.

And check Kestrel's actual position: the CM fee is $1,804,800 on a $47,500,000 GMP. $1,804,800 ÷ $47,500,000 = 3.80% margin on a 4.0% markup. Kestrel knows this. It is why the fee negotiation was about 4.0% and not "four points of profit," and why Nadia Haddad asks for the margin number, not the markup number, in every go/no-go meeting — Chapter 15 walks that meeting.

🔄 Check your understanding. A subcontractor's change-order proposal shows $84,000 of cost and says "plus 15% overhead and profit — $12,600, total $96,600." Your contract allows a maximum 15% markup on self-performed change-order work. Is their number right, and what is their actual margin?

12.9.4 Self-perform versus subcontracted — and why you need both numbers

Kestrel self-performs concrete, rough carpentry, and general trades, and subcontracts everything else. The estimating approach is genuinely different for each:

And here is the rule that surprises people: build an independent number for every major subcontracted trade anyway. Not a full detailed estimate — a defensible order-of-magnitude number from quantity and a historical unit rate. Kestrel does this for steel, curtain wall, mechanical, electrical, and drywall on every job.

Why spend the hours when you have quotes?

Because a quote is only meaningful against an expectation. If your independent number for drywall is $1.9M and the low bid is $1.35M, you have learned something urgent: either that bidder found an efficiency you should understand, or they missed scope, or they are buying the job. All three of those are things you need to know before award, and none of them are visible from a bid tab alone. If your independent number is $1.9M and the low bid is $1.87M, you have learned something else — that the market agrees with you, which is quiet, valuable confirmation that your documents are clear.

A cluster of bids close to each other and far from your number means your number is wrong. A single bid far below a cluster means that bid is wrong. That diagnostic — the bid-tab spread as an instrument — is worked in full in case-study-02.md, and it is one of the most useful tools in preconstruction. Chapter 16 turns it into a buyout process.

12.10 Where Estimates Go Wrong — and the Document That Prevents It

12.10.1 The error list, with money attached

Every one of these is a real failure mode. The dollar figures are illustrative but they are the right order of magnitude for a project of Northgate's size.

Look at the last two rows for a moment. Neither is a measurement error. Both are failures to price time — which is theme two of this book stated as a bookkeeping fact: the schedule and the budget are the same conversation. An estimator who does not have the schedule open is estimating half the job. This is exactly why Kestrel's preconstruction process runs the estimate and the CPM schedule together rather than in series; Chapter 14 builds the schedule side of that conversation.

Note also what unites the top two rows: nobody made an arithmetic mistake. Every trade priced its own scope correctly. The money fell into the space between scopes — which is where money always falls, and which no amount of careful measuring will find. Only a scope-by-scope reconciliation finds it.

12.10.2 The pre-bid estimate review

Kestrel runs this before every number leaves the building. It takes two people about ninety minutes on a job this size and it has paid for itself more times than anyone has counted.

Quantities 1. Does every quantity cite the sheet and detail it came from? 2. Was every quantity checked by a second method — count against schedule, area against $/SF, tonnage against SF of building? 3. Are the units labeled on every line? (Yes, every line.) 4. Have all addenda been incorporated, and has the drawing log been reconciled to the issued-for-bid revision of every sheet? 5. Do the quantities from the enlarged plans, details, and schedules agree with the overall plans?

Unit costs 6. Is every labor rate burdened, and does the burden use current rates from accounting? 7. Is the workers' comp class code correct for each classification in each crew? 8. Does every production rate have a stated basis and stated conditions? 9. Has waste been applied exactly once — in the quantity or in the price, not both? 10. Is freight in? Is sales tax in, and is the treatment correct for this jurisdiction and project type?

Subcontracts and quotes 11. Has every quote been read to the end, including the exclusions on the second page? 12. Does the bid tab compare scope-corrected numbers rather than face amounts? 13. Is there an independent in-house number for each major trade, and has any large variance been explained? 14. Is every quote logged with its validity date, and is escalation carried for anything bought after that date?

Indirects and markups 15. Do the general conditions match the current schedule duration, and were they re-run after the last schedule change? 16. Is contingency tied to a risk register with named risks? 17. Is escalation applied to the correct scope and the correct window? 18. Is the bond computed on the final price, using the divisor method? 19. Is the markup the markup that produces the margin you intend?

The document 20. Is there a basis-of-estimate page, and could a stranger price this job from it?

12.10.3 The basis of estimate — the threshold concept made concrete

This is the deliverable that §12.2 was about. The basis of estimate is not documentation of the estimate. It is the estimate. The number is the byproduct.

A basis of estimate has six sections, and none of them is optional:

1. Quantities and their source. Every major quantity, the sheets and details it came from, the drawing revision, and the method — hand takeoff, on-screen, model extraction (with the LOD), or a factor. Where a quantity is derived rather than measured, show the derivation.

Northgate's gypsum board is the perfect illustration, because 412,000 SF is a derived number and it must be shown as one:

   Partition length (S/A-series plans, floor by floor) ...  18,600 LF
   Average height to underside of deck or 6" above ceiling,
     per the partition schedule (WEIGHTED — assumption) ....  10.5 ft
   Two faces  ...........  18,600 × 10.5 × 2 ..............  390,600 SF
   Second layer, rated corridor and shaft walls
     (2,400 LF × 10.5 × 1 additional face)  ...............   25,200 SF
   Subtotal  .............................................   415,800 SF
   Deduct openings over 50 SF (doors, borrowed lights,
     per office convention — ASSUMPTION, state it)  .......  (18,900) SF
   Add soffits, exterior-wall furring, ceiling-height
     changes (not included in the 18,600 LF)  .............   15,100 SF
   ------------------------------------------------------------------
   TOTAL GYPSUM BOARD  ...................................  412,000 SF

Three of those six lines are assumptions, not measurements. The 10.5-foot weighted average height. The 50-square-foot deduction convention. The furring and soffit allowance. Change the average height by six inches and the total moves by about 18,600 SF — roughly $78,000 of installed board. A number like 412,000 SF is not a fact about the building. It is a fact about a set of decisions, and if you do not show the decisions, nobody — including you, in four months — can check them.

2. Unit costs and their source. Company historical data (which jobs, which years), published data (which source, which location factor), or a quote (whose, dated when, valid until when).

3. Assumptions. Access, sequence, crew sizes, shift, weather allowance, working hours, laydown availability, hoisting, site conditions, soil type, existing utility locations. Anything you had to decide because the documents did not say.

4. Inclusions. What is in the number. Be specific enough that a reader can check it against the specification.

5. Exclusions. What is not. Be specific enough that a reader cannot mistake the boundary. "Excludes hazardous material abatement" is an exclusion. "Standard scope per plans and specs" is not.

6. Risks priced and not priced. Which risks the contingency covers, which are excluded, and which are carried by allowance. This is the section that most estimates skip and every claim needs.

12.10.4 The ethical line: a qualified bid versus a hidden-exclusion bid

Every bid excludes something. That is not the issue. The issue is whether the reader can find the exclusions.

A qualified bid states its exclusions and clarifications on the face of the bid form, or on a clearly labeled attached page, in plain language, at the top level. The owner reads it, prices the gap, and compares bidders on equal footing. This is professional, it is expected, and on a complex project it is a sign of competence.

A hidden-exclusion bid buries the same exclusion where a busy reader will not find it — in the fine print of an attached vendor quote, in a clause that reads as boilerplate, in language ambiguous enough to defend both ways, or nowhere at all, with the exclusion existing only in the estimator's head as a change order they intend to write later.

The test is one sentence, and Tomás applies it to every bid Kestrel submits:

Would the owner, reading this bid for five minutes, know what they are not buying?

If the answer is no, you have not qualified your bid. You have set a trap. And it is worth being clear-eyed about why the trap is a bad business decision as well as a bad ethical one:

• It does not survive contact with a competent owner. Pri Sethi has been an owner's rep for eleven years. She reads the second page.
• It converts your project team's first ninety days into an argument. The change order you planned to write arrives on a job where the owner now believes you are not straight with them, and every subsequent legitimate change gets fought.
• It is a reputational asset you can only spend once. Owners talk to each other. Bid lists are short. Nadia Haddad's standing instruction at Kestrel is that the company will lose a job over a qualification before it will win one by hiding a number — not because she is sentimental, but because a place on a negotiated bid list is worth more than any single project's margin.

⚖️ What the contract says. Where you put the exclusion also determines whether it is legally operative. Most public bid forms are unforgiving: a bid that "qualifies" or conditions its price may be rejected as non-responsive, which is why on hard-bid public work like Rivermont Elementary #12 the mechanism for resolving an ambiguity is a pre-bid RFI and an addendum, not a qualification on your bid form. On a negotiated or CM-at-Risk procurement, clarifications are normal and expected. Know which kind of procurement you are in before you write a single exclusion, because the same sentence that demonstrates professionalism on one job disqualifies you on the other. Chapter 15 covers responsiveness, bid errors, and withdrawal in detail.

🪞 Learning Check-In

Stop here for two minutes. This one is worth doing honestly.

There is a specific and dangerous feeling that this chapter produces: the feeling of having followed the arithmetic. Every calculation in §12.7 was legible. You could see where each number came from. Nothing was hard. That feeling is not the same as being able to produce those numbers from a blank sheet, and the gap between the two is where most people's estimating education quietly stalls.

Test yourself against these four, and be strict:

1. Without looking back, write down the unit-cost formula and name what each of the three terms depends on. Which one does not change when the crew slows down?
2. Build a burdened rate for a $30.00/hr worker with a 45% burden factor. Now do it the hard way — from a $30.00 base with $4.85 health and welfare, 4% retirement, 5.5% vacation, 10.55% payroll taxes, 9.2% comp, 1.9% GL, and 2.5% small tools. Did you get the same answer? (You should get about $45.66, a 1.522 factor — so the "45% burden" shortcut was 2% light.)
3. A crew forms 465 SFCA a day. Production drops 15%. Does the unit cost rise by 15%? Explain the answer in one sentence without doing arithmetic.
4. You mark a $2,000,000 cost up by 12%. What is your margin? (10.71%.)

How to read your results. If you got all four, you have the mechanics and should now practise on work you have never priced. If you got the formula but not the burden build-up, you understand the structure and need the details — re-read §12.7.1 and do the 📋 Try it drill again with different wages. If question 3 required arithmetic, go back to the 🔍 Why this works box in §12.7; the asymmetry between time-based and quantity-based costs is the concept that makes the rest intuitive. And if you found yourself thinking "I understood it when I read it" — that is exactly the feeling this box exists to interrupt. Understanding a worked example and being able to generate one are different skills, and only the second one gets you paid.

Spaced Review

From Chapter 11 — conceptual estimating. Before reading on: at what point does a square-foot estimate stop being defensible, and what replaces it?

A conceptual or parametric estimate — $/SF, $/bed, $/parking stall — is appropriate when the design is too incomplete to measure, and its accuracy is bounded by the completeness of that design, not by the estimator's skill. Northgate's $360/SF is a check, not an estimate. The transition happens when there are enough drawings to measure real quantities, and what replaces the parametric number is exactly what this chapter built: quantity × unit cost, assembled from crews, rates, and production. Notice that the two are not rivals. You used the parametric method inside the detailed estimate in §12.4.5, when you converted 985 tons of steel into $36.19/SF to see whether the answer was plausible. Conceptual methods do not go away; they become your error-detection system.

From Chapter 8 — swell and compaction. Without looking: what does 44,000 CY of bank-measure cut become in the back of a truck, and what does 12,000 CY of that cut become once it is placed and compacted?

At 25% swell, 44,000 BCY becomes 55,000 LCY of loose material to handle. At 12% shrinkage, 12,000 BCY of reused cut yields only 10,560 CCY of compacted embankment — which is why §12.4.1 ended up importing 1,636 BCY of borrow at about $27,000. Bank, loose, and compacted are three different numbers for the same dirt, and the takeoff sheet must say which one it is reporting.

Deep callback to Chapter 7 — where quantities live. Recall the order-of-precedence principle: specifications govern quality and product; drawings govern quantity and location. That principle is the whole reason this chapter is structured the way it is. You measure quantities off the drawings, but you cannot price them until the specification tells you what you are measuring — which gauge of stud, which class of concrete, which reinforcing grade, whether the wall is rated. Dani's nine missing footings were a drawings problem (they were on a sheet nobody opened). The 412,000 SF of board is a specification problem (the partition schedule defines heights, layers, and ratings). Every estimating error is one or the other, and knowing which kind you are looking at tells you where to go find the answer.

Project Checkpoint: The Willow Street Quantity Takeoff

In Chapter 11 you produced a conceptual estimate for the Willow Street Community Center at a dollar-per-square-foot rate, along with a value-engineering log of eight priced options. That number was a judgment about the kind of building this is. Now you measure the actual building.

Your deliverable: a documented quantity takeoff of the Willow Street sitework, concrete, masonry, and structure, plus a basis-of-quantities page.

Work from the package in Appendix K — 24,000 SF, two stories, wood-framed second floor over a structural steel and CMU first floor, on a flat 2.1-acre site with one 8-inch water main to relocate.

Take off, in construction order:

Apply waste factors from §12.5, and show the factor you used on every line. Where you deviate from the table, say why.

For every single quantity, record four things: the number, the unit, the sheet and detail it came from, and the method (scaled, schedule-extended, counted, calculated, or factored). One row per quantity. If a row is missing a source, it is not finished.

Then write the basis-of-quantities page — one page, six sections, following §12.10.3. List every assumption you made: average wall heights, opening-deduction conventions, whether the topping is in the concrete or the steel package, what soil type you assumed for trench sloping, and which quantities are derived rather than measured. Aim for at least twelve stated assumptions. If you have fewer than eight, you have not been honest with yourself about how many decisions you made.

A discipline worth adopting now: highlight, in a different color, every quantity you are least confident in. On a real bid that list becomes the contingency conversation.

Next chapter you put prices on all of it. In Chapter 13 this takeoff becomes a full detailed estimate organized by CSI division, with subcontractor bid tabs, indirect costs, markup, and a final number — and the quality of that estimate will be capped, exactly and unforgivingly, by the quality of the takeoff you do this week.

Chapter Summary

The four formulas. Everything in this chapter reduces to these.

   Volume from area:      CY = SF × (thickness in inches ÷ 12) ÷ 27

                                 crew cost per hour
   Unit cost:             UC = ---------------------- + material/unit + equipment/unit
                                production per hour

   Markup ↔ margin:       markup = margin ÷ (1 − margin)
                          margin = markup ÷ (1 + markup)

   Circular percentages:  Price = Subtotal ÷ (1 − rate)      [bond, gross-receipts tax]

The decision framework — what to do when you are handed a scope you have never priced.

The nine things most likely to be wrong in your estimate right now:

1. A quantity that came from one sheet when it lives on two.
2. A schedule row read across the wrong line.
3. A production rate borrowed from a job with different access.
4. Waste applied twice, or not at all.
5. A unit conversion — bank/loose, SF/SY, ton/tonne, CWT/ton.
6. A quote whose exclusions you did not read to the end.
7. A markup that produces a margin you did not intend.
8. General conditions that do not match the current schedule duration.
9. An assumption that lives only in your head.

And the one sentence to keep: your product is not the number; it is the documented set of decisions the number rests on. The number will be wrong. The decisions can be right, defended, managed, and — when the drawings change — repriced.

What's Next

In Chapter 13 you build the full detailed estimate: every CSI division, subcontractor quotes reconciled on a bid tab, scope sheets that close the gaps between trades, indirect costs, markup, and a number you would sign. Then Chapter 14 does the other half of the same conversation — the critical path method — because the general conditions line you just estimated is a function of a duration nobody has calculated yet. Estimating and scheduling are one activity performed by two people; Chapter 14 is where they meet.