Quantity Takeoff Example, Worked Step by Step

What scope are we taking off?

The example is a single-story commercial shell on a level site, 60'-0" by 40'-0" out-to-out of the foundation wall. The structural set shows continuous strip footings around the perimeter, four interior column pads, and a 5-inch slab-on-grade reinforced with #4 bars at 18 inches on center each way. The foundation wall above is 8-inch CMU.

The scope boundary is CSI Division 03 only — cast-in-place concrete and its reinforcing. The CMU wall is Division 04 and is not in this takeoff. Neither is the excavation, the backfill, the column base plates, or the testing. That sounds obvious until you are two hours into a sheet and quietly measuring wall block because it was on the same drawing.

Four sheets do the work: the foundation plan for layout and dimensions, the typical footing section for the cross-section, the slab detail for thickness, reinforcing and vapor barrier, and the structural general notes for lap length and clear cover. The first real decision in a takeoff is not a measurement. It is writing the boundary down before you start, so the sheet does not drift.

Step 1: What do the dimension strings actually reference?

Out-to-out, face of wall, or centerline?. On this plan the 60'-0" and 40'-0" strings run to the outside face of the 8-inch foundation wall. That single fact sets every other number on the sheet. Had they run to the centerline of the wall, the footing perimeter and the slab area would both change, and nothing downstream in the takeoff would catch it. Find the datum before you find the scale.

Check the section before you trust the plan. The typical footing section shows a 2'-0" wide by 1'-0" deep strip footing, centered under the wall. Centered is something you read off the section, not something you assume. An eccentric footing — routine where a wall sits near a property line — shifts the footing centerline off the wall centerline, and the perimeter you measure changes with it.

The numbers that are not on the plan. Lap length, clear cover, and bar spacing tolerance live in the structural general notes, not on the foundation plan. On this set the notes call laps at 40 bar diameters and 3-inch clear cover at slab edges. Both change the reinforcing quantity by a measurable amount, and both are the first things a reviewer will check.

Step 2: Why measure footings along the centerline?

The wall is 8 inches thick and centered on the footing, so the footing centerline sits 4 inches — 0.333 ft — inside the outside face of the wall. The centerline rectangle is therefore 60 − 0.667 = 59.333 ft by 40 − 0.667 = 39.333 ft.

Perimeter: 2 × (59.333 + 39.333) = 197.33 LF. Volume: 2.0 ft wide × 1.0 ft deep × 197.33 LF = 394.7 CF, and 394.7 ÷ 27 = 14.62 CY.

Centerline is not a preference. At each corner the two footing runs overlap, and measuring four sides at the building line — 60 + 40 + 60 + 40 = 200 LF — counts every corner twice. Running the centerline resolves all four corners in one operation with no deductions to remember.

Here is the honest part: on this building the difference is 200 LF against 197.33 LF, about 1.3 percent, or roughly a fifth of a cubic yard. It disappears inside the allowance you will add later. On a 400-foot perimeter with 4-foot-wide footings the same mistake is worth roughly 16 LF of footing — at 1 ft deep that is about 2.4 CY, and on a deeper footing it is enough to change what you order. Build the habit on the job where being wrong costs you nothing.

The four interior pads are straightforward: 4'-0" × 4'-0" × 1'-0" = 16 CF each, 64 CF for four, 64 ÷ 27 = 2.37 CY.

Step 3: Why 2,400 square feet is not the slab

The architect calls this a 2,400 SF building, because 60 × 40 = 2,400. The slab is not 2,400 SF. It pours to the inside face of the foundation wall. This takeoff carries the 8-inch CMU at its nominal 8 in rather than its 7-5/8 in actual width, which is noted on the assumptions page; the difference is under an inch per side and is stated rather than hidden.

Inside dimensions: 60 − 2(0.667) = 58.667 ft, and 40 − 2(0.667) = 38.667 ft. Area: 58.667 × 38.667 = 2,268 SF.

That is 132 SF less than the footprint — about 5.5 percent. On the slab line alone, taking the architect's number would have bought roughly 2.0 CY of concrete you never place: 132 SF × 5/12 ft = 55 CF, and 55 ÷ 27 = 2.04 CY. Volume: 2,268 SF × 5/12 ft = 945.2 CF, and 945.2 ÷ 27 = 35.01 CY.

Four columns penetrate the slab, with base plates of roughly a square foot each. I did not deduct them. Four square feet on a 2,268 SF slab is well inside the noise of the pour, and chasing it costs more time than it saves. But the decision goes on the assumptions page, because "did not deduct" and "forgot to deduct" look identical on a spreadsheet, and only one of them survives a question.

The vapor barrier under the slab starts from the same 2,268 SF. Laps and turn-up are handled in Step 5, where they belong.

Step 4: How do you count reinforcing bars correctly?

Bar count is spaces plus one. Divide the effective width by the spacing to get the number of spaces, then add a bar. Dropping the plus-one is an easy error to make, and it is invisible in the total.

Step 5: Why is waste arithmetic rather than a percentage you pick?

Footing concrete follows the bucket, not a waste factor. Neat line on a 24-inch footing is 14.62 CY. Cut the trench with a 30-inch bucket and you place 30 ÷ 24 = 1.25 times the neat volume, or 18.28 CY. The overrun is proportional to trench width, and trench width is a question you can put to the excavation sub in one phone call. A footing cast against formed sides overruns far less, but not zero: form deflection, an uneven bottom under the forms, and spillage all add concrete. Carry a stated allowance on formed work — this takeoff uses 3 percent on the pads — rather than neat line. If nobody will answer, carry a stated trench width on the assumptions page and let the GC argue with a dimension instead of a percentage.

Slab overpour follows the subgrade. The slab is 5 inches. If the subgrade runs a quarter inch low on average you place 0.25 ÷ 5 = 5 percent more concrete: 36.76 CY instead of 35.01. Two caveats. High spots are trimmed, not credited, so overrun follows the low side of the distribution rather than its mean. And against the plus-or-minus 3/4 inch fine-grade tolerance in ACI 117, a quarter inch average is a well-graded subgrade — carry an overpour allowance from your own placement history on comparable subgrade, and tighten it only where the subgrade has been proof-rolled and fine-graded to a stated tolerance. Double the slab thickness and the same subgrade condition costs half the percentage. That is precisely why one blanket waste factor applied across different slab thicknesses is wrong — the physical cause scales with depth and the percentage does not.

Sheet goods follow the lap. A 6-inch side lap on a 20-foot-wide roll costs you coverage, not material, so the factor is 20 ÷ 19.5 = 2.6 percent — and 5.3 percent on a 10-foot roll. End laps where a roll runs out, cuts at column blockouts, and seam tape per ASTM E1643 all add on top of that; carry 5 percent unless you have laid out the roll pattern. The perimeter turn-up is a separate line again: the slab edge runs 2 × (58.667 + 38.667) = 194.67 LF, and at 6 inches up the wall that is 97 SF. So 2,268 × 1.05 = 2,381, plus 97, is about 2,480 SF.

The blanket percentage would have landed close here, and that is the trap. Neat line across the package is 52.0 CY. Item by item it is 57.4 CY — about 10.4 percent — but only under one specific set of assumptions: strip footings trench-cast with a 30-inch bucket, pads formed at 3 percent, slab at 5 percent for subgrade. Form the strip footings instead and the same package falls to roughly 54.5 CY while the formwork line climbs from 64 to 459 SFCA. The concrete allowance and the formwork assumption are a single decision recorded in two places, and they have to move together. A flat 10 percent would have produced almost the same total on this job while being wrong on every line inside it. On the next job, with a shallower slab or a narrower trench, the coincidence does not repeat and there is nothing in the sheet that tells you why.

What quantities can the drawings not give you?

Formwork is measured as contact area — SFCA, the surface the form actually touches, not the volume behind it. The four pads are simple: 16 LF of perimeter × 1 ft deep = 16 SFCA each, or 64 SFCA total.

The strip footings are not simple. Formed on both sides, they are 197.33 LF × 1 ft × 2 = 395 SFCA. Cast directly against the trench walls, they are zero.

Nothing on the drawings decides which. It is a means-and-methods call belonging to the concrete sub, and it swings the formwork line from 64 SFCA to 459 SFCA — a factor of seven on a line item that carries real labor.

The honest move is to state the assumption on the takeoff and flag it as an assumption, not to pick one silently and hope. This is a common reason two competent takeoffs of the same drawing set disagree, and it is not a measuring error on either side.

The same category catches you elsewhere: whether the pads pour monolithic with the slab or separately, whether a mud slab goes in first, whether the footings step to follow grade. Thickened edges and turndowns at door openings are a different animal — they are structural, they carry their own reinforcing, and if the set does not show them the answer is an RFI to the engineer, not an assumption on your sheet. All of it is decided off the drawings, in the field or in a submittal, after the takeoff is done.

Step 6: How should the takeoff be presented?

A takeoff nobody can check is a takeoff nobody will trust. The presentation is part of the work, not packaging applied afterward.

• Quantities in the units the trade actually buys in — concrete in cubic yards, reinforcing in linear feet and tons, formwork in square feet of contact area, vapor barrier in square feet, joints in linear feet (194.67 LF of perimeter isolation joint and roughly 330 LF of sawcut control joint here, at a 12-foot grid), slab edge form in linear feet, and finishing and curing in square feet. A Division 03 takeoff that reports only concrete, steel, form and barrier has left four priced lines off the sheet.
• Neat-line quantity and as-placed quantity in separate columns, so the reader can see the allowance and overwrite it with their own
• Sheet and detail reference on every line, so a disputed number can be checked against its source instead of re-measured from scratch
• An assumptions page: trench width, subgrade tolerance, lap length, formed versus trench-cast, and what was not deducted and why
• An exclusions page: excavation and backfill, CMU, base plates, testing, and anything the drawings deferred to a later submittal
• A marked-up plan set showing what was measured, in color, so the scope boundary is visible rather than described
• An editable file. A PDF that cannot be adjusted when the first addendum lands is a document, not a takeoff

What does this takeoff deliberately not do?

It carries no prices. A quantity takeoff stops at quantities. Turning 57.4 CY into a dollar figure requires a mix design, a delivery radius, a placement method, a crew rate, and a pour date — none of which appear on a structural drawing set, and all of which change the answer.

It is also not an estimate. An estimate adds labor, equipment, general conditions, overhead, markup, and a position on risk. This sheet is the input to that work, not a substitute for it. Treating a takeoff as a bid is how contractors end up defending a number they never built.

And it is not difficult. On a scope this size, a contractor with a scale, a calculator, and a couple of uninterrupted hours can work through the same sequence and land close to these numbers. If your jobs look like this one, you do not need to outsource the takeoff. You need to stop doing it at ten at night after a full day in the field, which is a scheduling problem rather than a measuring one.

Where outsourcing genuinely changes the math is a different shape of work: multi-trade packages where the takeoff runs days rather than hours, bid weeks where four sets are due and you have time for one, and drawing sets incomplete enough that the assumptions page matters more than the measurements. Outsourced estimating exists for that. It does not make a 2,268 SF slab any easier to measure.