How to Take Off Quantities From PDF Drawings (Without CAD)

Short answer: Open the PDF, calibrate the scale against a dimension printed on the drawing rather than the stated ratio, measure in the unit the item is priced in, record deductions explicitly, and keep wastage out of the measurement. Most takeoff errors are not measuring errors — they are unit, scale and revision errors.

Takeoff is measurement. Estimating is money.

The single most useful discipline in this job is refusing to mix the two.

Takeoff answers how much: 46 linear metres of skirting, 112 square metres of plasterboard, 8 nr doors. Estimating answers what it costs: the rate per metre, the labour constant, the waste allowance, the margin.

Firms that blend them end up with quantities that cannot be checked and rates that cannot be reused. Firms that separate them can reprice a whole job in an afternoon when a merchant puts prices up, because nothing needs remeasuring.

Step 1: get the right drawing, and prove it

Before any measuring:

• Check the revision. The revision letter and date live in the title block. Measuring rev C when rev E is issued is the most expensive mistake on this list, and it is invisible until someone else prices rev E.
• Check the sheet is complete. Plans that have been cropped to a detail lose their title block and their dimension strings, which are the two things you need.
• Prefer a vector PDF over a scan. A PDF exported from CAD has real geometry. A photographed or scanned drawing has been through a lens or a roller, and may be very slightly stretched along one axis.

Step 2: calibrate — do not trust the printed scale

This is the step people skip, and it is the one that determines whether everything else is right.

A drawing labelled 1:50 is only 1:50 if the sheet you are holding is the size it was drawn for. In practice, drawings get exported A1 to A3, "scaled to fit" by a print driver, or dropped into an email as an image. The scale note in the corner does not update when that happens.

So calibrate:

Find a printed dimension on the drawing — an actual number written on a dimension string, ideally the longest one available.

Measure that same span with your measuring tool and tell it the true distance.

Now find a second, independent printed dimension elsewhere on the sheet and measure it. If it comes back correct, your calibration holds. If it does not, the sheet is distorted or you are looking at two different scales on one page.

That second check is not optional. It is the only thing that catches a stretched scan or a detail inset at a different scale — and details at a different scale on the same sheet are extremely common.

Calibrate on the longest dimension you can find. A 1 mm reading error on a 500 mm span is a 0.2 per cent error on everything you measure afterwards. The same 1 mm error on a 10,000 mm span is 0.01 per cent.

Step 3: measure in the unit the item is priced in

The unit is part of the measurement. Getting the number right in the wrong unit produces a confidently wrong price.

The recurring trap is the mixed item. A partition is measured in m², but its head and base track are metres and its studs are a count driven by centres. If your rate is a composite per m², measure m² only. If your rate is built from components, measure the components. Mixing the two double-counts.

Step 4: deduct openings — and say that you did

Every trade has its own convention on deductions, and every convention is defensible right up until nobody records which one was used.

The rule that survives scrutiny: measure gross, list deductions separately, show the net. So a wall is 24.6 m² gross, less 1.9 m² for the door, less 2.4 m² for the window, net 20.3 m². The next person can see what you did and disagree with the convention rather than with the arithmetic.

Where the trade convention is not to deduct small openings — many finishes rates ignore anything under a set area because the labour is unchanged or worse — write the convention down next to the measurement. "Openings under 0.5 m² not deducted" costs one line and settles an argument later.

Step 5: keep wastage out of it

Measure the net quantity. Apply wastage at the rate.

A tiled area is 112 m². It is not 123 m². Ten per cent wastage on tiles is a pricing decision that belongs against the tile rate, where it is visible, adjustable, and different for a mosaic than for a 600 × 600 porcelain.

Bake it into the measurement and three things go wrong: nobody downstream knows whether the quantity is net or gross, a variation gets remeasured against a different basis, and when the client asks why 112 m² of floor needs 123 m² of tile, there is no answer in the paperwork.

The four errors that actually cost money

In order of how often they bite:

Wrong revision. You measured a superseded drawing. The fix is procedural: record the revision against every measurement, and re-check the revision before submitting.

Uncalibrated or singly-calibrated scale. Everything is out by a constant percentage, which is invisible because it looks internally consistent. The second calibration check catches it.

Detail insets at a different scale. A 1:50 plan with a 1:10 detail in the corner. Measure the detail with the plan's calibration and you are out by a factor of five.

Unit drift between takeoff and rate. Measuring a linear item that is priced per m², or measuring in millimetres against a rate built per metre. Always the same symptom — a price out by exactly 1,000 or exactly the wrong shape.

A fifth, less common but worse: measuring an existing-condition survey drawing as though it were the proposal. Refurbishment sets frequently contain both, drawn identically, distinguished only by the sheet title.

Where software actually helps

Digital takeoff tools do three things a scale rule cannot.

They remember. Every measurement stays attached to the drawing, at the coordinates where it was taken. When the drawing is revised, you can see what you measured and where, rather than starting again.

They keep the audit. Who measured it, when, on which revision, with what calibration. That matters far more on a remeasured contract or a disputed variation than it does on the original price.

They can propose the estimate lines. This is the newer, more useful capability. Once the geometry is measured and the item type is known, the tool has enough to suggest the estimate lines the measurement implies — the area, the unit, the item — leaving you to check and price a proposed list rather than typing it out from scratch. Human judgement stays where it belongs, on whether the list is right and what it should cost.

ScopeKit works this way: measurements taken off a drawing propose the estimate lines they drive, so measurement and pricing stay linked rather than being two disconnected exercises in two different files. If a measurement changes, the line it drives is traceable back to it.

A workable order of operations

For a typical refurbishment or fit-out package:

Log the drawing set and revisions.

Calibrate each sheet, and check with a second dimension.

Measure by element, not by room — all the plasterboard, then all the skirting. Fewer tool changes, fewer unit slips.

Record deductions and conventions inline.

Total by item and unit.

Hand the net quantities to pricing, where wastage, labour constants and margin get applied.

Keep the measured basis. When the variation arrives, you remeasure the difference instead of the whole job.

Related reading

• How to price a building job — what happens to these quantities once they reach the rate.
• How to write a construction quote — presenting the priced result so it gets accepted.
• Dayworks explained — pricing the work that cannot be measured in advance.
• Takeoff and measurement in ScopeKit — measuring on the drawing and turning it into estimate lines.