Sharp Pencil

Sharp Pencil › Guides

How to price a TPO roof

A complete walkthrough, from reading the plan set to the number you hand the GC. Written for people who actually do this, so it includes the parts that go wrong.

Start with what the spec says, not what you assume

Before any measuring, find Division 7 and read it. You are looking for six things that set the entire assembly:

The warranty trap. A lot of estimators assume a 20-year NDL requires a coverboard. Read the manufacturer's actual requirements. Coverboard is usually driven by hail riders, EPS substrates, or terms past 30 years, not by NDL itself. The real gate on an NDL is installer certification level, and some manufacturers will not issue one at all to a base-tier contractor.

The takeoff, in the order that avoids mistakes

Field area

Get it from printed dimension strings wherever they exist. Scaling is a fallback, not a method. Break the roof into sections if levels or systems differ, because a single blended area hides a lot of money.

Perimeter and heights

Perimeter comes off the plan. Heights do not. Parapet and wall heights live on the elevations and sections, and this is where estimators lose the most money: they assume a parapet height, price base flashing to it, and find out in the field it was four feet taller for half the run.

Drainage

Count drains, overflows, scuppers, and gutters. Then reconcile against the plumbing sheets, because interior drains appear on both and the counts do not always agree. A roof with zero interior drains and a perimeter gutter is a finding worth writing down, not an absence to skip past.

Penetrations and curbs

VTRs, gas lines, conduit, equipment curbs. Mechanical and plumbing sheets, not the roof plan. Count what the tags say. A heat pump tagged HP-1 is not an RTU, and pricing it as one is a real error.

The downspout trap. Downspout count comes off the plan. Downspout length does not. You need the vertical run, which means reading top-of-wall minus grade off the elevations. Twenty-two downspouts on a 33-foot wall is roughly 730 linear feet of metal that a plan-only takeoff prices at zero. Watch for grade ramps and bump-outs that make some runs shorter.

Material, in the units you actually buy

The bridge from quantity to cost is the coverage rate, and this is where sloppy math compounds:

ItemHow it convertsWatch for
MembraneRoll coverage, minus seam overlapWaste factor differs by roof complexity, not a flat 10%
InsulationBoards per square by thicknessMultiple layers mean staggered joints and more boards
Fasteners and platesPattern per square by wind zonePerimeter and corner zones are denser than field
AdhesiveCoverage per unit by substratePorous substrates drink more than the data sheet says
Bonding adhesiveCoverage per unit, two surfacesBoth sides get coated, so it is half the stated rate

Fastener length is derived, not chosen

The fastener has to reach through the full insulation stack plus the required embedment into the deck. So the fastener length follows the stack thickness, which follows the R-value, which follows the spec. Change the insulation and the fastener line changes with it. And if the job has taper, the average stack thickness changes across the roof, which means the fastener schedule and the freight both move.

Labor, from production rates you can defend

The single most common error in roofing labor is confusing crew-day rates with man-day rates. If your notes say "we do 20 squares a day" with a six-man crew, that is roughly 3.3 squares per man-day, not 20. Price it as 20 per man-day and your labor is off by a factor of six.

Build labor from three inputs:

  1. Production rate in squares per man-day for the specific system and condition. Tear-off is slower than recover. Adhered is slower than mechanically attached.
  2. Crew size and burdened hourly rate. Burdened, meaning wages plus taxes, insurance, and workers' comp, which in roofing is not a rounding error.
  3. Flashing and detail labor separately, usually in linear feet per man-day, because a roof with 40 penetrations and 900 feet of parapet is not the same job as an open field of equal area.

The lines people forget

Markup, and why the split matters

Overhead and profit are not one number. Overhead covers what it costs to keep the doors open, and it applies to your direct costs. Profit is what the risk is worth. Blending them into one percentage is fine on simple jobs and dangerous when subcontracted work is involved.

Here is why: if a job includes $40,000 of subcontracted taper and you apply a full blended markup to it, you have priced your own risk on money that passes straight through. Competitors who break it out will beat you on that line every time. Carry sub costs separately with their own markup.

Before you send it

Send us the next one

Your first takeoff is free, on a real live bid. Full quantity takeoff, RFI register, material list, and a client-ready proposal in 48 hours.

Upload a plan set