How to Price Drywall Jobs: Per-Sheet and Per-Square-Foot Rates With AI

Price drywall per sheet or per square foot of board when the plans give a firm quantity, and by the hour for repairs, patches and jobs with unknown access. Whichever unit you pick, build it from your crew's cost per hour and your own hanging and finishing times, then test it with AI against past jobs before quoting.

The unit matters less than what sits under it. A per-sheet rate that works on a house of plain walls loses money on a basement with a full ceiling, because ceiling sheets take longer to hang and finish, and a Level 5 skim coat adds more again. AI is quick at the arithmetic, if you check its sums, and at spotting that pattern in your old jobs. The hours per sheet have to come from your crew.

Follow me on Instagram@sagnikteaches

Every cost, hour and price below is invented for illustration: a two-person crew, one 61-sheet basement and a handful of past jobs. Swap in your own figures before relying on any result.

Connect on LinkedInSagnik Bhattacharya

Which unit fits which kind of drywall job

A standard sheet is 4 ft wide and most often 8 ft long, so it covers 32 square feet. Longer sheets such as 12 ft (48 square feet) cut down on end joints, and 1/2 inch and 5/8 inch are the common thicknesses. That makes per sheet and per square foot of board the same method in two units: $88 a sheet is $2.75 a square foot. They only come apart when someone quotes per square foot of floor area, which is where the basement example goes wrong later.

Subscribe on YouTube@sagnikteaches

My rule: per sheet or per square foot of board when you have measured the job and can state the finish level, by the hour when you cannot see what is behind the problem. Even for new work I would build the price from hours underneath, as the next sections do, and only then express it per sheet. The unit is how the customer reads the price. The hours are how you know it works.

Start with what one hour of your crew costs

Everything else multiplies this number, so settle it first. Add wages, the on-costs on top of them (payroll taxes, insurance, benefits) and a share of yearly overhead (van, tools, insurance, software, accounting), divide by the hours that actually get worked on jobs, then add your margin. The fuller build is in the tutorial on working out a tradesman's hourly rate. For drywall, this is the prompt I would paste into ChatGPT or Claude:

I run a two-person drywall crew. Work out our cost and our selling price per person-hour.
Inputs:
- Lead hand wage $34 an hour, helper $24 an hour
- On-costs (payroll taxes, insurance, benefits): 22% on top of wages
- Yearly overhead (van, tools, insurance, software, accounting): $26,400
- Productive hours per person per year: 1,320 (leave out travel, quoting, weather days, holidays)
- Target margin: 25% of the selling price
Show every step as a formula with the numbers in it.
Selling price = cost / (1 - margin), not cost x 1.25.
State any assumption you had to make.

Here is the table I would expect back (illustrative):

Two traps sit in that table. The first is the hours. If you leave them out, an assistant has to assume some, and the obvious assumption is 2,080 paid hours a year (52 weeks of 40). That drops overhead to $6.35 an hour and the selling price to $55.63, about 8% light, on every job you ever quote from it. The second is margin against markup. Adding 25% to $45.38 gives $56.73, and the profit on that price is only 20% of it. To earn 25% of the price, divide by 0.75.

Hours per sheet: take them from your own jobs

A job log with five columns is enough: job, sheets hung, how many were ceiling sheets, finish level, and person-hours split into hanging, finishing and stocking. After a dozen jobs you have rates. After thirty you have rates you can defend to a customer. Paste the log into an assistant and ask for hours per sheet by surface and level, plus the spread, not just the average. If the log is a mess of notes and half-filled sheets, the tutorial on cleaning messy data covers the tidy-up first. For the imaginary crew, the log works out like this:

Check that the log reconciles before you trust it. The hours the assistant adds up for last month should match your timesheets within a few per cent. If they do not, jobs are missing or hours are landing on the wrong job, and every rate built from the log inherits the error.

One basement priced per sheet, per square foot and by the hour

The job is a 20 ft by 40 ft basement, 800 square feet of floor, with 8 ft walls and one 12 ft partition boarded on both sides. Wall area is 120 ft x 8 ft = 960 square feet around the edge plus 12 ft x 8 ft x 2 sides = 192 for the partition, so 1,152 square feet, or 36 sheets. The ceiling is 800 square feet, or 25 sheets. That is 61 sheets. I have not deducted the door and windows, because cutting around openings costs more time than the board saves. With 10% waste, 61 x 1.10 = 67.1, and you round up to 68 sheets to order.

I would quote $6,100. Expressed in the units customers ask about, that is $100 a sheet, $3.13 a square foot of board (1,952 square feet) and $7.63 a square foot of floor. Now compare what you would have charged using the rates from the crew's last job, a 1,200 square foot apartment with 74 sheets, 12% of them ceiling, which billed $6,512.

The two borrowed rates fail for different reasons, and the floor-area one fails worst because it hides two variables at once: wall height and ceiling share. A basement with a full ceiling and long walls carries far more board per square foot of floor than the apartment did. Per sheet at least counts the board; it just prices every sheet as if it were a wall.

Now the time risk. Suppose the job takes 20% more person-hours than planned: 73.9 instead of 61.55. At the fixed $6,100 the margin falls from 25% to about 16%. If you had agreed to bill by the hour, the margin stays at 25%, but the customer's bill ends near $6,840 instead of $6,100, which is exactly why customers ask for a cap on hourly work. Fixed prices suit measured work. Hourly suits work where neither side can see the finish line.

Why ceilings and finish levels move the price more than sheets do

The Gypsum Association's finish levels, in GA-214 as summarised by CertainTeed, matter because each level is a different amount of work. Level 3 (tape and an extra coat) suits medium or heavy textures. Level 4 gives two coats over flat joints and three over fastener heads for smooth walls under flat paint. Level 5 is Level 4 plus a skim coat over the whole surface, recommended for non-flat paint, dark colours and critical lighting.

In the crew's log, finishing a wall sheet takes 0.38 person-hours at Level 3, 0.50 at Level 4 and 0.82 at Level 5, because the skim adds 0.32 hours and about $4 of compound a sheet. Ceilings need 1.2 times the finishing hours of a wall. Priced from those logs at a 25% margin, with board (plus 10% waste) and consumables included, about $2 less compound at Level 3, and stocking and delivery left out, a sheet comes out like this:

Two readings matter. A Level 5 wall costs 29% more than a Level 4 wall, so a customer who asks for gloss paint or a long window wall with raking light is asking for a different price, not an upgrade you absorb. And a ceiling sheet needs 44% more labour than a wall sheet (1.15 person-hours against 0.80 at Level 4) yet prices only 25% higher, because board and consumables cost the same either way. Notice that the wall Level 4 figure of $85 sits close to the $88 the crew had been charging on everything. That is why the old rate looked fine on walls and failed on ceilings.

If you mix 8 ft and 12 ft sheets, a per-sheet rate stops meaning much. $100 a sheet is $3.13 a square foot on 32-square-foot sheets and $2.08 on 48-square-foot sheets. Convert to square feet of board first, then quote, and state the sheet length on the quote.

Repairs and patches: hourly pricing with a floor under it

Repairs are small, so what costs money is the visit. Each coat of compound needs time to dry, which means most patches take two or three visits, and every visit includes a drive. Price the hours at the crew rate, add materials at the same margin, and add a visit charge. Here the visit charge is $40: half an hour of driving at $45.38 plus about $7 of vehicle cost, divided by 0.75.

The first row gives you the minimum call-out: $135, however small the hole. The formula behind each row is hours x $60.51, plus materials / 0.75, plus visits x $40. A customer who balks at $135 for a nail pop is comparing it with the five minutes of work they can see, not with two drives and a drying wait. I would write that into the quote in one sentence: "Minimum charge $135, which covers two visits so each coat of compound can dry before the next."

Water damage is the exception to flat pricing. Quote an inspection visit by the hour, and make closing up a wet wall conditional on the cause being fixed and the framing being dry. The drywall is the small part of that job, and a price quoted before you know the cause will be wrong in one direction or the other.

Back-test the per-sheet rate on six past jobs

Suppose the crew has charged $88 a sheet on every job for a year. Six of those jobs go into a spreadsheet and then into an assistant, with the prompt below. The point is not to get a new rate from the AI. It is to find out which job features the single rate fails on.

Here are six finished drywall jobs as CSV with these columns:
job, sheets, ceiling_sheets, finish_level, charged, actual_person_hours, materials_cost.
Our cost per person-hour is $45.38 (wages, on-costs and overhead).
For each job calculate cost, profit, margin on the charged price, and ceiling share of sheets.
Then say whether ceiling share or finish level explains the differences in margin.
Show a spreadsheet formula for every column so I can rebuild it myself.
For the all-jobs figure, use total profit divided by total charged, not an average of percentages.

The result, sorted by ceiling share (illustrative):

Among the Level 4 jobs, margin slides from 22.4% with no ceilings to 7.0% at 61% ceilings. Job 3 is the exception: only 17% ceilings, but it is the one Level 5 job, and it lost money. Across all six, the crew charged $29,656 against a cost of $25,014, a margin of 15.7% when the target was 25%.

Watch the headline figure. An assistant that averages the six margin percentages will report 14.1%, because the 28-sheet job counts as much as the 95-sheet job. Total profit over total charged is the number that reflects the business, which is why the prompt says so. Check also that the assistant used your cost per person-hour rather than a figure it found for itself, and that it did not treat the finish level as an explanation without testing it: remove job 3 and the ceiling-share pattern still holds.

What I would do with the result: split the rate into a wall rate and a ceiling rate, price Level 5 as its own line, and keep something near $88 only for Level 4 walls. The tutorial on checking building rates against past jobs uses the same method on mixed trades.

Four arithmetic slips to catch before the quote goes out

1. Rounding before waste. 61 sheets with 10% waste is 67.1, so order 68. Rounding 67.1 down to 67 leaves you a sheet short on the day. Ask for the working, not the answer.
2. Floor area standing in for board area. The basement has 800 square feet of floor and 1,952 square feet of board. Any figure labelled "per square foot" should say which of the two it means.
3. Mixed sheet lengths. The same $100 a sheet is $3.13 per square foot on 8 ft sheets and $2.08 on 12 ft sheets. State the length and keep the conversion in one cell.
4. Markup mistaken for margin. $45.38 x 1.25 is $56.73, and the margin on that is 20.0%. Dividing by 0.75 gives $60.51.

Whatever the assistant calculates, ask it for formulas, paste them into a spreadsheet and check two rows by hand. Chat assistants write good formulas and occasionally slip on multi-step arithmetic, and the formula is what you can audit. If you want a step-by-step on that, the guide to using ChatGPT with Excel covers it.

Putting the unit and the finish level on the quote

The price is only as clear as the line that carries it. This is the wording I would send for the basement, with the unit, the sheet length, the finish level and the extras rate stated:

Supply, hang, tape and finish 61 sheets of 1/2 inch drywall (8 ft lengths, measured from drawings) to Level 4: smooth, ready for flat paint. $6,100 fixed. Not included: priming and painting, a Level 5 skim coat, removal of existing ceilings, moving furniture. Sheets needed beyond the measured count are charged at $100 each, hung and finished.

Then ask an assistant to read it as the customer would. A prompt such as "List three things a customer could reasonably assume are included in this quote that are not" usually returns something useful, and the answers tend to be the ones that start arguments: who removes the old ceiling, who supplies access equipment, whether the second coat of compound counts as a separate visit. Add the ones that matter. The drywall estimate template carries the full scope and exclusions wording, and the tutorial on bidding drywall from plans covers quantity checks when a main contractor supplies drawings.

Drywall pricing questions that come up next

Do I charge separately for stocking and carrying sheets?

Either works if the hours are counted. Put stocking in as its own line when stairs, long carries or no lift are involved, or fold it into the per-sheet rate for easy sites. In the basement example it was 4 person-hours for 61 sheets. Log those hours on every job so the rate reflects real access, not a guess.

Can ChatGPT read a plan and count the sheets for me?

It can read dimensions that are written clearly on a drawing, but it can misread scale, miss rooms or confuse floor area with wall area. Treat any sheet count as a draft and check it against your own measurements. Plan takeoff software exists for bigger jobs, and the series has a separate tutorial comparing it with photo-based takeoff.

How do I word the finish level so the customer understands it?

Say what they will see and what comes next: Level 4, smooth and ready for flat paint. Then name what is not included, such as priming, painting and a Level 5 skim coat for gloss paint or strong side lighting. A one-line plain-English description beats a bare number, and it gives you a clean price for the upgrade.

Further reads

• Plastering Quote Template: Skim, Render and Patch Lines With AI — Skim coat and patch lines, if you also plaster.
• How Small Builders Use AI to Price Variations and Change Orders — Price extra sheets and scope changes once the job is running.
• How to Price Carpentry Work: Test Day Rates Against Job History With AI — Test a day rate against job history, not memory.
• Painting Estimating Software: Which Ones Use AI and What They Cost — What the painters who follow you use to price.
• How Much Does Estimating Software Cost? Price Bands With and Without AI — What estimating software costs once a spreadsheet stops being enough.
• AI Drywall Estimating Software: Plan or Photo Takeoff for a Small Crew? — Plan or photo takeoff tools for counting sheets on bigger jobs.
• AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.

Sources: CertainTeed's guidance on recommended levels of finish (summarising the Gypsum Association's GA-214); a general drywall reference for standard panel widths, lengths and thicknesses. Every crew cost, production rate and price in the examples is invented for illustration.