How to build an estimate you can actually stand behind

Most estimating advice is written by people who have never had to defend a number to a homeowner across a kitchen island. This is written by a renovation contractor who prices jobs every week, and it is about the part nobody teaches: not how to add up a quote, but how to build one that still holds when the job is finished.

An accurate estimate is not one that turns out exactly right. It is one where you know why every number is what it is, and where being wrong tells you something instead of just costing you money.

1. Know what an hour actually costs you

This is where most estimates are already wrong before a single quantity is measured.

You pay a carpenter $32 an hour, so you price labour at $32 an hour. But your real cost is that wage plus employer CPP, EI, WSIB at a construction rate, and vacation accrual. In Ontario that lands closer to $38 an hour — roughly 19% above the wage.

Price at $32 and you lose about $6 on every hour you sell. On a job with 100 crew hours that is $600 of margin gone, and you will never see it as a line item. It just shows up as a year that felt busy and did not pay.

The burdened rate is the number that belongs in your estimate. Work it out once, write it down, and use it everywhere. If you have never calculated it, that single afternoon is worth more than any software.

The same applies to the things nobody bills: driving between sites, the trip to the supplier, the hour you spend on the phone with the tile rep. If it is 6 hours a week and it is not in your rate, it is coming out of your profit.

2. Separate the measuring from the pricing

An estimate is two completely different jobs wearing one hat.

The first is measurement. How many square feet of wall tile. How many sheets of drywall. How many hours of framing. This is arithmetic, it is tedious, and it is where most of the evening goes. It is also the part that can be automated without risk, because the answer does not depend on your business.

The second is pricing. What you charge for those quantities, based on your supplier, your crew, your market and your appetite for the job. This part should never be automated, because the moment it comes from a national average database it stops being your number.

This distinction matters more than it sounds. A tool that hands you a fully priced estimate from someone else's cost data has not saved you time — you now have to check every line against what you would actually charge, which is slower than writing it yourself. A tool that does the measuring and then looks up your prices has genuinely removed the tedious half.

It is also why "how accurate is AI estimating" is the wrong question. Ask instead: whose prices is it using?

3. Price in phases, not in one number

A single total tells you nothing and defends nothing.

Break the job the way you build it — demolition, framing, plumbing rough-in, electrical, drywall, tile, finishes. Each phase gets its own labour, its own materials, its own hours.

Three things happen when you do this:

• You can find your own mistakes. A bathroom where tile setting comes out at four hours is obviously wrong. A bathroom that comes out at $18,400 is not obviously anything.
• The client can read it. Phases match how a homeowner thinks about the work. A 200-line breakdown invites an argument about your markup; six phases invites a decision.
• The estimate becomes the budget. Mid-job you can ask whether drywall came in at what you priced, while there is still time to react.

That last point is the one people miss. An estimate that stops being useful the moment it is accepted is a document. An estimate that becomes the yardstick you measure the job against is a tool.

4. Name the unknowns instead of padding for them

Every renovation has things you cannot see. Plaster instead of drywall. A joist in the wrong place. Knob-and-tube behind the vanity.

The instinct is to pad — add 10% and hope. The problem is that padding is invisible to you and to the client. If the surprise does not happen you have quietly overcharged; if it does you have usually not padded enough; and either way you have learned nothing about your own estimating.

Name them instead. A 1960s house with plaster walls usually costs an extra day of prep — say so, price it as a line, and let the homeowner see the assumption. Two things follow: your base number is honest, and when the wall opens up you are having a conversation about something you already flagged rather than delivering bad news.

The estimates that cause arguments are almost never the expensive ones. They are the ones where something appeared that was never mentioned.

5. Markup and margin are not the same thing

This one costs contractors real money every year.

If a job costs you $10,000 and you add 20% markup, you charge $12,000 and make $2,000. But $2,000 on $12,000 of revenue is a 16.7% margin, not 20%. To actually keep 20% you need to charge $12,500.

The gap widens as the numbers grow, and if you have been quoting on markup while thinking in margin, you have been running thinner than you believe on every job you have ever priced.

Decide which one you are working in, write it down, and be consistent. Then check it against reality — which is the next point, and the one that matters most.

6. Close the loop — or none of it improves

Here is the uncomfortable truth about estimating: you cannot get better at it if you never find out whether you were right.

Most contractors quote a job, build it, invoice it, and move on. The company was profitable this year, so things must be fine. But ask which of last year's jobs actually made money and the honest answer is usually a shrug — because the estimate lives in one place, the receipts in the truck, the hours in a group chat and the invoices somewhere else.

So you keep bidding bathrooms and basements the same way, with no idea that one makes you 34% and the other makes you 11%.

Closing the loop means comparing, for every finished job, what you priced against what it cost. Not roughly. Line by line, phase by phase, with labour costed at the burdened rate and every receipt attached to the job it belonged to.

Do that ten times and your estimating improves more than any technique on this page. You stop guessing at hours because you know what the last four bathrooms actually took. You stop underpricing basements because the numbers told you that you were.

This is the reason job costing and estimating belong in the same system. When they are separate — and in most software they are, because the financial side gets handed off to accounting — the comparison never happens, and the estimate never gets better.

A worked example

A 8×10 bathroom gut in a 1960s house. Remove the tub, install a 3×5 tiled shower with a glass door, new vanity, toilet, tile floor, move the plumbing two feet.

Structured properly, that becomes:

• Demolition & disposal — hours plus bin
• Plumbing rough-in — relocating the drain two feet is the real cost here, not the fixtures
• Shower — 92 sq ft of wall tile, waterproofing, niche, glass
• Floor — 80 sq ft, plus levelling if the subfloor needs it
• Fixtures, paint & finish
• Flagged, not priced: plaster walls typically add a day of prep

Each line carries its own hours at your burdened rate and its own materials at your supplier's price. The total is a consequence, not a guess. And when the job finishes, every one of those phases has a real cost sitting against it.

Doing this without spending your evenings on it

Everything above is just good practice — you can do all of it with a spreadsheet and discipline. Plenty of good contractors do.

What is hard is doing it consistently, on every job, at nine o'clock at night after a full day on site. The estimate that gets skipped is never the one you meant to skip.

Contrax was built to remove the tedious half. Describe the job the way you would to a foreman and it works out the quantities, the phases and the hours — then prices every line from your rate book, never a national average. If it cannot match a line to your rates confidently, it says so rather than inventing a figure.

Then the estimate becomes the budget. Crew hours cost to the job at the burdened rate. Receipts photographed at the supplier counter attach to the job the same day. And when it finishes, you get the comparison that makes you better at the next one — because the books are in the same system, not on the other side of an export.

It was built by a renovation contractor in Oshawa who got tired of pricing jobs at 9pm, and it runs his own company. Ontario HST, WSIB in the burdened rate, Construction Act holdback, prices in CAD.

Common questions

How accurate should a construction estimate be?

For renovation work quoted from a walkthrough, most experienced contractors aim to land within about 10% of final cost. What matters more than the percentage is whether you can explain the variance afterwards. An estimate that came in 15% under and told you exactly why is more useful than one that landed by luck.

What is the most common estimating mistake?

Pricing labour at the wage instead of the burdened cost. It is invisible, it applies to every hour on every job, and it compounds quietly for years. Second is confusing markup with margin.

Should I show the client a detailed breakdown?

Show phases, not line items. Six to eight phases match how a homeowner thinks about the work and let them make decisions. A 200-line breakdown invites a negotiation about your markup on individual materials, which is a conversation nobody wins.

How long should an estimate take to write?

For a typical bathroom or kitchen, an experienced contractor working from a good rate book and a clear scope should be looking at well under an hour. If it routinely takes an evening, the bottleneck is almost always working out quantities and typing them in — which is the part worth automating.

Do I need takeoff software?

Only if you regularly bid from full drawing sets. Takeoff tools measure off plans, and most renovation work is quoted from a walkthrough, photos and measurements. Different problem, different tool.

How do I know if my estimates are getting better?

Compare priced against actual on every finished job, with labour at the burdened rate and every receipt attached. After ten jobs the pattern is obvious — which work pays, which does not, and where your hours estimates run short. Without that comparison you are guessing with more confidence, which is worse than guessing.