How to Price Welding and Fabrication

Quick answer: price from your own shop rate against realistic hours, use weight only as a sanity check, and never quote by the pound alone on complex work. The reason is simple. Weld inches do not scale with weight, and weld inches are what consume the day.

The Four Ways Shops Price, and When Each Works

Basis	Works when	Fails when
Per hour	Almost always, if hours are estimated honestly	The estimate is a guess with no history behind it
Per weld inch	Repeatable weldments, similar joint sizes	Heavy prep, positioning or fit-up dominates the job
Per pound	Heavy, simple, repeatable structural work	Light complex work with many short welds
Per foot	Handrail, pipe spool, repeated linear product	Anything with variable fit-up

Per hour is the honest default. The others are shortcuts that work inside a narrow band of similar work, and every shop that uses one has a band where it stops working.

Why Pricing by the Pound Quietly Loses Money

Weight tells you about material. It tells you very little about labour.

Consider two jobs of the same weight. One is a heavy plate weldment with four long straight welds. The other is a light frame with forty short welds, more fit-up, more positioning and more cleanup. Same pounds. The second takes considerably longer.

A per-pound price collects the same money for both. So the shop makes money on heavy simple work, loses it on light complex work, and the two average out to a thin overall margin that nobody can explain. Then the customer who buys the light complex parts keeps coming back, because you are cheap, and the mix drifts toward the work you lose money on.

That drift is the real damage. It is slow, it looks like growth, and it is invisible without job-level cost data.

Building the Number

1. Material. Include the drop. Charge the job for what gets issued, including the offcut you cannot use on the next job. Add a contingency on jobs quoted more than a few weeks out, because steel prices move and the quote does not.

2. Labour, by operation. Estimate separately and you will estimate better:

• Cut and prep, including squaring, bevelling and deburring
• Fit and tack, which is where complex assemblies actually consume time
• Weld, being the arc-on time
• Finish, meaning grinding, cleanup and inspection

3. If you price by weld inch, you need a deposition assumption, meaning inches per hour at a given size and process. Establish yours from your own jobs instead of from a table, because it varies with position, access, joint prep and the welder. Then add the prep, positioning and cleanup around it, which routinely exceed arc-on time.

4. Consumables. Gas, wire, discs, tips. Small per job, real per year, and almost always written to overhead and never recovered.

5. Outside processing. Galvanising, plating, heat treat, machining. Include freight both ways, and include the calendar time in your schedule.

6. Apply your shop rate. From how to calculate your shop rate. Do not use a rate you got from someone else, because it recovers their overhead.

7. Margin, and mind the trap. Divide by one minus your target margin. Do not mark up. A 30 percent markup on cost yields about a 23 percent margin, and the difference is money you meant to earn.

Reverse-Engineer Your Own Estimator

This exercise is worth an afternoon and it is the single most useful thing in this guide.

Take the last three quotes you sent. For each one, write down how every number was produced. Not what the number was. How it was arrived at.

You will find one of three things:

A consistent method. Excellent. Write it down so it survives the estimator taking a holiday, and so a second person can quote.

A method with gaps. Material is calculated and labour is a feel. Consumables are never included. That is fixable this week, and it is usually worth more than any software.

No method. The number comes from experience and a look at what the customer paid last time. That works while the person doing it is present and stops working the moment they are not, which is a continuity risk before it is a pricing problem.

Anyone selling you estimating software who has not asked to see three real quotes is guessing at your business. Our Edmonton software planning guide covers what to establish before any vendor conversation, and spreadsheets vs an operations platform covers how far a spreadsheet takes this.

Close the Loop, or None of This Improves

Pricing gets better only when quoted cost is compared against actual cost, by line.

A shop that compares totals learns that a job made or lost money. A shop that compares by line learns that the fitting was fine and the welding ran double, which tells the estimator exactly what to change on the next one like it.

That requires recording actual hours by operation while the job runs. Our job costing spreadsheet does this for free, with estimate against actual and a variance column, and the metal fabrication guide covers what happens when the spreadsheet runs out.

One honest caution. Estimate against actual is only as good as what the floor records. If weld time is written down at the end of the week from memory, the variance column tells you what people remember. Decide whether your shop will record daily before you build anything on top of it.