Setting Up a Welding Shop by the Numbers
Chapter 1 of 10 · 10 min read
Deciding what work the shop will do
Choose the work a new welding shop will do, estimate the hours it can sell from two people's time, and see how the work mix decides the building, the equipment and the prices.
Every other decision in setting up a shop follows from one: what work will it do? A shop that builds trailers needs a long floor, a big door and a lifting beam; one that repairs machinery needs a lathe, a press and a bench for small parts; one that does site work needs a vehicle and a generator welder more than floor space. The work also decides who the customers are, how they find the shop, how jobs are priced and how much cash the business needs. A shop set up for the wrong work wastes money on things it does not use and turns away work it cannot do.
This chapter looks at how the early repair shops chose their work, sets out a method for estimating the mix of work for a new shop, calculates how many hours two people can actually sell, and introduces the shop whose setting up the rest of the book follows.
The repair shop of 1917
The custom welding shops of the early oxy-acetylene era knew their work. Willis wrote in 1917 that probably about three-quarters of the welding done in a custom repair shop was the welding of castings: cracked cylinder blocks, broken gear housings, crank cases, machine frames and farm implements, mended at a fraction of the cost of a new part. Dunham’s Automobile Welding of 1917 was written for garages and repair shops doing exactly that work on cars. Their equipment, layouts and prices were built around castings: preheating fires, brick furnaces, a heavy table and patient work on one-off repairs.
The work has changed. Castings are fewer, parts are cheaper to replace, and many repairs that once needed a welder are now done by fitting a new part. But the method the early writers used, building the shop around the work it would actually get, is the right one.
Kinds of work
A small welding and fabrication shop can aim at several kinds of work, each with its own demands.
| What it needs | How it is usually priced | |
|---|---|---|
| Repair and maintenance | flexible skills, small-parts bench, machining help | time and materials, minimum charge |
| Jobbing fabrication | floor space, cutting, bending, lifting | quoted per job |
| Product manufacture | jigs, repeatable process, storage | price per unit |
| Site work | vehicle, generator welder, site safety systems | hourly or day rate plus travel |
| Small structural | qualified procedures and welders, documentation | quoted to drawings and specifications |
Repair work is varied, urgent and often small, and customers judge it on speed and on whether it lasts. Jobbing fabrication, such as gates, trailers, brackets and frames, is quoted and needs room and handling equipment. Product manufacture, such as a range of trailers or stands, repeats and rewards jigs and efficient methods. Site work brings the welder to the customer, with travel, access and safety systems. Small structural work brings standards, qualified welders and documentation, as Testing Welded Joints by the Numbers and Welder Training by the Numbers describe.
The running case
This is an illustration. Mara, a welder with fifteen years in a fabrication shop, plans to start her own business in a regional town, with one employee, Sam, a qualified welder she has worked with. She knows the local market well: farmers who need trailers, gates, stock handling equipment and machinery repairs; small manufacturers who need brackets, frames and maintenance; transport operators who need truck body and trailer repairs; and builders who sometimes need small structural steel, stairs and balustrades. She has found an industrial unit for rent and has savings and a small loan to equip it. Before signing the lease, she wants to know what she is setting up for.
Estimating the work mix
Mara’s first step was to estimate how many hours a year of each kind of work she could expect, from her knowledge of the local market, conversations with past customers, and work she had seen turned away by her former employer. This is an illustration, and the estimates are hers, not data.
The total, about 2,340 hours, became the target for the business’s first year. Two things stood out. About a third of the work, farm fabrication, needed floor space for trailers up to about 6 m long and a door big enough to get them out. And about a sixth was site work, which needed a vehicle with a generator welder rather than anything inside the shop.
How many hours can two people sell?
The hours a business can sell are far fewer than the hours its people are paid for. A full-time person works about 38 hours a week, but after public holidays, annual leave and sick leave, a year has about 1,800 to 1,850 working hours. Not all of those can be charged to customers: time goes on quoting, ordering, cleaning, maintenance, deliveries, waiting, training and running the business.
- Sam, working hours a year
- about 1,800 at about 75% utilisation
- Mara, working hours a year
- about 1,800 at about 55% utilisation
- Sam: 1,800 × 0.75 = 1,350 billable hours
- Mara: 1,800 × 0.55 = 990 billable hours
- total ≈ 2,340 billable hours a year
Mara’s own utilisation is lower than Sam’s because she will also quote, order, invoice, pay bills, chase payments, deal with customers and plan the work. The total of about 2,340 billable hours matches her estimate of the work, which told her two things: that two people were about right for the expected work, and that if the work came in as estimated, the shop would be full, with little slack. Chapter 8 uses the billable hours to set the hourly rate, which depends heavily on them.
The utilisation figures are estimates too, and they are among the most important numbers in the plan. A new owner who assumes that every working hour will be billed will set prices too low and find, by the end of the first year, that the business has paid its costs out of the owner’s wages. Mara used the figures her former employer achieved for its welders and supervisors as a guide, and planned to measure her own from the job cards from the first week.
Testing the estimate
An estimate of work built from one person’s knowledge can be badly wrong, so Mara tested hers before committing. She rang a dozen past customers and asked, without pressing, how much work of each kind they might send in a year; she asked the rural store and the truck repairer how often they were asked for a welder; and she looked at what her former employer had turned away because it was too small or too far out of town. The answers were rough, but they were independent of her own hopes, and they confirmed the farm fabrication and repair work while suggesting that the structural work was optimistic. She cut the structural estimate in half in her planning, which is reflected in the figure above, and kept the original number as an upside rather than a target.
Busy and quiet seasons
Work does not arrive evenly through the year. Farm work peaks before harvest and in the months when machinery is prepared, falls away at harvest itself, and has a second, smaller peak before winter. Manufacturers’ maintenance work clusters around their shutdowns, often over the Christmas period. Site work slows in wet weather. Mara sketched the months in which each kind of work would be busiest and found that the shop could expect a crush in late winter and spring and a quiet patch in midsummer. That pattern shaped three decisions: to take on Sam from the start rather than later, to plan her own leave and the shop’s maintenance for the quiet weeks, and to build a cash reserve for the quiet months, which Chapter 10 sizes.
Work the shop will not do
Deciding what not to do is as useful as deciding what to do. Mara listed the work she would refer elsewhere: pressure vessel repairs, which need qualifications and inspection she would not have at the start; large structural jobs beyond two people and a forklift; precision machining beyond a simple lathe; and any work on fuel tanks or containers that had held flammable materials, unless they had been made safe and certified gas-free by someone qualified to do it. Having a list meant she could say no quickly and refer the customer to someone suitable, which kept goodwill and kept the shop out of work that could hurt it.
The work decides the shop
From the work mix, Mara drew up the shop’s main requirements:
- floor space for a 6 m trailer on stands with room to work round it, plus a smaller bay for repairs and brackets;
- a door at least 4 m wide and high enough for a trailer on a forklift or a ute;
- lifting, by forklift or a beam with a hoist, for trailer chassis and heavy repairs;
- cutting and forming, a saw, a plasma cutter, a press brake or bender, and a drill press;
- welding with solid-wire and flux-cored processes for steel, manual metal arc for site and heavy repairs, and tungsten arc for stainless and occasional aluminium;
- a site vehicle with a generator welder, cutting gear and a trailer;
- a machine shop only as far as a lathe and a press for repairs; larger machining would be subcontracted.
Each item has a cost, and the following chapters size the floor, services and equipment to these needs.
Practise before selling
The Imperial handbook of 1916 warned new owners against soliciting work before they could do it. It described novices who had sought jobs before their equipment had even arrived, and others who announced themselves ready for castings of any metal without having connected their equipment, and it advised setting up in the back of the shop and practising until competent, because practising on a customer’s job earns that customer’s enmity. Mara’s version of the advice was to offer only work she and Sam could already do well, and to practise anything new, such as aluminium repairs, on scrap before offering it to customers.
Specialise or stay general?
A new shop must decide how wide to cast its net. A general shop takes whatever comes, which keeps it busy but means equipment and skills for many kinds of work, small varied jobs, and prices that are hard to set. A specialised shop, say in trailers, does fewer kinds of work but does them efficiently, with jigs and repeatable methods, and can build a reputation in one market. Many successful small shops start general, to pay the rent, and specialise gradually as they find which work they do best and which pays best. Mara planned to start general within her mix and to review the job records at six months to see which work earned most per hour.
The customers and how they will find the shop
Work does not arrive just because the shop opens. Mara listed where each kind of work would come from: past customers who had followed her from her former employer, farmers through the local rural supplies store and word of mouth, manufacturers by direct approach, transport operators through a truck repairer she knew, and builders through tenders and referrals. Each source suggested a different first step, from a sign at the rural store to visits to manufacturers, and each had a different lead time. Chapter 10 builds the slow start of the first months into the cash flow forecast.
A failure: the aluminium boat
This is an illustration. In her first month, Mara was asked to repair a cracked aluminium boat hull, and took the job because the shop was quiet. She had welded aluminium occasionally years before. The repair took three times as long as she had estimated, the first attempt cracked beside the weld, and the second, after she had bought a different filler and practised on scrap, was sound but untidy.
She charged for the time she had estimated, not the time it took, and lost money on the job. The lesson she drew was the Imperial handbook’s: practise before selling. She added aluminium repairs to the list of work to practise on scrap in quiet periods, and did not quote another until she could weld test pieces that passed her own break test.
Exercise: a third person?
If the work grows by about 40% in the second year, about 940 more billable hours, how many more people at 75% utilisation would the shop need, and what would happen to Mara’s utilisation if she also had to supervise them?
At 1,800 working hours and 75% utilisation, each new person adds about 1,350 billable hours, so the extra 940 hours need about 940 ÷ 1,350 ≈ 0.7 of a person: a third full-time welder, with some spare capacity, or a part-time one.
Supervising a third person would take more of Mara’s time. If her utilisation fell from 55% to 45%, she would lose about 1,800 × 0.10 = 180 billable hours, so the shop would gain about 1,350 − 180 = 1,170 hours from the third person, still more than the 940 needed. Growth brings management work with it, and the hourly rate in Chapter 8 must allow for it.
