Every estimate answers two questions: what will this cost, and where does the money go. The first is a number. The second is your cost structure, and most shops never write it down. It lives in the shape of an old spreadsheet, in the columns somebody added ten years ago and nobody dares delete.
A cost bucket is simply a heading you want a total for on every scope: Material, Shop Fabrication, Finish, Field Erection. Get the buckets right and every estimate reads the same way, reviews go faster, and comparing what you bid with what the job cost becomes possible. Get them wrong and you either drown in twenty-five columns nobody looks at, or you hide your biggest risk inside a single line called "Labor".
Four rules for choosing buckets
1. A bucket should match a decision
Make something a bucket if, during review, someone would look at that total and decide something: "shop hours look light", "freight is too high for one trip", "why is finish a third of the scope?". If nobody would ever look at the total on its own, it does not need to be a bucket. It can be a line inside another one.
2. A bucket should match how you buy or schedule
Material is bought from suppliers. Shop time is scheduled on the floor. Field time is scheduled with crews. Galvanizing is sent out. Each of those is managed by a different person on a different calendar. When your buckets follow the same lines, your estimate can talk to purchasing, the shop and the field without translation.
3. A bucket should have one unit of risk
Combining shop and field labor into one "Labor" bucket is the classic mistake. Shop hours are controlled and predictable. Field hours depend on access, weather, the GC's schedule and other trades. Lumped together, the risky number hides behind the safe one.
4. Fewer is better, up to a point
Most fabrication shops land between eight and twelve buckets. Under six, important risks get hidden. Over fifteen, nobody reads the scope summary anymore; it becomes a wall of small numbers.
A starting set for a fabrication shop
The set below is a sensible starting point for a misc metals, structural steel or stair and railing shop. It is an example, not a standard. Rename, merge or split to fit how your shop actually works.
| Bucket | What goes in it | Why it is separate |
|---|---|---|
| Material | Shapes, plate, bar, pipe, grating, bolts, anchors, hardware | Bought from suppliers; prices move between bid and buy |
| Shop Fabrication | Cutting, fitting, welding, assembly hours | Your biggest controllable labor cost |
| Finish | Galvanizing, shop primer, paint, powder coat | Often sent out; easy to apply to the wrong item |
| Shop Drawings | Detailing and drafting | In-house or bought; often forgotten on small scopes |
| Engineering | Delegated design, calculations, stamps | Specified on some jobs and not others |
| Field Measure | Site visits before fabrication | A real cost that disappears inside "Erection" |
| Field Erection | Installation and field welding hours | The riskiest labor; deserves its own total |
| Cranes & Rigging | Crane, lift and rigging equipment days | Rented by the day; driven by the site, not the steel |
| Freight to Site | Trucking and trips | Depends on distance and piece size, not weight alone |
| Project Management | PM hours, submittals, coordination | Scales with job complexity, not tonnage |
An example only. Your shop's names, splits and rates will differ.
A few judgement calls you will face:
- Finish in or out of Material? Keep it out. Galvanizing is priced by weight, sent out, and scheduled separately. When a spec changes from galvanized to primed, you want to see exactly what moved.
- One Field bucket or two? If you send a separate person to measure, keep Field Measure. If the erection crew measures as part of the job, fold it into Field Erection.
- Where do subcontractors go? If you regularly sub out specific work (say, fireproofing touch-up or a specialty finish), a Subcontractors bucket keeps those quotes visible. If it is rare, put the quote in the bucket it replaces.
- Site costs like fire watch? Some jobs require them, most do not. A small bucket that is zero on most scopes costs nothing and stops it being forgotten on the jobs that need it.
Buckets versus cost components
Buckets are where totals land. They are not how numbers are calculated. That is the job of cost components underneath them. Keeping the two apart is what makes a structure both readable and accurate.
A cost component describes one way a cost gets into a scope: is the number typed by the estimator or pulled from the library? What does it measure (time, days, weight, quantity)? How is it calculated (hours × rate, days × rate, weight × rate, quantity × rate)? What is the default rate? Which bucket does it land in?
For example, your Shop Fabrication bucket might be fed by a single component: shop hours typed by the estimator, multiplied by your shop rate. Your Finish bucket might be fed by a galvanizing component that multiplies the material weight from the library by your galvanizing rate. Your Cranes & Rigging bucket might be fed by crane days times your day rate.
Two consequences matter in practice. First, the estimator types the judgement (hours, days, trips) and the structure applies the rates, so every estimator uses the same rates. Second, when a rate changes, you change it once in the component, and new estimates pick it up. In ScopeDive this all lives in the Engine Room: each company builds its own cost buckets (Material is always one) and up to 100 active cost components, with its own default rates.

The software should not calculate your fabrication. Your hours and your rates should. Be wary of any estimating tool, spreadsheet template included, that produces shop or erection hours from someone else's production table. Use your own history.
Mistakes to avoid
Copying the accounting chart of accounts
Your accountant's cost codes are built for tax and job costing, not for reviewing a bid. They are often too fine (forty labor codes) or too coarse (one "direct materials" line). Map between them later if you need to; design your buckets for the estimator.
Hiding contingency inside a bucket
Padding shop hours "just in case" makes the bucket useless for review and invisible to the person approving the bid. If you carry contingency, carry it openly, as an adder or a clearly named line, so it can be discussed.
Changing the structure every month
A cost structure is only useful if bids can be compared over time. Settle on a set, use it for a quarter, then adjust based on what reviewers actually looked at. Rename freely; restructure rarely.
Building buckets nobody fills in
If a bucket is zero on nearly every scope and nobody misses it, retire it. Every unused heading makes the real numbers harder to see.
Test it on a bid you already sent
The best test of a new structure is an old bid. Take a recent project you priced in your spreadsheet, ideally one you won and built, and price it again in the new structure. Then look at three things:
- Does every cost have an obvious home? If the estimator hesitates over where something goes, the buckets are not clear yet.
- Do the totals tell you something? Read the bucket totals for each scope. If Field Erection on one stair is double the other two, you should be able to see why immediately.
- Does it match the job cost? If the job is built, compare bucket by bucket with what it actually cost. That comparison is the start of better rates next year.
Once the structure holds up on a real bid, write it down and share it with every estimator. A shared structure is the difference between a shop with a few good estimators and a shop with an estimating system.
Related reading: the misc metals, structural steel and stairs and railings pages show example setups, and the glossary defines the terms used here.