Cost Accounting

How to Set Standard Costs That Reflect Production Reality

How to set standard costs that match your shop floor: material yield, real run times, labor and overhead rates, setup, and when to update. Worked example.

Two professionals reviewing financial spreadsheets and "Costing Reports" at a desk with a calculator and laptop.

If your standard costs don't match what happens on your shop floor, your financials are wrong in a way that's hard to see. Inventory is off. Gross margin by product is off. And every quote built on those standards is off by the same amount.

The fix isn't complicated. It's mostly a matter of building standards from your own production data instead of from hope, and then keeping them current.

This post walks through how we build standards for manufacturers, step by step, with one worked example that runs all the way through. If you're new to the idea, start with standard costing explained and come back.

What Are Standard Costs, and Why Do They Keep Breaking Down?

A standard cost is what one unit should cost to make under normal conditions. It's built from material, direct labor, and overhead, each as a quantity times a rate.

Standards break down for a simple reason. The inputs change and the standard doesn't.

Steel goes up. You give raises. You buy a faster machine, or lose your best setup person. Volume drops 20% and your overhead rate is still based on last year's busy plant. None of that updates itself.

The other reason is that many standards were never right to begin with. They came from the original quote, an engineering estimate, or an ERP default someone typed in during implementation.

Why Standard Cost Systems Fail Manufacturers

The same few causes tend to show up again and again.

They're built on best-case times. Engineering says the part runs in six minutes. On a good day it does. Across a real month, with tool changes, first-piece inspection, and the occasional bad bar, it runs closer to seven or eight.

Material yield is ignored. The standard says one pound because the finished part weighs one pound. But you bought 1.25 pounds to make it.

Setup is left out. Setup cost per part depends on lot size. If the standard ignores setup, short runs look much more profitable than they are.

The overhead rate is set on the wrong volume. Too high a volume and every part looks cheap. Too low and every part looks expensive.

Nobody owns them. Standards are set once by whoever implemented the ERP, then left alone. Variances pile up in one line and get written off at year-end.

The cost of all this is in your pricing. If the standard is low, you underquote and win work you lose money on. If it's high, you overquote and lose work you should have won.

Step-by-Step: How to Build Standard Costs That Hold Up

Here's the process. The example is a machined part made in a four-machine CNC cell. All numbers are illustrative.

Step 1: Start with the parts that matter

Don't try to fix 2,000 item numbers at once. Pull a sales report for the last 12 months and sort by revenue.

In many shops a fairly short list of parts carries a large share of revenue. Start there. Fix the top 20 or 30, then work down the list over the next few months.

Step 2: Set material standards from real yield

The material standard is what you consume, not what ends up in the part. That includes cut-off, bar ends, trim, and normal scrap.

In our example, the finished part has 1.00 lb of aluminum in it. Looking at six months of job records, the shop buys 1.25 lb for every good part. That's an 80% yield.

At $4.00 a pound, the material standard is 1.25 times $4.00, or $5.00. A standard built on the finished weight would say $4.00.

Use a recent price you can actually buy at, including inbound freight if it's material. If you import, duties belong in the material cost too. Our post on tariffs in your manufacturing cost structure covers that piece.

Step 3: Set run times from job history

Pull the actual run times from the last 10 or 20 times you made the part. Use the median, not the best run and not the average. One disaster job shouldn't set your standard, and neither should one perfect one.

In our example, engineering quoted 6 minutes a part. The median from job history is 7.5 minutes, or 0.125 hours.

If you don't collect run times by job, that's the first thing to fix. You can't build a real labor standard without it.

Step 4: Build a labor rate that matches what you pay

The labor rate should include what an hour of direct labor actually costs you. That's base pay plus payroll taxes and benefits.

In the example, operators average $24 an hour. Payroll taxes, health insurance, and other benefits add about $6. So the standard labor rate is $30 an hour.

Decide up front how you'll treat overtime premium. I think it's usually cleaner to leave it out of the standard and let it show up as a rate variance. That way you can see it. For more on building the rate, see calculating labor and overhead cost and how to calculate labor cost per unit.

At 0.125 hours, run labor is $3.75 a part. The engineering time would say $3.00.

Step 5: Set the overhead rate on normal capacity

The overhead absorption rate is budgeted overhead divided by budgeted hours. The hours you pick matter a lot.

Say the CNC cell's overhead for the year is $600,000. That covers depreciation on the machines, maintenance, tooling, power, the lease share, and supervision.

The cell has four machines. At normal volume, each runs about 2,000 hours a year. That's 8,000 machine hours, and a rate of $75 an hour.

Here's how much the choice of hours moves the cost of our part:

Hours used for the rateOverhead rateOverhead per part (0.125 hr)
16,000 (every machine, two full shifts)$37.50$4.69
8,000 (normal capacity)$75.00$9.38
6,000 (last year's slow year)$100.00$12.50

I'd use normal capacity. The theoretical number makes every part look cheap. The slow-year number makes every part look expensive and pushes your quotes up right when you need the work.

For financial statements this lines up with GAAP too. Fixed overhead is generally allocated based on normal capacity, and the per-unit rate shouldn't go up just because production was abnormally low. We cover the mistakes in more depth in overhead absorption mistakes destroying your margins.

Step 6: Spread setup over a realistic lot size

Setup on this part takes 3 hours. During setup you pay the operator and the machine sits, so setup costs 3 hours times ($30 + $75), or $315.

What that adds per part depends on how many you run at once:

Typical lot sizeSetup per partTotal standard cost per part
50$6.30$24.43
100$3.15$21.28
500$0.63$18.76

Use the lot size you actually run, not the one you wish customers ordered. If one customer orders in 50s and another in 500s, it's worth quoting them differently.

Step 7: Put it together and compare

Here is the realistic standard at a lot size of 100 next to the optimistic one built from engineering time, finished weight, and no setup.

ElementOptimistic standardRealistic standard
Material$4.00$5.00
Run labor$3.00$3.75
Overhead$7.50$9.38
Setup$0.00$3.15
Total$14.50$21.28

The gap is $6.78 a part, or about 47% on top of the optimistic number.

Now look at what happens in quoting. If you price for a 30% margin off the optimistic standard, the price is $14.50 divided by 0.70, or $20.71. Your realistic cost is $21.28. You're losing $0.57 on every part and the P&L will only show it as a pile of unfavorable variances.

That's the whole case for doing this work. Once the standard is right, see pricing and margin analysis for how we use it in quoting.

The Mid-Point Problem: What to Do When Standards Go Stale

Many shops update standards once a year, usually with the budget. That's a reasonable rhythm. The hard part is what to do in the middle of the year when something big moves.

I'd update mid-year when one of these happens:

  • A major material moves enough that the price variance on it is large every month.
  • A routing changes, like a new machine or a new fixture that cuts run time.
  • Wage rates change across a department.
  • Volume shifts enough that your normal capacity assumption no longer holds.

If the change is small, another option is to leave the standard alone and explain the known variance each month. That keeps the year consistent. It just means everyone has to remember why that variance is there.

If you do change standards mid-year, inventory on hand gets revalued to the new standard. That creates a one-time adjustment in the month you change it. Plan for it, explain it in the monthly package, and let your CPA know, so it doesn't look like a surprise at year-end.

Keep a log of every standard change: what changed, when, and why. Six months later nobody remembers.

How to Use Variance Analysis to Improve Operations, Not Just Close the Books

Once standards are right, variances become useful. They tell you where cost went off plan and who can fix it.

A good standard makes a variance mean something. If a part runs 20% over on labor hours against a realistic standard, that's a real problem worth a conversation on the floor. Against a fantasy standard, it's just Tuesday.

We cover the monthly review process in manufacturing variance analysis. For the specific variances, see material price vs. usage variance and labor rate vs. efficiency variance.

One habit I'd add. When a variance points at a standard that's wrong, fix the standard. When it points at a process that's wrong, fix the process. Don't fix the standard to make a process problem disappear.

When Your ERP and Your Reality Don't Match

In many systems, the standard cost is rolled up from the bill of materials and the routing. If either one is wrong, the standard is wrong, no matter how carefully you set the rates.

These are common gaps:

  • The scrap or yield field on the bill of materials is blank.
  • Units of measure don't match. You buy in pounds, issue in feet, and the conversion is off.
  • The routing is missing steps like deburring, inspection, or packaging.
  • Outside processing, like plating or heat treat, isn't on the routing at all.
  • The overhead rate in the system is from the year it was implemented.

A quick test: pick five of your highest-volume parts. Walk each one through the floor with the routing in hand. Then compare the standard to what the last few jobs actually cost. You'll know quickly whether the system can be trusted.

If you're also weighing a new system, our comparison of QuickBooks, NetSuite, and Fishbowl covers how each handles costing.

It also helps to know which of your costs are fixed and which move with volume. See fixed vs. variable costs for that split, and the role of standard costing in manufacturing for how it fits the bigger picture.

How Accounovation Helps Manufacturers Build Accurate Standard Costs

We act as the finance team for manufacturers in the $3M to $30M range, and costing is a big part of that work.

We pull your job history, sit down with production, and rebuild standards for the parts that drive your revenue. Then we set up the monthly variance report and review it with you, so the standards stay current.

If you want help, it's part of our manufacturing accounting and outsourced controller work. Good standards also feed better cash planning, which we cover in cash flow forecasting best practices, and better product margin reporting, covered in margin analysis in manufacturing. Happy to jump on a call if helpful.

Frequently Asked Questions

How often should manufacturers update their standard costs?

Once a year with the budget is a reasonable base. Update sooner when a major material price, a routing, wage rates, or volume changes enough that the same variance shows up every month.

Should I use standard cost or actual cost for pricing?

Use a realistic standard for quoting, because you need a cost before the job runs. Then check it against actual job costs regularly. If actual runs above standard month after month, your prices are probably too low.

Why is my overhead always under-absorbed?

Usually because the rate was set on more hours than you actually run. If you produce less than the volume behind the rate, part of your overhead never makes it into product cost, and it hits the P&L as an unfavorable variance.

Should overtime be built into the labor standard?

I'd usually leave the overtime premium out and let it show as a labor rate variance. If you run planned overtime every week all year, though, it's part of your real cost and belongs in the standard.

Should outside processing be part of the standard cost?

Yes. If every part goes out for plating or heat treat, that cost belongs on the routing and in the standard. Leaving it out makes the part look more profitable than it is.

More on: Cost Accounting · Pricing & Margins

Keep reading

Related articles

Get a finance partner in your corner

Book a free 30-minute call. We'll talk through your cash, your margins and the decisions ahead, as a partner, not just your accountant.

Book a free call