Understanding Standard Costing in Logistics
What should one order cost?

There's an exchange you hear constantly in warehouses, and it's almost always answered wrong.

"What does it cost us to fulfill one order?" "Take last year's center operating cost, divide by order count — about 2,300 KRW."

The arithmetic is fine. The answer is useless. Inside that 2,300 KRW sit single-item orders that finish in three minutes and multi-line orders that take twelve; November overtime and February idle time; one shipper's profit quietly covering another shipper's loss. Averages are convenient. But not a single order in your warehouse actually costs the average.

That's what always happens when you count cost afterward. A month-end close is an autopsy report on something that already happened, and autopsies don't save patients.

This series treats logistics cost the other way around — set it in advance, then read the gap against actual. Part 1 starts where that begins: standard costing.

A standard cost is not a forecast

Accounting textbooks define standard cost simply.

Translated into logistics: how many minutes should one order take (standard time), and what is one of those minutes worth (standard rate). Multiply the two and you have the standard cost of one order.

Three things need separating first. They get mixed up constantly in practice, and once mixed, standard costing goes limp fast.

A forecast exists to be accurate. A budget exists to be respected. A standard cost exists to be compared against. Which is why the real output of standard costing isn't the cost figure — it's the variance against actual.

That variance splits one sentence into two. "Outbound cost rose 18% this month" issues no instruction to anyone. But "we spent 12% more time per unit of volume, and each hour cost 6% more" goes to two different desks — the first to the floor, the second to HR and procurement.

Why standard costing matters more in a warehouse

Manufacturing has a BOM. Which materials, how many, how many minutes — it exists as a document from the design stage. Setting a standard cost is comparatively easy.

Logistics has no BOM. And "one order" turns out to be a far more uneven unit than it sounds. The same single order might have one line or seven, loose eaches or full cases, stock at the end of a Zone A aisle or on the top level of Zone D. In manufacturing that degree of variation is a defect. In logistics it's the normal workload.

The cost structure, meanwhile, is unsentimental. Study after study converges on one point — more than half of a warehouse's operating cost is labor. Recent industry sources put labor's share of DC operating cost somewhere between 45–57% and 50–70%. And labor always decomposes into time × rate. Which means more than half of warehouse cost stands on a physical quantity called time.

If time isn't measured, half the cost isn't measured. And the unmeasured half always disappears the same way — revenue prints clearly on the invoice, cost scatters across the floor and evaporates.

For a 3PL there's one more layer. A 3PL sells by the unit and buys by the hour. Billing goes out per order and per pallet; spending goes out per hour. The exchange rate between those two units is standard time. How often people sign a rate card without knowing that exchange rate is proven by the question that surfaces at every renewal: "why is this shipper so painful?"

In-house logistics (1PL) is no different. Without a basis for allocating internal logistics cost to business units and products, logistics stays a "cost department" forever. Only an organization that can break cost down to the unit gets to treat logistics spend as something to explain rather than something to cut.

A standard cost is assembled from three parts

Building one means deciding three things: activities, standard time, standard rate. That order is also the correct order.

① Activities and cost drivers — half the work is decided here

An activity is the unit you attach cost to. Receiving, inspection, putaway, replenishment, picking, packing, shipping, returns processing, and value-added work (labeling, inspection, kitting) covers most centers.

Each activity gets one cost driver — the physical quantity that represents its workload. This choice is half the job of standard costing. Choose wrong and everything downstream is precisely wrong.

The wrong answer you see most often is measuring picking by order count. When a one-line shipper and a six-line shipper are assigned the same cost, the latter is profitable on paper and loss-making in reality. And that loss shows up nowhere in the reporting — because you designed the count that way in the first place.

② Standard time — normal time plus an allowance

Standard time comes out of industrial engineering, and the formula is plain.

Normal time is how long a trained worker takes at a normal pace, uninterrupted, doing only that task. On top of it you add a PF&D allowance — Personal needs, Fatigue, and unavoidable Delay. Industrial engineering practice typically uses a 9–15% range, and for a warehouse floor something near 15% is realistic.

Skipping the allowance is the beginner's first mistake. A 0% allowance assumes a person moves at the same speed for eight straight hours without a bathroom break, and a standard built on that assumption spits out nothing but unfavorable variances every month. The floor then stops believing the standard, and a standard nobody believes is the same as no standard.

There are three ways to build standard times, and all three have a place.

One piece of practical advice here. Trying to build engineered standards from the start usually means never starting at all. You collect consulting quotes, schedule time studies, and half a year is gone. Setting engineered standards for your top three activities by volume and leaving the rest at historical P50 beats attempting full precision and delivering nothing.

The trap to watch is what a historical baseline actually is. An average is not a standard. An average blending a third-week hire with a five-year veteran is an observation of "this is what our center currently does," not a benchmark of "this is what it should do." So even when using history, work from the median and above, and separate out tenure bands. For reference, floor practice applies the learning curve in steps — 70% of standard in weeks 1–3, 85% in weeks 4–8, and full standard from week 9. Without that split you'll blame the standard when efficiency variance spikes in a month with a lot of new hires.

③ Standard rate — where most people get it wrong

The standard rate is "what one hour is worth." It's also the most frequently understated piece in practice. Put the hourly wage in as-is and your entire cost lands low.

A direct-labor standard rate has to be stacked up at minimum like this.

• Base hourly wage — Korea's 2026 minimum wage is 10,320 KRW per hour (+2.9% year over year), or 2,156,880 KRW per month at 40 hours a week. Warehouse floor roles typically start above that.
• Employer share of social insurance — national pension, health insurance, employment insurance, industrial accident insurance. Roughly 10–12% depending on industry.
• Severance accrual — one twelfth of annual wages, about 8.3%.
• Paid time — annual leave, paid holidays, and the like. A few percent depending on how you run it.

Stack all of that and the loaded hourly rate lands around 1.2× the base wage. A 12,000 KRW base means a real standard rate closer to 14,700 KRW. Price a contract off a standard cost that omits that 20% and the deal is loss-making from the moment you sign.

Overhead goes on as a separate absorption rate. Pool rent, depreciation, utilities, supervisory salaries, and system costs, then divide by normal capacity (monthly direct labor hours).

Let's run the numbers — the standard cost of one order

You need figures to get a feel for it. Let's assemble one for a single-item B2C order.

Standard time first.

Standard rate. 12,000 KRW base + 23% burden = 14,760 KRW/hour. Center fixed cost of 35,000,000 KRW per month ÷ normal capacity of 5,000 hours = an overhead rate of 7,000 KRW/hour.

Standard cost.

Now run the same method on a five-line multi-item order. Picking adds about 1.1 minutes per line, check and pack grow too, normal time reaches 10.5 minutes, and standard time lands at 12.1. Standard cost: 4,388 KRW. That's 2.2× the single-item order.

Which reveals what that opening "average of 2,300 KRW" actually was. It sits close to the single-item order (about 2,000 KRW) and understates the real cost of a high-line shipper (about 4,400 KRW) by more than half. A 3PL billing off average unit cost is quietly subsidizing its high-line shippers.

The real output of standard costing is the variance

The month after you set the standard is when the real work starts: splitting the gap between actual and standard. Accounting calls it variance analysis, and it carries over to logistics unchanged.

In numbers. A center running the standard above (5.5 min / 14,760 KRW) ships 20,000 orders this month, burns 2,050 actual hours, and posts 32,000,000 KRW of actual labor cost.

Split those 4.94 million KRW into two pieces.

The actual rate was 32,000,000 ÷ 2,050 = 15,610 KRW, 5.8% above standard. Actual hours ran 11.8% over the allowance. And of the 4.94 million KRW overrun, roughly 65% came from time and 35% from price.

That one line is the reason standard costing exists. "Cost overran by 18%" produces no action; "two thirds of the overrun is time" does. Instead of negotiating overtime rates down, you go look at why 20,000 orders consumed 2,050 hours. Too many new hires? Slotting drifted? Multi-line share up? Or is 5.5 minutes simply no longer true?

A third variance attaches here: volume variance. If you planned to absorb 35,000,000 KRW of fixed cost across 5,000 normal-capacity hours but actual activity stopped at 4,300, then 700 hours × 7,000 KRW = 4.9 million KRW of fixed cost sits unabsorbed.

Five ways standard costing fails

Standard costing took a serious beating in manufacturing. The lean accounting camp went as far as arguing that standard costs obstruct improvement. Read that critique rather than dismissing it and you find the problem was usually not standard costing itself but how it was operated. The failure patterns repeat almost identically in logistics.

① The standard goes stale. The most common and most damaging. Best practice recalibrates every 18–24 months; reality is once every 3–5 years, or never. The warehouse changes in the meantime. Industry analysis attributes to e-commerce order complexity a 3–4× increase in distinct DC task types since 2018, with omnichannel centers going from roughly a dozen task types to 40–50. Install a conveyor, add a put wall, switch to batch picking, and keep using the old standard — and that standard is measuring a warehouse that no longer exists.

② Calling the average a standard. A historical average is an observation; a standard is a benchmark. Treat them as the same thing and improvement reads as zero forever, because matching yesterday means a variance of zero.

③ Using the standard only as a scorecard. The primary purpose of standard costing is cost attribution; individual evaluation is a by-product. Reverse that order and the floor games the standard — cherry-picking easy orders, deferring hard ones, and skipping anything the system doesn't record.

④ Treating non-productive time as if it didn't exist. This is the quietest leak in standard costing. Industry analysis puts roughly 25% of total paid hours into non-productive activity: travel, shift handoffs, waiting, manual check-ins. Leave that 25% out of the standard and it reappears every month as an unfavorable variance with no identified cause. For scale: improving labor utilization by just 5% is estimated to save a mid-size DC 400,000–700,000 USD a year. Vanished time is not small money.

⑤ Setting standards without a ledger. Even with a standard time, you cannot calculate a variance if actual time isn't recorded. In a warehouse where who did what, when, and how much never lands in a ledger at the activity level, a standard cost is just paper on a wall. The order is always the same — measurement first, standards second.

What's reshaping cost calculation in 2026

Standard costing is an old tool. Several reasons to look at it again have appeared.

① Aggregate indicators can fall while unit cost doesn't.

Per the 2026 CSCMP State of Logistics Report (authored by Kearney, released June 16), U.S. business logistics costs came to 2.4 trillion USD, or 7.8% of GDP, down from 2.6 trillion and 8.7% the prior year. Don't read that as "logistics got cheaper." An aggregate ratio moves because the freight market and the GDP denominator both move; it's a different story from the cost of one order in your building. The proof is that the same report flags labor market and productivity constraints as structural pressures in the same year. Never judge your own unit cost from a macro indicator. That's taking someone else's temperature to check your own fever.

② Labor is still the biggest line, and 3PLs absorb the pressure directly.

In the annual benchmark survey of U.S. 3PL warehouses (Extensiv), 70% of respondents reported rising labor costs and 53% said labor exceeds 40% of total business cost. Korea points the same way. The 2026 minimum wage is 10,320 KRW per hour (+2.9%), and there is no mechanism automatically indexing private 3PL rates to it. Labor rises by index; rates rise only by negotiation. What closes the gap between them is a costing basis.

③ Variable cost is migrating into fixed cost.

Automation converts labor (variable) into depreciation (fixed). Industry estimates put process improvement at 15–30% labor savings with almost no capital outlay, while automation delivers 30–50% but demands an 18–36 month payback. The character of cost management changes here. A center with a heavier fixed base takes the hit through volume variance in months when volume drops. The more automated the warehouse, the more it needs to know its break-even volume — and break-even comes from dividing fixed cost by unit contribution. Both come out of standard costing.

④ Tariffs pushed cost from a "rate" into a "scenario."

The U.S. effective tariff rate climbed to roughly 17% in 2025, the highest sustained level since the 1930s. Static rate comparisons lost their force in sourcing decisions, and modeling total landed cost across three to five tariff scenarios spread in their place. Tariffs themselves aren't a warehouse cost, but the ripple reaches the warehouse. Move sourcing and inbound lot sizes, lead times, safety stock, and storage turns all change — which is to say standard times and capacity utilization change. Setting cost as a set of scenarios rather than a single number has become a necessary habit.

⑤ For AI to judge anything, a baseline has to exist.

The same CSCMP report assesses that AI has "made the crossover from a technology to try, to one that delivers measurable commercial returns in specific, well-defined applications." There's a catch in that. Any model, AI or otherwise, can only say "this is abnormal" if normal has been defined. On data with no standard, what a model produces isn't anomaly detection — it's the inertia of the past. Standard costing isn't an obsolete tool in the age of automation; it's the baseline that automated judgment leans on.

Korea already has a standard — the Corporate Logistics Cost Guidelines

There is one domestic framework you have to know when handling logistics cost: the Corporate Logistics Cost Calculation Guidelines (기업물류비 산정지침). It rests on Article 26 of the Framework Act on Logistics Policies, and the current version is MOLIT Notice No. 2016-182, effective April 7, 2016. It was developed jointly with the Korean Institute of Certified Public Accountants, and application is voluntary, not mandatory.

The guidelines define logistics cost as "the economic value incurred or consumed in carrying out logistics activities." They then classify it along six axes. The distinction matters: the first four exist to establish the facts, the last two to manage.

Two calculation methods are offered as well. The general basis computes logistics cost from separate cost records using cost accounting; the simplified basis estimates it from ledgers and financial statements using financial accounting. The effort differs, so most companies take the simplified route.

Something needs pointing out here. The guidelines standardized the containers that hold logistics cost; they do not hand you a unit standard cost. A figure derived on the simplified basis will tell you "logistics is X% of our revenue," but never "this is what one order costs." You end up knowing the total and not the unit.

Put differently, only the last two of the six axes connect directly to standard costing. Without the capacity axis (fixed vs. variable) you can't build break-even or volume variance; without the management-item axis (customer, activity) you can't attach cost to a shipper. The two axes the guidelines labeled "for management" are exactly where standard costing lives.

What you can do in the first month

Building standard costs properly is a project of several months. Starting takes one. Getting one full lap in matters more than completeness.

1. Cut down to five activities. Receiving, putaway, picking, packing, shipping. Value-added work joins in round two.
2. Pick one cost driver per activity. Lines for picking, boxes for packing, pallets or cases for receiving. Rather than hunting for the perfect driver, choose on whether you can still use the same driver next month.
3. Set standard times provisionally at historical P50. Time studies come later. Do write down that these values are provisional.
4. Always load the burden onto the standard rate. Wage × roughly 1.2, plus the overhead rate. Skip that one line and the entire first month is meaningless.
5. Run a month, then split the variance into rate and efficiency. This is where usable information appears for the first time.
6. If the variance exceeds ±30%, suspect the standard. Look at the standard before you look at the floor. First standards are usually wrong, and that's normal.

One precondition for all of it. If activities aren't recorded, a standard cost cannot be calculated at all. No actual time means no variance, and no variance means a standard cost is just a budget under another name. Whether activity-level records are accumulating in your warehouse ledger — that is the one real prerequisite for this work.

Where the series goes next

This piece was about building the ruler. The remaining four are about where you hold it.

The activity-based costing of Part 2 isn't a competitor to standard costing. ABC defines the path cost travels, and standard costing defines the unit value carried along that path. Only together do they complete the sentence "this shipper cost us X this month, and Y of it fell outside standard." Looking at per-shipper margin without that sentence is something we already covered in "3PL Shipper Profitability: Why Revenue Alone Misleads You."

One last word on the nature of this work. It looks like accounting, but what it actually is, is an agreement. The moment the floor and management both accept that one order should take 5.5 minutes, those 5.5 minutes stop being a cost figure and become a shared benchmark. And only an organization with a benchmark can prove improvement. Without proof of improvement, there's no basis for defending your rates at the table next year either.