Warehouse labor productivity metrics answer one question: how much work did you get for every paid hour? The five that matter are units per hour (UPH), lines per hour (LPH), orders per hour (OPH), cost per line, and direct labor utilization. UPH and LPH tell you how fast the floor moves. Cost per line tells you what that speed costs. Utilization tells you how many of your paid hours went to work that moves product.
Most warehouses track one of these, usually a blended UPH, and wonder why it never explains the labor line on the P&L. The fix is not more metrics. It is the right denominator, measured by function, against total paid hours rather than scanned hours. For reference, warehousing and storage labor productivity (output per hour) rose 1.4% from 2024 to 2025 after three straight years of declines, according to the Bureau of Labor Statistics. If your building is not beating that, the metrics below will show where the hours are going.
The five metrics that cover most of what you need
Run the program on five numbers.
- Units per hour (UPH). Units handled divided by labor hours in that function. Best for receiving, putaway and replenishment.
- Lines per hour (LPH). Order lines completed divided by labor hours. Best for picking, because a line is one trip to one location no matter how many units come off the shelf.
- Orders per hour (OPH). Orders completed divided by labor hours. Best for packing and shipping.
- Cost per line. Fully loaded labor cost divided by lines shipped. The number finance recognizes, and the one that ties the floor to the invoice.
- Direct labor utilization. Direct task hours divided by total paid hours. This explains the gap between “my pickers hit 120 LPH” and “my labor cost per order went up.”
How to calculate each one without fooling yourself
The formulas are simple. The denominators are where people go wrong. Write this table into your KPI glossary as the definition of record.
| Metric | Formula | Use for |
|---|---|---|
| Units per hour | Units processed ÷ direct hours in the function | Receiving, putaway, replenishment |
| Lines per hour | Lines picked ÷ direct picking hours | Picking (each, case, pallet) |
| Orders per hour | Orders packed or shipped ÷ direct hours | Packing, shipping |
| Cost per line | Loaded labor cost ÷ lines shipped | Monthly P&L, client pricing |
| Direct utilization | Direct hours ÷ total paid hours | Every function, every shift |
Two rules keep the numbers honest. First, the denominator is paid hours from the time clock, never scanned hours from the WMS. A picker who scans for six hours and is paid for eight was 75% utilized, and that other quarter of a shift has to show up somewhere. Second, measure by function. A UPH that mixes pallet receiving with each picking is not actionable.
Why lines per hour beats units per hour for picking
A pick is a trip. The picker travels to a location, confirms it, and takes product. Whether it yields one unit or twelve, the trip costs about the same. UPH rewards whoever got the deep lines that day. LPH measures the work.
Here is the test. Pull a week of pick data and sort by picker. If the UPH and LPH leaderboards put different people on top, your UPH is tracking order profile, not performance. In a 3PL with several clients on one floor, a client at 1.1 units per line and one at 4.0 will make the same picker look slow one day and fast the next.
If you want one picking number that survives profile shifts, weight lines by pick type (each, case, pallet) and report a weighted LPH. That is the first step toward engineered labor standards, but you do not need standards in place to start reporting it.
Cost per line: the metric that ties the floor to the P&L
Start with a wage you can source. The BLS reports average hourly earnings in warehousing and storage (NAICS 493) of $26.74 for all employees and $26.02 for production and nonsupervisory employees in August 2026, with production employees averaging 39.3 hours a week (BLS, Warehousing and Storage: NAICS 493, preliminary). Median 2025 wages on the same page run from $21.49 an hour for stock clerks and order fillers to $23.16 for industrial truck operators. Your own payroll is the input; these tell you whether your rate is in range.
Now a worked example with round numbers, say a 40,000-order month at 2.5 lines per order, which is 100,000 lines:
- Direct picking, packing and shipping hours for the month: 9,000.
- Indirect and support hours (receiving, putaway, replenishment, cycle counts, leads, cleanup): 4,000.
- Total paid hours: 13,000. Direct utilization on outbound work is 9,000 ÷ 13,000, about 69%.
- Average wage $26.00, plus an assumed 30% burden for taxes, benefits and temp markups, gives a loaded rate of $33.80.
- Loaded labor cost: 13,000 × $33.80 = $439,400.
- Cost per line: $439,400 ÷ 100,000 = $4.39. Cost per order: $439,400 ÷ 40,000 = $10.99.
Outbound ran 100,000 lines in 9,000 hours, and cost per line is still high because 31% of paid hours were indirect. Chasing pick rate would be the wrong move. Fixing indirect hours, usually replenishment timing, receiving backlog and lead ratios, is where the money is. You only see that when cost per line uses total paid hours.
What the public data says, and why it is not your benchmark
Industry productivity numbers are a direction, not a target. The BLS series for warehousing and storage shows labor productivity falling 11.5% in 2021, 11.0% in 2022 and 4.4% in 2023, then flat at 0.1% in 2024 (BLS, NAICS 493). The 2025 results show a 1.4% gain, with output up 1.8%, hours up 0.4% and unit labor costs up 0.6% (BLS, Productivity and Costs by Industry, 2025).
That does not tell you what your pick rate should be. A building shipping single-unit apparel orders and one shipping mixed pallets to retail DCs are both “in the industry” and their LPH will differ by a factor of ten. Your benchmark is your own trailing 13 weeks, by function and shift.
The one external check worth running is on rate. Production and nonsupervisory earnings in transportation and warehousing reached $31.50 an hour in September 2026 (BLS Employment Situation, preliminary). If cost per line is rising and your wage tracks the market, the problem is hours. Rate is a market problem. Hours are an operations problem, and the one you can fix on the floor.
Measure by function, not by building
A plant-wide number hides the function that is bleeding hours. Give each one the metric that fits.
| Function | Primary metric | Watch alongside | A bad number usually means |
|---|---|---|---|
| Receiving | Units or cases per hour | Dock-to-stock hours | ASN quality, trailer scheduling, staging congestion |
| Putaway | Pallets or cartons per hour | Travel distance | Slotting, directed putaway rules, equipment |
| Replenishment | Replens per hour | Pick-face stockouts per shift | Min/max settings, wave timing |
| Picking | Weighted lines per hour | Pick accuracy | Slotting, pick path, batch size, profile shift |
| Packing | Orders per hour | Cartons per order | Station layout, cartonization, dunnage |
| Shipping | Orders per hour | Carrier cut-off misses | Manifest timing, dock door assignment |
| All functions | Direct utilization | Overtime as % of hours | Scheduling, cross-training, absenteeism |
Overtime as a share of hours is the early warning that your labor plan missed the forecast. Pick accuracy sits next to LPH so nobody buys rate with errors. If LPH rises and accuracy falls in the same week, you moved cost to returns and customer service.
Mistakes that wreck the numbers
- Blended UPH as the headline. It moves with order profile, not performance, and hides the function that is losing hours.
- Scanned hours as the denominator. The WMS only knows time between scans. Breaks, meetings and waiting on replenishment are all paid and all invisible unless you use the time clock.
- Letting rate stand in for safety. The BLS reports 4.8 recordable injuries per 100 full-time workers in warehousing and storage for 2024 (BLS, NAICS 493). A building that pushes rate without tracking recordables is trading one cost for another.
A 30-day plan to put these metrics in place
- Days 1–5: define the denominators. Pull time-clock hours by employee and map each one to a function and shift. Decide once how floaters and leads are counted.
- Days 6–10: pull the numerators. Units received, lines picked, orders packed and shipped, by day, from the WMS. Tag each pick line with pick type so you can weight LPH.
- Days 11–15: build the baseline. Four weeks of all five metrics by function. No targets yet.
- Days 16–20: find the biggest gap. It is almost never pick rate. Look at utilization and indirect hours first, then overtime, then the widest shift-to-shift spread.
- Days 21–25: set one target per function. Use your best consistent shift from the baseline, not an industry number.
- Days 26–30: start the weekly review. Fifteen minutes, one page, five metrics by function against the trailing 13 weeks. Ask what changed and who owns the fix.
Brands evaluating a 3PL can ask for that same page. A provider that shows LPH and cost per line by client, by week, is running the floor on numbers. One that can only quote a plant-wide UPH is not, and your 3PL scorecard should reflect that.
Frequently asked questions
What is a good units per hour rate in a warehouse?
There is no single good number, because UPH depends on pick type, order profile, equipment and slotting. Compare each function against its own trailing 13-week baseline and use industry figures only as a check on direction.
Should I track lines per hour or units per hour for picking?
Lines per hour, weighted by pick type. A line is one trip to one location, so it measures the work the picker did. Units per hour mostly measures how deep the orders were that day.
How do I calculate cost per line in a warehouse?
Divide fully loaded labor cost for the period (wages plus taxes, benefits, temp markups and overtime premiums, across all paid hours) by order lines shipped. Base wage alone or direct hours alone understates it.
Why did productivity go up while labor cost per order went up too?
Usually because the productivity number used scanned hours or covered only pickers, while the cost number used all paid hours. Direct utilization and cost per line will show where the gap is.
If your labor numbers and your P&L tell two different stories, that is a measurement problem before it is a performance problem, and it is fixable in about a month. 3PL operations advisory is where I help operators build the metric set, find the hours and put a weekly review in place that holds.



