How to Run a Leaner Warehouse (Without Hiring More People)

Volume is up, the hiring budget is not, and the people you do have are already busy. The question on the floor is some version of "how do we get more out of the same crew?"

The short answer: take hours out of the work, not effort out of the people. Find where paid hours actually go, measure it before you change anything, remove the walking, waiting, searching and rework nobody would pay for, and then set a realistic standard for the work that is left. Overtime and "just push harder" are not on that list, for reasons below.

Is warehouse productivity going up or down?

Down for most of the last decade, then partly back. The US Bureau of Labor Statistics (BLS) publishes an index of output per hour for the warehousing and storage industry (NAICS 493), with 2017 = 100. It peaked at 122.6 in 2014, fell to 100 in 2017 and to a low of 65.6 in 2022, then recovered to 70.2 in 2025, still about 30 percent below 2017. Between 2014 and 2022, hours worked in the industry rose about 2.5 times, from 1.50 billion to 3.69 billion. Since 2022 hours have been roughly flat (3.64 billion in 2025) while output per hour partly recovered.

Two cautions. "Output" here is industry revenue adjusted for inflation, not units shipped. And the price measure behind it is built around storage: a 2021 methods paper for the Voorburg Group of national statisticians says US general warehousing prices are an average price per cubic foot, total warehouse revenue divided by occupied storage space. Our reading: a measure built that way may not track labor-heavy fulfillment work well. These figures are BLS releases as published on FRED (productivity updated August 31, 2026; hours June 3, 2026). BLS revises them, and older bls.gov tables can show different numbers.

None of it describes your building. Only your own hours-by-process numbers do that.

Where do the hours actually go on a warehouse floor?

Most of them go to picking, and much of picking is walking. A widely cited review of order-picking research by de Koster, Le-Duc and Roodbergen (European Journal of Operational Research, 2007) estimates the cost of order picking at "as much as 55% of the total warehouse operating expense," and calls it "the most labour-intensive and costly activity for almost every warehouse." Its chart of a typical picker's time, reproduced from a facilities-planning textbook (Tompkins et al.), splits the day into travel, search, pick, setup and other, with travel at about half.

The same review adds the caveat most summaries drop: travel "is often the dominant component," but case studies show "other activities than travel may substantially contribute." A floor with good pick paths can still lose hours to searching for stock that is not where the system says it is, to waiting at the one label printer, or to repacking orders that were picked wrong.

So list what happens in a paid hour, not just in a pick:

• Travel: walking or driving with nothing in hand.
• Search: looking for a location, an item or a tote.
• Waiting: for replenishment, a forklift, a printer, a supervisor's override, the next wave.
• Double handling: product set down in staging and picked up again later.
• Rework: mispicks caught at pack, repacks, relabels, returns caused by your own errors.
• Admin: paperwork, keying, chasing exceptions.
• The work itself: picking, packing, receiving, loading.

Only the last line is what the customer pays for. Some of the rest is unavoidable; the job is to shrink it, biggest category first.

How do you measure where the time goes before cutting anything?

Measure the process first and the person last. Three cheap ways to do it:

1. Hours by process. Split last month's paid hours across receiving, put-away, replenishment, picking, packing, shipping, returns and indirect time. Assign floating or borrowed labor to the process it actually worked, from scans or a sign-in sheet, not to its home department.
2. Units per paid hour, by process. Units, lines or cartons divided by the paid hours of that process, not only the hours someone was scanning. The gap between paid hours and scanning hours is time nobody is measuring.
3. Work sampling. A supervisor walks the floor at pre-generated random times, across every shift, and notes what each person is doing at that moment, using the categories above. Our recommendation is several hundred observations, not a few dozen. People work differently when watched, so run it long enough that the clipboard stops being news.

Tell the crew what you are measuring and why before you start, and record categories, not names. The first goal is to find the waiting and the walking, not to rank people. For the numbers worth tracking afterwards, see warehouse KPIs that matter.

Which wastes cost a warehouse the most hours?

The National Institute of Standards and Technology's manufacturing extension program describes lean manufacturing as "a continuous improvement model that strives to eliminate non-value-added activities." Our translation to a warehouse floor:

• Walking empty. Fast movers stored far from pack, pickers returning to the start after every order. The fix is mostly slotting and batching; see what warehouse slotting is.
• Waiting at the pick face. A location that is empty at pick time means a picker waits or skips. Trigger replenishment on min/max levels or on the next wave's demand, use top-offs where a face cannot hold a whole wave, and keep forklift replenishment out of aisles while pickers are working them.
• Searching for stock. Every "it says it's here but it isn't" costs a walk, a search and often a short shipment. That is mostly an inventory accuracy problem, not a labor problem; see what inventory accuracy is.
• Touching it twice. Product staged on the floor "for now" gets moved again later. Count how many times a pallet is handled between the dock and its rack location.
• Rework at pack. A mispick caught at pack costs the pick, the check, the walk back and the second pick. A mispick that ships costs all of that plus the return.
• Checks nobody reads. A count, signature or second scan that no report uses may be removable, but confirm with QA, compliance and your customers' routing guides first, and track the error rate afterwards.

Most of these can start without new hires or capital, though re-slotting costs hours up front.

Should we just run more overtime instead?

For a peak week, maybe. As the plan for the year, no. BLS reports that all employees in warehousing and storage averaged between 39.4 and 40.7 hours a week from May to August 2026 (August preliminary). A 2004 NIOSH review of 52 studies across industries, reporting findings as the original authors stated them, found that in 16 of 22 studies on general health, overtime was linked to poorer perceived health, more injuries, more illness or higher mortality. It also reported "a pattern of deteriorating performance on psychophysiological tests," with a similar pattern for injuries, strongest with very long shifts and with 12-hour shifts combined with more than 40 hours a week.

Overtime borrows hours from the same people who are already doing the walking and the waiting. It does not remove any of it. For context, the warehousing industry recorded 4.8 injury and illness cases per 100 full-time workers in 2024, per BLS.

Do we need labor standards, and are they legal?

You need a planning number for each process: how many units an hour a trained person can do under normal conditions, with breaks included. Without it, you cannot tell whether next Tuesday needs eight pickers or eleven. Build it from your own measured hours after the obvious waste is out, and update it when the layout, the order mix or the equipment changes.

How you use it matters. The California Labor Commissioner's FAQ (January 2022) on the state's warehouse quota law, AB 701, defines a quota as "a work standard under which an employee is assigned or required to perform at a specified productivity speed, or perform a quantified number of tasks, or to handle or produce a quantified amount of material, within a defined time period and under which the employee may suffer an adverse employment action if they fail to complete the performance standard." The law covers employers that "directly or indirectly control 100 or more employees at a single warehouse distribution center," or 1,000 or more at warehouse distribution centers in California, as defined in Labor Code section 2100. Covered employers must give each employee, and new employees when hired, a written description of each quota. A quota cannot prevent meal or rest periods, "use of bathroom facilities (including reasonable travel time to and from bathrooms)," or compliance with occupational health and safety standards. Current employees can request their quota descriptions and their most recent 90 days of work-speed data, with a response due within 21 days; a former employee gets one request, covering the 90 days before separation. The FAQ says employers cannot retaliate for these requests.

Other states have their own laws, with different thresholds: Washington (RCW chapter 49.84), New York (Labor Law Article 21-A, the Warehouse Worker Protection Act) and Minnesota (section 182.6526, which starts at 250 employees at a single center). Elsewhere, check your state.

Our caution: if a number is shown to employees and missing it could lead to any adverse action, treat it as a quota. Disclose it in writing at hire, make sure breaks, bathroom trips and safety rules fit inside it, and have counsel review it. A planning standard used only for staffing math, and never applied to individuals, is a different thing.

Where does doing more with the same crew stop working?

Lean takes out waste; it does not create capacity that is not there. Signs you have reached the limit:

• Units per paid hour stopped improving after the walking, waiting and rework came out.
• The constraint is physical: dock doors, pack stations, one conveyor, the number of trucks.
• Peaks run well above the average day, so a crew sized for Tuesday is wrong for the week before a holiday.
• Error rates or injuries rise as volume rises.

At that point the answer is more people, another shift or more equipment, and your measurements make the case. BLS puts the 2025 median wage for stock clerks and order fillers in warehousing and storage at $21.49 an hour, before benefits and overtime premiums.

Where do you start on Monday?

1. Weeks 1 and 2: split last month's paid hours by process, and run the work sampling.
2. Week 3: pick the one waste category with the most hours and change one thing. Do not start five projects.
3. The next two to three weeks: leave it alone. Compare units per paid hour for that process on like-for-like days, and track lines per order alongside it, because a shift in order mix can move units per hour on its own.
4. Then: set or adjust the planning standard and pick the next category.

Start with the hours between the work, not the people doing it.