The warehouse KPIs worth tracking are the ones that tie to your current business problem. For most operations that means on-time shipment, order accuracy, dock-to-stock time, inventory accuracy, labor cost per unit and capacity utilization. Pick five to eight, give each one an owner and a target, and review them on a fixed cadence. Anything beyond that is usually reporting, not management.
This guide covers the metrics, formulas and benchmarks for each stage of the building, from receiving to shipping. It also shows how to choose a short list and keep it from turning into a dashboard nobody reads.
TL;DR
- Start with 5 to 8 KPIs that map to one business goal. Margin problem: labor cost per unit and inventory carrying cost. Retention problem: order accuracy and on-time shipment.
- Current best-in-class marks from the 2025 WERC DC Measures Report: on-time shipments at 99.5% or better, order-picking accuracy at 99.68% or better, dock-to-stock under 3.5 hours.
- A KPI earns its place only if a bad number triggers a specific action. If it doesn't, it's noise.
- Benchmark against your own history first, then against industry figures.
- Every KPI needs a named owner, a target and a weekly review, or it becomes an orphan metric.
Table of Contents
- What a warehouse KPI is for
- The metrics the industry watches most
- KPIs by stage of the warehouse
- What bad accuracy costs
- How to choose your 5 to 8
- Perspective: why most KPI programs stall
- Where TKD Consulting fits
- Making the numbers stick
- FAQ
- Sources
What a warehouse KPI is for
A warehouse KPI is a number that measures how well a facility receives, stores, picks, packs and ships. The useful ones work as diagnostics. Zoho's inventory guide makes the point that a KPI is only valuable if it analyzes a specific process and can be compared with your own historical performance. If the number doesn't tell you to retrain pickers or re-lay the dock, it isn't doing a KPI's job.
That is also why a list of 30 or 40 metrics, which is what most top-ranking articles offer, is a poor starting point. A list is a menu. What you need is a short dashboard built around a decision.
The metrics the industry watches most
The 2025 WERC DC Measures Report is the most widely used benchmarking survey for distribution centers in the US. It ranks the most critical metrics, with these best-in-class marks:
| Metric | Best-in-class benchmark (WERC 2025) |
|---|---|
| On-time shipments | 99.5% or higher |
| Average warehouse capacity used | 90% or higher |
| Order-picking accuracy | 99.68% or higher |
| Dock-to-stock cycle time | Under 3.5 hours |
Two cautions when you read these. First, best-in-class means the top performers in the survey, not the average. A facility at 98.5% on-time shipment is not failing, but it has room. Second, the benchmarks move. WERC notes that order-picking accuracy's best-in-class mark dipped slightly to 99.68%, which shows how hard it is to hold near-perfect numbers as complexity rises. Don't panic over small slippage in an already elite metric. Focus on process stability.
The same report shows how priorities shift. Total order cycle time jumped from 20th in importance in 2024 to 8th in 2025. WERC frames the lesson as combining vision with vigilance: re-check regularly which KPIs matter to the current strategy instead of defending a legacy dashboard.
KPIs by stage of the warehouse
Organizing metrics by where the work happens makes it clear who owns what.
Receiving
Receiving sets the quality of everything downstream. A typical flow runs like this: the supplier sends an advance ship notice, the carrier books a dock appointment, the team inspects the trailer and unloads, counts and checks against the purchase order, records discrepancies, labels and puts away, and the system updates inventory. Each handoff is a place where time and accuracy leak. Three KPIs cover it:
- Dock-to-stock time. Hours from the truck arriving (or the unload starting, define it once) to stock being available in the system and on the shelf. WERC's best-in-class mark is under 3.5 hours. ReadySpaces uses a looser under-24-hours target for smaller operations.
- Receiving accuracy. Share of receipts that match the PO on item, quantity and condition. Track discrepancies by supplier so you have evidence for a vendor conversation.
- Dock appointment adherence. Trucks arriving in their window divided by trucks scheduled. Low adherence usually explains idle receiving labor in the morning and a pile-up in the afternoon.
Dock-to-stock deserves extra attention. When it runs long, your system may show inventory that sales can't yet use. That creates phantom stockouts: the product is in the building, but the customer sees a delay or a cancellation. High performers prioritize same-day availability for critical SKUs.
A simple receiving checklist that supports these KPIs: confirm ASN and PO match, check the seal and trailer condition, count and inspect against the PO, log discrepancies with photos before the driver leaves, label, put away, and confirm the system quantity updates the same shift.
Storage and inventory
- Inventory accuracy. System quantity versus physical count, by location. ReadySpaces cites a working range of 95% to 99% for smaller operations. Cycle counting is how you hold it; see our guide to cycle counting best practices.
- Capacity used. Average and peak space used. WERC's top performers average 90% or higher. Run much above that at peak and congestion starts to cost you in travel time and errors.
- Inventory carrying cost and turnover. Storage, capital, insurance, obsolescence and shrink, relative to inventory value. These matter most when revenue is growing but margin isn't.
Picking, packing and shipping
- Order-picking accuracy. Lines picked correctly divided by lines picked. WERC's best-in-class is 99.68% or higher.
- Order accuracy rate. Orders shipped with the right items, quantities and address, divided by total orders shipped. BoxWise puts the best-in-class target at 99.5% or higher.
- Pick rate. Lines per labor hour. ReadySpaces gives a general range of 60 to 120 lines per hour, but your number depends on SKU profile, slotting and method. Compare against your own trend, and see our comparison of batch and wave picking if the number is stuck.
- On-time shipment rate. Orders leaving by the committed cutoff divided by orders due. WERC's top benchmark is 99.5% or higher.
- Order cycle time. Order release to shipment confirmation. It tells you whether the building flows or batches up.
Labor and cost
- Labor cost per unit (or per order). Total warehouse labor cost divided by units shipped. This is the first number to watch when margin is the problem.
- Units per labor hour. Productivity measured against standard. Pair it with an accuracy measure, or you reward speed that creates errors.
- Overtime percentage. Often the earliest sign that staffing doesn't match the forecast. Our warehouse labor planning guide shows how to turn a forecast into a staffing plan.
Safety
Safety numbers are increasingly a compliance matter, not just a culture one. Minnesota, for example, has a law for warehouse distribution centers with 250 or more employees at one site, or 1,000 or more across the state. The state's OSHA annual report describes it. The law requires employers to track injury rates, and a rate at least 30% above the NAICS-code average can prompt a state investigation. That rule is Minnesota's, not national, but it shows where regulation is heading. Track recordable incident rate and near-miss reports whether or not your state requires it.
What bad accuracy costs
Order accuracy is the metric where small percentages hide large dollars. BoxWise estimates that a 1% error rate on 100 orders a day means about 365 errors a year, and each can cost $15 to $60 once you count return freight, rework, replacement product and customer service time. Run that against your own volume.
That math is why picking errors deserve a root-cause review, not another reminder email. We cover the practical fixes in Picking Errors: What Warehouse Managers Can Fix This Week. If you are also trying to see where manual rework and handoffs hide in the process, this overview of automating business processes is a reasonable place to start thinking about what to automate first.
How to choose your 5 to 8
The mistake we see most often is tracking everything available. ReadySpaces and other practitioner guides converge on a small set tied to the current goal. A working method:
- Name the business problem in one sentence. "Revenue is up 20% but margin is flat" or "we're losing repeat customers to late shipments."
- Map it to the process. Margin problems usually sit in labor, carrying cost and rework. Retention problems sit in accuracy and on-time shipment. Cash problems often sit in inventory and dock-to-stock.
- Pick one lagging and one leading KPI per problem. On-time shipment lags; dock-to-stock and pick backlog lead.
- Set the baseline from your own last 90 days. Industry benchmarks come second. You can only judge improvement against yourself.
- Write the trigger. For each KPI, state the number that forces a response and what the response is. "If pick accuracy falls below 99.3% for two days, the shift lead audits the three most error-prone zones."
- Assign one owner. Not a department. A person.
For how to build the ownership side, see our 90-day KPI ownership plan. To connect the warehouse numbers to company-level goals, read about cascading KPIs.
Example dashboards
| Business problem | Core KPIs |
|---|---|
| Margin is flat while revenue grows | Labor cost per unit, overtime %, inventory carrying cost, order accuracy, capacity used |
| Customers complain about late or wrong orders | On-time shipment, order accuracy, pick accuracy, order cycle time, dock-to-stock |
| Stockouts despite full shelves | Dock-to-stock, inventory accuracy, receiving accuracy, appointment adherence |
| Peak season strain | Capacity used (average and peak), units per labor hour, overtime %, safety incident rate |
If you also track cost at the job or order level, the discipline is similar to what contractors use when they compare field data against WIP; this job costing article shows how cost data collected in real time changes the conversation about margin.
Perspective: why most KPI programs stall
We don't think warehouses have a measurement problem. Most have plenty of numbers. The failure we see is between the number and the person on the floor. The report goes to the executive team monthly, nobody on the dock can say what the target is, and the supervisor who could change the result never sees it until it's too late.
That comes from running distribution centers and carrying a P&L, not just diagramming them. A KPI that matters is one a floor manager can read at the start of the shift, understand in ten seconds and act on before lunch. Handing a manager 30 metrics on Monday morning guarantees nothing gets fixed. Handing them five, with targets and a named trigger, gets work done.
So our rule is simple. Fewer metrics, shorter feedback loops, and a weekly meeting where the owner explains the number and the action. When a KPI has been red for a month and nothing has changed, we either fix the process or retire the KPI. Leaving it on the dashboard is the worst option.
— David
Where TKD Consulting fits
If your KPIs exist but results don't move, the gap is usually ownership and cadence, not data. Our 90-Day Operations Audit is a fixed-price diagnostic at $3,500 that maps people, processes and performance systems against your goals and ends in a prioritized action plan with scorecards and meeting cadences your team can run. If you'd rather talk first, a paid discovery call is $500. We work with owner-operated and mid-market companies, and the plan is written to be handed to a floor manager.
Making the numbers stick
A few habits separate dashboards that work from those that don't:
- Define each KPI once, in writing. Where does the dock-to-stock clock start? Does a split shipment count as one order or two? Disputes about definitions kill trust.
- Review weekly at the point of work. Monthly reviews are for trends. Weekly and daily reviews are for fixes.
- Show trend and target together. A single number without either tells a supervisor nothing.
- Retire on purpose. Re-check the list each quarter. Priorities change, which is exactly what the WERC data about shifting rankings shows.
- Choose tools after the metrics. Spreadsheets are enough to begin. When you outgrow them, our WMS guide for mid-market operations and our piece on operational dashboards help you decide what to buy. An operations scorecard is a lighter first step.
FAQ
What are the most important warehouse KPIs?
For most operations: on-time shipment, order or picking accuracy, dock-to-stock time, inventory accuracy, labor cost per unit and capacity used. WERC's 2025 ranking puts on-time shipments, average capacity used and order-picking accuracy at the top. Your own top five should follow your current business problem.
How many KPIs should a warehouse track?
Five to eight on the main dashboard. More than that dilutes attention and makes it unclear who owns what. You can keep a deeper set of supporting metrics for diagnosis, but managers should act on the short list.
What is a good order accuracy rate?
Best-in-class targets are 99.5% or higher for order accuracy, and WERC reports 99.68% or higher for order-picking accuracy. Smaller operations often set an interim goal near 99%. Start from your own baseline and measure the cost per error, since that drives how fast you need to improve.
Do distribution centers use different KPIs than warehouses?
The core metrics are the same. A distribution center moving high order volumes puts more weight on cycle time, on-time shipment and pick productivity, while a storage-heavy warehouse leans on capacity, inventory accuracy and carrying cost.
Sources
- WERC 2025 DC Measures Report release
- BoxWise: warehouse KPIs and order accuracy cost
- ReadySpaces: warehouse productivity metrics and KPIs guide
- Zoho Inventory: warehouse KPIs
- Minnesota DLI: State OSHA annual report, FFY 2023