Order Picking Methods

Picking method selection is driven almost entirely by the order profile - specifically lines per order and the mix of B2C eaches, B2B cases, and full pallets. The wrong method for your order profile produces low productivity no matter how well everything else is executed.

The Five Core Methods

Discrete (Single-Order) Picking

One picker fulfills one order completely from start to finish. Picker travels the full warehouse for each order.

When to use: Low daily order volume. High lines-per-order (6+). Operations where order accuracy is paramount and sortation infrastructure doesn’t exist.

Weakness: Maximum travel distance per picker. Least productive at high order volumes. Travel accounts for 50–60% of total pick time in most manual operations (De Koster, Le-Duc & Roodbergen, 2007, EJOR 182(2):481–501; DOI: 10.1016/j.ejor.2006.07.009). The 70% upper bound appears in practitioner sources but is not confirmed in peer-reviewed literature.

Batch Picking

One picker fulfills multiple orders simultaneously, picking all units of a SKU at once across several orders, then sorts them downstream (put wall or conveyor sortation).

When to use: High daily order volume with low lines per order (1-3 lines typical). Reduces travel by consolidating trips to the same pick location.

Mechanics: WMS groups orders into a batch (typically 12-48 orders). Picker carries multi-tote cart or picks to conveyor induction. Downstream sortation (put wall or automated sorter) separates units back to individual orders.

Batch size tradeoff: Larger batches = more travel reduction per pick, but higher WIP and longer time-to-ship for any individual order. Optimize batch size based on carrier cutoff windows and order cycle time targets.

Wave Picking

Orders are released in coordinated waves across the building - all zones pick simultaneously, and outputs converge at packing/shipping at the same time.

When to use: Multi-zone operations with conveyor sortation. Operations with defined carrier cutoff windows that require synchronized output.

Mechanics: Wave planner (WMS) groups orders by carrier, zone, ship-by time, or order type. Releases all zones simultaneously. Sorter or put-to-light consolidates zone outputs into complete orders.

Wave planning is a discipline unto itself. Poor wave parameters (wrong order grouping, wrong release timing) cause sorter backup, dock congestion, and missed carrier cuts even when all individual picks are on time.

Zone Picking (Pick-and-Pass)

Warehouse divided into geographic zones. Each picker owns one zone. Orders travel through zones sequentially - each picker adds their items, passes to the next zone.

When to use: High lines-per-order operations (6-15 lines). Multi-category facilities where SKUs are naturally separated by zone. Any operation where a single picker completing the full order would require too much building travel.

Two configurations:

• Sequential (pick-and-pass): Order container moves from Zone 1 → Zone 2 → Zone 3. Simple but creates bottlenecks if zones aren’t balanced.
• Simultaneous: All zones pick in parallel; downstream sortation consolidates. Requires sortation infrastructure but eliminates zone-balancing bottleneck.

Cluster Picking

Picker carries multiple orders simultaneously in a cart with individual tote slots (one slot per order). Picks are directed by RF/voice/light to the correct tote without downstream sortation.

When to use: Low lines-per-order (1-4 lines). Each-pick operations where sortation infrastructure isn’t available. Bridge between discrete (too slow) and batch (needs sorter).

Cluster size: Typically 4-12 orders per cluster depending on cart design and average order cube. WMS optimizes cluster grouping based on geographic proximity of pick locations.

Bucket Brigades (Self-Balancing)

A self-balancing variant rather than a sixth fixed method: workers sequenced slowest to fastest on a flow line hand off orders so the line finds its own optimal balance, with no engineered standards or zone re-balancing. A real DC recorded a 34% productivity gain after adopting them. Best when SKU velocity churns (no re-slotting needed to keep zones balanced) or labor turnover is high. See Bucket Brigade Order Picking.

Method Selection by Order Profile

The cardinal rule: Never run high-volume 1-line B2C orders through the same zone structure as 22-line B2B wholesale orders. They need different methods, different equipment, and often different physical zones.

Productivity Benchmarks by Method

Slotting Impact on Picking

Pick method and slotting are inseparable. ABC slotting in a 1,000-foot warehouse — illustrative example based on de Koster, Le-Duc & Roodbergen (2007) travel-time modeling:

• Unslotted: ~1,000 ft average travel per pick
• ABC slotted: ~340 ft average travel per pick
• Result: 66% travel reduction, same people, same equipment, different slot assignments (Illustrative — actual reduction depends on warehouse geometry and ABC velocity curve. The directional result — A-item proximity to the I/O point materially reduces picker travel — is supported by peer-reviewed travel-distance models.)

A-items in the golden zone (30”-60” shelf height) produce 10-20% higher pick productivity vs. floor or top-shelf picks (Petersen, Siu & Heiser 2005, International Journal of Operations & Production Management 25(1):17-34; DOI: 10.1108/01443570510619491). Every slotting decision compounds across thousands of picks per shift.

Goods-to-Person vs. Traditional Picking: ROI Comparison

Goods-to-person (G2P) brings the product to a stationary picker via automated storage (AS/RS, cube storage, shuttle). Traditional zone or batch picking requires the picker to travel to the product.

Productivity comparison:

G2P achieves 2–4× the productivity of modernized RF-scan discrete picking, primarily by eliminating travel (inVia Robotics goods-to-person benchmark: 2–4× documented across configurations; GoASRS 2026). Compared to very-low-baseline manual picking (60-100 picks/hr in high-travel environments), G2P achieves 3–6×. Actual improvement depends on the facility’s baseline picking productivity.

ROI decision framework:

Realistic payback for G2P systems (2026):

• AutoStore / cube storage: 2–4 years (Kardex 2024 verified range: 2.6–3.7 years; individual cases as fast as 18 months depending on labor rates and throughput utilization)
• Mini-load AS/RS (goods-to-person): 4–6 years
• AMR-based G2P (pods-to-person): 2–4 years (lower capital cost)

G2P is a labor-efficiency play, not a labor-elimination play at most operations. Maintenance, WMS, and WCS overhead increases. See Warehouse Execution System for software stack requirements.

Pick Zone Technology Sizing: PTL vs. Voice vs. RF

Head-to-head comparison:

Zone sizing methodology:

Pickers needed = Zone throughput required (uph) / Technology UPH benchmark

Example (PTL zone):
  Zone required throughput: 1,800 uph
  PTL benchmark: 280 uph per picker
  Pickers needed: 1,800 / 280 = 6.4 -> plan for 7 (buffer for breaks, exceptions)

PTL module count rule: One light module per active pick face. For 150 active SKU locations, you need 150 modules. Size the zone to have no more than 15–20% inactive locations at any time - above that, re-slot or relocate slow movers.

Voice cold chain: Voice is the standard for frozen/cold storage picking. Scan gloves are impractical at sub-zero temperatures. Voice profiles require 20–40 min of training per picker; factor into new hire onboarding.

RF accuracy requirement: Scan-first confirmation (scan location barcode, then scan item) is required to achieve 99%+ accuracy. Item-only scanning produces 98–99% at best.

Academic grounding

The picking methods and travel-time figures on this page are grounded in the peer-reviewed reviews collected in warehouse-or-literature-canonical, principally de Koster, Le-Duc & Roodbergen (2007) for manual picking and Boysen, de Koster & Weidinger (2019) for e-commerce picking.