Data first. Design second. Always.
The layout is not the product. The layout is the answer to a question: given what this operation needs to do, how should the space be organized to do it efficiently, safely, and with room to grow?
Minimum Data Package
Never start designing without at least 80% of this data:
Order & Throughput:
- 12 months of order history (orders/day, lines/order, units/line, cases/order)
- Order profile by channel (B2B wholesale and DTC have completely different profiles)
- Inbound trailer volumes (receipts/day, pallets/receipt)
- Monthly on-hand inventory snapshots - at peak, not average
SKU Master:
- Number of active SKUs (by channel)
- Item dimensions (L × W × H) and weight per SKU
- Unit of measure, packaging type, special storage requirements
Velocity Data:
- Annual units sold per SKU (drives ABC classification)
- Cubic velocity (units sold × cube per unit) - drives slot sizing
- Seasonal peaks per SKU
Growth & Planning Horizon:
- 5-year growth plan by channel
- Seasonal spike magnitude and duration
Order Profile Analysis: Most Important Dataset
Lines Per Order Distribution
| Lines Per Order | Process Implication |
|---|---|
| Primarily 1-line | Each-pick automation, goods-to-person, or dedicated single-line zones |
| 2-5 lines | Zone picking, batch picking, pick-and-pass systems |
| 6-15 lines | Full warehouse tours, batch/cluster. Travel path optimization critical. |
| Mixed (all) | Multi-channel op - almost certainly needs zone separation |
Real project example: CPG company running B2B wholesale (avg 22 lines/order) and DTC e-commerce (avg 1.8 lines/order) from the same building. Result: dedicated DTC pick module with carton flow and pick-to-light, completely separated from bulk pallet pick lanes. Two operations under one roof, each designed for its actual order profile.
ABC/Pareto Analysis
| Category | % of SKUs | % of Total Picks | Placement |
|---|---|---|---|
| A items | ~20% | ~80% | Prime pick locations - golden zone, nearest shipping |
| B items | ~30% | ~15% | Secondary locations |
| C items | ~50% | ~5% | High-density storage, less-accessible |
The 80/20 Pareto pattern for ABC classification is the standard industry benchmark (Frazelle, World-Class Warehousing and Material Handling 2002; de Koster, Le-Duc & Roodbergen 2007, European Journal of Operational Research 182(2):519-558). Actual ABC percentages vary by operation; always validate against your own velocity data.
Velocity calculation methods:
- Basic: units sold / days (default ABC input)
- Weighted: (units sold × order frequency) / days (sporadic demand)
- Volume-based: (units sold × unit volume) / days (space-constrained ops - use for slot sizing)
- Value-based: (units sold × unit value) / days (loss/shrink risk)
Minimum analysis period: 3-6 months. Always analyze multiple windows for seasonality. A C-item in March may be an A-item in October.
Seasonal Peak Design
Always design to peak, not average.
| Vertical | Peak:Average Ratio | Peak Period |
|---|---|---|
| E-commerce (general) | 2.0-3.5× | Nov-Dec (BFCM/holiday) |
| Toy / gift | 3-5× | Oct-Dec — supported by FreightAmigo 2025 seasonal planning data |
| Apparel | ~2.0× | Spring/Fall + holiday — practitioner estimate for store-replenishment DC; no named benchmark source confirmed |
| Consumer electronics | 2-4× | Nov-Dec |
| Outdoor / garden | 2-3× | Mar-May |
| Food & beverage (seasonal) | 1.3-2× | Varies |
| B2B / Industrial distribution | 1.2-1.5× | Lower peaks; steadier |
Standard rule: Design to peak × projected Year 3-5 growth. Add 15-25% safety stock buffer during peak (supplier lead times stretch).
Functional Area Sizing
Each zone in the facility is sized independently based on its specific throughput driver:
Receiving: Dock door count = f(trailer volume, throughput, dwell time, turn time). Use the Dock Door Calculator.
Storage: Storage positions = f(inventory at peak, SKU count, pallet size, rack type, cube utilization target). Always size to 85% utilization at peak.
Pick area: Forward pick sizing = f(SKU count in forward, days of supply target, cubic velocity per SKU, replenishment frequency).
Pack/VAS: Workstation count = f(orders/hr, pack time per order, ergonomic constraints).
Shipping staging: Staging lanes = f(trailer volume, staging dwell time, outbound door count).
Planning Horizons by Investment Type
Physical infrastructure (building shell, dock doors, utilities, rack columns) → size to Year 5 peak. These are the hardest and most expensive to change. A dock door costs $40-80K to add post-construction. (Practitioner estimate; cost varies significantly by region, building construction type, and whether a pit leveler, HVLS fan, and dock seal are included. Get contractor quotes for any specific project.)
Automation and mechanized systems (conveyors, sorters, AS/RS) → size to Year 3. Technology changes; don’t over-commit 5 years out.
Labor standards and process design → size to Year 3 peak. Labor is the most flexible resource.
Greenfield benchmark: Nestlé USA’s Arvin, CA DC (700,000 sq ft, $330M, opened May 2026) demonstrates these planning principles at CPG scale: crane-aisle AS/RS, laser-guided vehicles, and layer-picking robotics were integrated into the building design before construction. The facility handles shelf-stable goods and was explicitly built for peak-season resiliency without proportional headcount scaling. See Research - Nestle Arvin DC Greenfield Automation 2026.
Design Tool Stack
- Excel - velocity pivot tables, order profile distributions, Pareto analyses, the Data Request Template
- SQL - extract from WMS/ERP. Key tables: ORDER_LINES, INVENTORY
- Power BI / Tableau - peak/valley visualization, ABC charts for stakeholder presentations
- AutoCAD / Visio - facility layout drawings; layering convention matters (rack, MHE, column grid, fire suppression all on separate layers)
- Alteryx - AI-enabled data cleaning for large/messy datasets
WERC Best-in-Class Benchmarks for Layout Sizing
| Metric | Best-in-Class | Median |
|---|---|---|
| Dock-to-stock time | <3 hours | - |
| Warehouse location utilization | 92-95% | - |
| On-time shipping | >99% | - |
| Lines received/put away per hour | >68.9 | - |
| Orders picked/shipped per person per hour | ≥35 | 10 |
The gap between median (10 orders/hr) and best-in-class (35 orders/hr) is where good layout design lives. Zone structure, slotting, and properly sized pick locations drive a significant portion of that difference.
Non-Traditional Layouts
Conventional orthogonal aisles are not the only option. Flying-V and fishbone designs add diagonal cross-aisles aimed at a single pickup/deposit point and cut single-command unit-load travel by 10-20%, with important caveats for AGV navigation. See Non-Traditional Aisle Layouts.
Academic grounding
The layout sizing and functional-area content here aligns with the canonical warehouse-design review literature collected in warehouse-or-literature-canonical, principally Gu, Goetschalckx & McGinnis (2007, 2010).