Quick Answer: Goods-to-person (G2P) delivers 3-5x higher pick rates and 99.9%+ accuracy but requires $2M-$15M+ in capital. Person-to-goods (P2G) with AMR assistance costs 70-80% less upfront and works well under 5,000 orders per day. Most mid-size operations should start with AMR-assisted P2G and add G2P for high-velocity SKUs as volume grows.
The Core Distinction
Person-to-goods (P2G): Workers walk to inventory locations, pick items, and bring them to packing stations. Traditional warehouse operation. Even with AMR assistance (robots carry totes while guiding pickers), the human still walks — just more efficiently.
Goods-to-person (G2P): Automated systems retrieve inventory and deliver it to stationary pick stations. Workers stand at ergonomic workstations, pick items as they're presented, and the system returns inventory to storage. Walking time: zero.
The difference seems simple. The operational implications are enormous.
Performance Comparison
| Metric | Person-to-Goods (Manual) | P2G with AMR Assist | Goods-to-Person | |--------|--------------------------|--------------------|--------------------| | Picks per hour per worker | 60-100 | 150-250 | 300-600 | | Pick accuracy | 97-99% | 99.5-99.8% | 99.9-99.97% | | Walking as % of labor time | 50-65% | 20-30% | 0% | | Space utilization | 40-55% | 45-60% | 75-90% | | Capital cost (50K sq ft) | $50K-$200K | $300K-$800K | $2M-$8M | | Labor cost per pick | $0.25-$0.45 | $0.12-$0.22 | $0.06-$0.15 | | Deployment time | 1-4 weeks | 4-12 weeks | 6-18 months | | Scalability | Add workers | Add robots | Add modules/stations |
When Person-to-Goods Wins
Lower Volume Operations
Below 2,000-3,000 orders per day, the capital investment for G2P systems rarely pencils out. A 25,000 sq ft facility processing 1,500 orders daily can operate efficiently with 8-12 workers using AMR-assisted picking at a capital cost of $250,000-$500,000. The same throughput via G2P would require $1.5M-$3M in automation.
High SKU Variety with Low Velocity
If you carry 50,000+ SKUs but most items ship fewer than 5 units per day, G2P systems spend more time shuttling inventory than enabling picks. The retrieval-to-pick ratio becomes inefficient. P2G lets pickers batch multiple slow-moving SKUs in a single walk path.
Facilities with Irregular Products
Oversized items, non-conveyable goods, and products requiring special handling (hazmat, fragile, refrigerated) often don't fit G2P system constraints. Most shuttle-based and cube storage G2P systems have strict dimensional and weight limits — typically under 35 kg and under 600mm × 400mm × 350mm.
Rapid Deployment Needs
If you need automation operational before peak season in 8 weeks, P2G with AMR assistance is your only realistic option. G2P systems require 6-18 months for design, manufacturing, installation, and commissioning.
When Goods-to-Person Wins
High-Volume Fulfillment
Above 5,000 orders per day, G2P systems dominate the economics. A single G2P pick station served by a cube storage or shuttle system processes 300-600 picks per hour — replacing 3-6 manual pickers. At scale, the capital investment amortizes rapidly.
Example: A 100,000 sq ft e-commerce facility processing 15,000 orders/day with 40,000 SKUs:
- P2G with AMRs: 65 pickers across two shifts, $1.2M in AMRs, $3.8M annual labor
- G2P (cube storage): 18 pick station operators across two shifts, $6M system cost, $1.1M annual labor
- G2P breakeven: 26 months. Year 3+ savings: $2.5M annually.
Labor-Constrained Markets
When you can't hire enough pickers — and in 2026, this is most markets — G2P lets you do more with fewer people. A facility that needs 60 pickers in a P2G model needs only 15-20 operators with G2P. In markets where warehouse labor costs $18-$25/hour and turnover exceeds 100% annually, reducing headcount by 60-70% is transformative.
Space-Constrained Facilities
G2P systems, particularly cube storage like AutoStore and Exotec Skypod, achieve 75-90% space utilization vs. 40-55% for conventional racking with aisles. If you're running out of room, G2P can double effective storage density — potentially delaying or eliminating the need for a larger facility.
A 50,000 sq ft G2P facility can store what would require 90,000-120,000 sq ft of conventional racking. At $8-$12/sq ft annual lease rates, the space savings alone can justify the investment. Learn more in our Cube Storage Systems Guide.
Accuracy-Critical Operations
If pick errors cost you significantly — pharmaceutical distribution, high-value electronics, automotive parts — G2P accuracy of 99.9-99.97% reduces error costs dramatically. A facility shipping 10,000 lines per day dropping from 99% to 99.95% accuracy eliminates 95 errors daily. At $15-$50 per error (returns processing, reshipping, customer credits), that's $500,000-$1.7M in annual savings.
G2P Technology Options
| Technology | Throughput/Station | Storage Density | Investment (50K sq ft) | Best For | |---|---|---|---|---| | Cube storage (AutoStore, Exotec) | 350-650 picks/hr | Very high (90%+) | $3M-$8M | Small-to-medium items, high SKU count | | Shuttle AS/RS (Dematic, TGW) | 250-500 picks/hr | High (80%+) | $4M-$12M | Medium items, very high throughput | | Pod-carrying AMRs (Geek+, HAI) | 200-350 picks/hr | Medium (65-75%) | $1.5M-$4M | Lower cost G2P entry point | | Carousel/VLM (Kardex, Modula) | 150-250 picks/hr | Medium-high (70-80%) | $200K-$1M | Small parts, tool cribs, low-medium volume |
The Hybrid Approach: Best of Both Worlds
The highest-performing fulfillment centers don't choose one strategy exclusively. They segment inventory by velocity and deploy the right strategy for each segment.
ABC Velocity Segmentation:
- A-SKUs (top 20% by velocity, ~80% of picks): Goods-to-person. These items justify the capital investment because they generate most of the throughput.
- B-SKUs (next 30%, ~15% of picks): AMR-assisted person-to-goods. Moderate velocity doesn't justify G2P capital, but AMRs optimize pick paths.
- C-SKUs (bottom 50%, ~5% of picks): Manual person-to-goods. Low-velocity items in conventional racking picked on demand.
This hybrid model captures 80-90% of the labor savings of full G2P at 40-50% of the capital cost.
Making the Decision
Step 1: Calculate Your Cost Per Pick
Determine your current fully-loaded cost per pick (labor + facilities + error costs + equipment). If it's above $0.25, there's significant room for automation savings.
Step 2: Model Both Scenarios
Use the TCO Calculator to compare 5-year total cost of ownership for P2G with AMRs vs. G2P for your order volume, SKU count, and labor costs.
Step 3: Consider Your Growth Trajectory
If you're at 3,000 orders/day growing 30% annually, you'll cross the G2P threshold within two years. Planning for G2P now — even if you deploy AMR-assisted P2G first — avoids costly rework later.
Step 4: Evaluate Labor Availability
In labor markets with unemployment below 4% and warehouse wages above $20/hour, weight the analysis toward G2P. The ability to grow without proportionally growing headcount is a strategic advantage that pure ROI calculations undervalue.
Use the Robot Finder to explore specific AMR and G2P systems that match your facility's requirements.
Frequently Asked Questions
What is goods-to-person fulfillment?
Goods-to-person (G2P) uses automated systems — robots, shuttles, or AS/RS — to bring inventory to stationary workers at ergonomic pick stations. The worker stays in place while products are delivered, eliminating the walking time that consumes 50-60% of a traditional picker's shift.
Which is more cost effective?
It depends on volume. For operations above 5,000 orders/day, G2P delivers lower cost per pick ($0.08-$0.15 vs $0.18-$0.35) due to 3-5x higher pick rates. Below 2,000 orders/day, AMR-assisted P2G typically offers better ROI due to lower capital requirements.
Can you combine both approaches?
Yes, and it's the recommended approach for most mid-to-large facilities. Use G2P for fast-moving A-SKUs and P2G for slow-moving C-SKUs. This captures most of the labor savings at roughly half the capital cost of full G2P.