4. Modes of freight — the decision surface
Picture three shipments leaving the same factory cluster in Vietnam. The first is a single pallet of pharmaceutical samples that has to be in a US lab in 48 hours. The second is 1,500 cartons of finished apparel for a retail program that lands on shelves in 10 weeks. The third is 200 crates of dense industrial parts for a plant in Mexico.
Same origin. Three completely different freight decisions. The first one is air express, no question. The second is ocean (almost certainly via a buyer’s consolidator). The third is probably cross-border road from a different supplier entirely; if it stays Vietnamese, it’s ocean again with a Mexican-coast port-of-entry.
Mode selection is usually framed as a four-variable tradeoff (cost, speed, reliability, capacity) with a little table showing that air is fast/expensive and ocean is slow/cheap. In practice, the decision is much more constrained. Commodity class, value density, shipment size, origin-destination geometry, customs complexity, and carbon targets narrow the set of viable modes long before cost enters the conversation.
This page lays out every major mode on the same axes and gives you a filter to surface which ones actually fit a given lane profile. The deep-dive chapters that follow (Ocean, Air, Road, etc.) go into the economics of each.
Starting cold? Most cross-border lanes have only 2–3 modes that fit at all once you account for shipment size, urgency, and commodity. The filter below narrows the list to what’s actually viable for your lane.
The modes, at a glance
Section titled “The modes, at a glance”There are six canonical modes plus several hybrids and specialties. Treat the six as primitives; real lanes almost always combine two or more.
- Ocean: bulk, FCL , LCL , reefer, Ro-ro , breakbulk, project cargo.
- Air: general cargo, express / integrator, charter, sea-air hybrid.
- Road: FTL , LTL , partial, drayage, cross-border, dedicated.
- Rail: intermodal (containerized), carload, unit train, international intermodal ( IPI , MLB ).
- Parcel / courier: integrator small-parcel networks, postal, same-day.
- Pipeline: crude, refined products, natgas, specialty (ammonia, CO₂, hydrogen).
And the hybrids worth naming:
- Intermodal: any two-mode move using standardized containers (almost always ocean+rail, or rail+truck).
- Multimodal: three or more modes under a single BoL.
- Sea-air: ocean to a hub (typically Dubai, Los Angeles, or Singapore), air to final destination. Cuts 40% of ocean time at 50% of air cost.
- Cross-border trucking: its own operational animal due to customs at the line (US-MX, US-CA, intra-EU post-Brexit GB-EU).
Use the filter
Section titled “Use the filter”Select the attributes of your lane. Modes that fit every selected filter stay highlighted; modes that violate any one get dimmed. This is a first-pass screen, not a final recommendation; the deep-dive pages cover the exceptions.
| Mode | Typical cost ($/kg, intercon.) | Typical transit | Reliability (CoV) | gCO₂e / tonne-km | Best for |
|---|---|---|---|---|---|
| Ocean FCL | $0.10–$0.40 | 14–45 days port-to-port | Low (CoV 20–40%) | 8 | Full-load, non-urgent, intercontinental, any commodity |
| Ocean LCL | $0.20–$0.80 | 21–55 days door-to-door | Lowest (CoV 30–50%) | 15 | < 15 CBM, non-urgent, intercontinental |
| Air general cargo | $3–$8 | 3–7 days door-to-door | Mid (CoV 10–20%) | 500 | High-value, time-sensitive, intercontinental |
| Air express / integrator | $6–$20 | 1–3 days door-to-door | High (CoV 5–10%) | 1000 | Sub-48h critical, samples, spares, documents |
| Road FTL | $0.04–$0.15 (intra-regional) | Same-day to 5 days | High (CoV 8–15%) | 60–80 | Full trailer, intra-continental, any commodity |
| Road LTL | $0.10–$0.50 (intra-regional) | 2–7 days | Mid (CoV 15–25%) | 100 | 150–10,000 lb, intra-continental, palletized |
| Rail intermodal | $0.03–$0.10 | 4–10 days (domestic) | Mid (CoV 15–25%) | 20 | Long-haul cross-continental, non-urgent |
| Rail carload / unit train | $0.02–$0.06 | 5–15 days | Mid (CoV 15–30%) | 20 | Bulk commodities (grain, coal, chemicals), unit-volume |
| Parcel / courier | $5–$30 per parcel | Same-day to 5 days | Highest (CoV 3–8%) | 600 | < 70 kg per piece, high-velocity, high-visibility |
| Ro-ro | $0.08–$0.25 | 14–40 days | Low (CoV 20–35%) | 10 | Wheeled cargo: vehicles, machinery, trailers |
| Breakbulk / project | $0.15–$1.00+ | 21–60+ days | Low (CoV 25–50%) | 10 | Oversize, indivisible loads that won’t containerize |
| Pipeline | $0.005–$0.02 | Continuous (days in flight) | Highest (once commissioned) | 5 | Crude, refined products, natgas, fixed point-to-point |
| Sea-air hybrid | $1.50–$4 | 10–18 days | Mid (CoV 15–25%) | 250 | Mid-value goods where ocean is too slow and air is too expensive |
No modes match every filter. Try relaxing one. Most commonly, shipment size or commodity are over-constrained when this happens.
How to read the axes
Section titled “How to read the axes”A few axes need unpacking:
- Reliability ( CoV ) is the coefficient of variation of transit time (the standard deviation divided by the mean). Low CoV (5–10%) means you can plan against the mean; high CoV (30%+) means you need material safety-stock buffer, which is itself a hidden cost.
- gCO₂e / tonne-km is a well-to-wake figure using GLEC Framework methodology [GLEC Framework]. Real emissions depend heavily on load factor, routing, and fuel. Treat these as orders of magnitude, not point estimates.
- Value density is the commodity-level axis that matters more than most planners realize. A
$1,500/kgelectronics shipment pays air freight with pocket change; a$1/kgcommodity chemical can’t justify even ocean LCL on a short hop.
The decision geometry
Section titled “The decision geometry”Once you’ve filtered, the remaining choice is usually a tradeoff between two surviving modes. Three recurring pairs:
- Ocean FCL vs. Rail intermodal. For US or EU inland destinations, the ocean container can ride rail inland (intermodal) or transload to truck. Rail is cheaper and greener, truck is faster and more reliable.
- Air general vs. Sea-air. For mid-value goods that can tolerate 10–14 days but not 30+, sea-air splits the difference at roughly half the cost of full air.
- LTL vs. Parcel. The breakpoint sits around 150–300 kg for a multi-piece shipment, but dim weight rules can push small-but-bulky shipments out of parcel economics entirely.
We unpack each of these in the mode deep-dives.
Putting it together
Section titled “Putting it together”Mode selection is a constraint-satisfaction problem disguised as an optimization problem. Most lanes collapse to two viable modes once you filter on commodity, size, urgency, and geography. The remaining judgment call is usually about what risk you’re willing to carry — transit-time variance, single-mode dependency, carrier concentration, or carbon exposure.
How to think about your own lanes
Section titled “How to think about your own lanes”Five decisions worth revisiting:
When you inherited a lane structure 5+ years ago: rerun the filter with current commodity, urgency, and volume profile. A meaningful share of lanes are still on the mode they were assigned to under different constraints (different volume, different sourcing geography, different fuel prices).
When you’re choosing between Ocean FCL and rail intermodal on a long inland leg: rail saves 20–35% on freight at the cost of 2–5 more transit days and slightly higher CoV. The math depends on your inventory carrying cost and your tolerance for variance.
When air is the reflex pick: ask whether sea-air via Dubai, LAX, or Singapore would split the difference. Sea-air runs 40–55% of full air cost at 50–70% of full ocean transit. Most forwarders don’t proactively quote it.
When you’re sizing the LTL/parcel boundary: the breakpoint is around 150–300 lb per multi-piece shipment, but dim weight rules can push small-but-bulky shipments out of parcel economics entirely.
When you’re picking a mode for a new origin-destination pair: start with the filter, then check whether anything obviously disqualifies a mode (cabotage on cross-border road, customs complexity on air, ramp positioning on rail). The filter is a first screen; the deep-dive chapters cover the exceptions.