Why freight emissions numbers can still mislead, and how to avoid it
The false confidence problem
At most mid-market shippers, the person putting together the freight emissions number isn't a dedicated sustainability team, it's someone in operations, procurement, or logistics who has this on top of their actual job, working from carrier invoices, a spreadsheet, and whatever figure a forwarder sent over. Increasingly that number gets asked for anyway: a large customer needing it for their own Scope 3 disclosure, or a single voluntary emissions claim in a tender that's enough to bring a company into scope under CountEmissionsEU.
A dashboard showing 1,247 tonnes CO₂e across 3,400 shipments looks exactly like what that effort should produce: specific, itemised, easy to defend if anyone asks. That specificity is not the same thing as being right. A calculation that applies ISO 14083's rules incorrectly to every one of those 3,400 shipments still produces a precise-looking number, just a wrong one, and there's no in-house methodology reviewer to catch it before a customer's auditor does.
Four gaps account for most of the distance between a calculation that cites ISO 14083 and one that actually applies it. All four are invisible from the inside; they surface only when someone with the standard memorised, an assurance reviewer, sometimes a customer checking their own supply chain, looks for them specifically.
Gap 1: Distance
Say a shipment runs from Piraeus to Rotterdam. The shortest road route cuts through Serbia and North Macedonia, both outside the EU; many carriers instead take a longer route through Bulgaria, Romania, Hungary, and Austria to stay within the EU and avoid customs stops. A generic mapping tool defaults to the shortest option. The GLEC Framework's own preferred road-distance basis, the Shortest Feasible Distance found via route-planning software, explicitly allows for exactly this kind of deviation: actual routes commonly differ from the planner's default to avoid tolls, reach rest points, or, as here, avoid customs delays. The same tool is also the wrong basis entirely for a rail or sea leg, since it only knows road driving. A platform can apply the right distance rule automatically once it knows the mode and the corridor, it just can't know the corridor on its own; that still has to come from the carrier.
Gap 2: Load factors
A half-empty return leg burns nearly as much fuel as a full one, so its emission intensity per tonne-kilometre is far higher once that fuel use is spread across less cargo. The GLEC Framework's default load factors cover this well when nothing better exists, but many calculations keep using the default even once a carrier could supply the truck's actual primary data, discarding real information. A platform can guarantee primary data always overrides a default the moment it's supplied; it can't invent payload data the carrier never captured. A quieter, purely internal version of the same mistake is an outdated emission factor database, which a platform can and should just keep current on its own.
Gap 3: Allocation logic
ISO 14083 splits a shared truck or container between co-loaded customers using mode-specific rules: mass-based for road and rail, TEU-based for container shipping, volume-based for some air freight. The common mistake is applying one rule across every mode in a chain, so a forwarder quoting the same lane to five customers can end up with five different, equally confident-looking totals for what is physically the same shipment. Applying the right rule is calculation logic a platform can automate the moment it knows each leg's mode. Having the right inputs, the full manifest of everyone else's cargo sharing that vehicle, not just the one shipment being invoiced, still depends on the carrier running the consolidated load actually sharing it.
Gap 4: System boundary
ISO 14083 requires the full transport chain: pre-carriage, main carriage, last-mile delivery, and hub operations, not just the leg the primary carrier invoices. Skipping pre-carriage and last-mile is often the single largest source of unaccounted emissions in a reported total, and it stays invisible because nobody asked the smaller operators running those legs for their numbers. A related mistake covers the fuel's energy pathway instead: reporting only TTW (tailpipe combustion) and leaving out WTT (producing and distributing the fuel), which the GLEC Framework's default emission factors put at roughly a quarter of TTW for a diesel HGV (0.029 vs. 0.119 kgCO₂e per tonne-km). A platform can enforce the WTT/TTW rule automatically; closing the missing-leg half of the gap still needs the shipper or their carriers to supply the figure.
Why this matters even without a CSRD mandate
Most shippers reading this aren't required to file a CSRD report themselves. Omnibus narrowed mandatory reporting to companies over 1,000 employees and €450 million in turnover, which rules out most shippers and forwarders in this market, with no confirmed date for that to change. That doesn't make freight emissions data optional. It changes who's asking for it, and why.
Since July 2026, the EU's Voluntary Sustainability Reporting Standard for SMEs (VSME) has applied alongside a legal "value chain cap": a large, CSRD-obligated customer cannot require a supplier with under 1,000 employees to hand over more than the VSME's roughly 50-datapoint Basic Module. That protects suppliers from an ever-expanding, bespoke questionnaire, but it doesn't stop a large customer asking for more voluntarily, and most will keep asking, because they still need shipment-level freight data for their own Scope 3 Category 4 number.
That's where the real leverage sits for an SME shipper or forwarder. An answer ready in the form a large customer or a bank already expects settles that conversation in one email; without one, a company either negotiates a bespoke request from scratch every time, or reads as the least reliable link in someone else's supply chain. Producing a clean, ISO 14083-consistent number here isn't compliance for its own sake, it's what keeps a smaller company easy to work with, not harder.
It's worth being precise about what that claim is and isn't. A correct emissions number is not a lower one, and CO2Path doesn't reduce anyone's emissions, it makes sure the number describing them is one a customer, a bank, or an auditor can rely on. That distinction is not just honesty for its own sake: the EU's Empowering Consumers for the Green Transition Directive applies from September 2026 and specifically targets vague, unverifiable environmental claims. "We're a sustainable partner" is exactly the kind of statement it's built to catch. A shipment-level record with its data quality tier and calculation basis stated is not, because it's a claim that can be checked.
What "methodology-consistent" means in operational practice
None of this is about hiring a transport-emissions specialist or applying stricter standards. It's about applying the same standard correctly at every step, which is hard to do by hand when it's one person's side project:
- Primary activity data used where available, with clear documentation of where defaults are applied and why
- Current emission factor versions with version number and date documented in the calculation output
- Mode-specific allocation applied correctly for each leg in a multi-modal chain
- System boundary defined, documented, and consistent across reporting periods, with pre-carriage, last-mile, and WTT all included
These are the errors that surface in Scope 3 Category 4 reviews, whether the review is a formal CSRD assurance engagement or simply a large customer checking the number a smaller supplier sent them. Every gap above has the same shape: a rule half that's a matter of discipline, applying ISO 14083 consistently across every shipment, every week, and a data half that only the client, or the carriers and forwarders they work with, can supply. For a forwarder or carrier, getting the rule half right is also a competitive point: a consistent, defensible number is what lets a shipper trust a quote without re-deriving it themselves.
CO2Path's job is the rule half. It enforces ISO 14083 logic at every calculation step, mode-specific distance and allocation applied automatically, primary data preferred over a default the moment it's supplied, emission factors kept current, WTT included alongside TTW, so nobody has to catch these by hand. What it won't do is guess at data nobody gave it: a missing route, payload, manifest, or transport leg shows up as a flagged gap, not a silent assumption. Closing every gap for good stays a joint effort, CO2Path holds the rules steady, the shipper and their carriers supply the rest.
See where your own freight emissions data still has gaps → Request a demo
Frequently asked questions
The four most common gaps are: (1) using a mapping tool's default route instead of the mode-specific or actual-corridor distance ISO 14083 requires, most often applying road-style driving distance to a rail or sea leg, or missing real routing choices such as EU-only border-crossing corridors; (2) using outdated default load factors or emission factor databases; (3) incorrect mass-based allocation for co-loaded shipments; and (4) incomplete system boundary, excluding WTT (fuel production) and reporting only TTW (tailpipe combustion).
A mapping tool's default route is built for road driving, so it's the wrong basis for any leg that isn't road, and even for road-only lanes it may not reflect the corridor actually driven. Piraeus to Rotterdam, for example: the shortest route crosses Serbia and North Macedonia, both outside the EU, but many carriers route via Bulgaria, Romania, Hungary, and Austria instead, a longer distance chosen specifically to avoid non-EU customs stops, which a generic query has no way to know. The GLEC Framework's own preferred road-distance basis, the Shortest Feasible Distance (SFD) found via route-planning software, explicitly recognises this kind of deviation: actual routes commonly differ from the planner's default to avoid tolls, reach rest points, or, as here, avoid customs delays. A narrower, related mistake shows up in high-volume automated calculators that skip routing entirely and estimate distance directly from coordinates: a 2026 study of European road networks puts the typical gap between that straight-line shortcut and actual routed distance at a median of 1.25 (roughly a quarter longer). Uncorrected, any of these versions of the gap compounds across thousands of shipments per year into a materially incorrect annual total, while appearing specific.
Allocation splits a vehicle's total emissions between co-loaded customers. ISO 14083 requires mass-based allocation (cargo weight multiplied by distance). Using incorrect allocation bases, such as volume without distance weighting or flat per-consignment splits, redistributes emissions incorrectly between customers, which affects both internal reporting accuracy and external CSRD disclosures. Applying the correct rule is calculation logic; having the correct inputs depends on the carrier sharing the full manifest of what else was on the vehicle, not just the one customer's own shipment.
Well-to-Tank (WTT) covers the carbon cost of producing and distributing fuel, before it reaches the vehicle. ISO 14083 requires that both TTW (tailpipe combustion) and WTT are included in every calculation (WTW boundary). Official UK government conversion factors put diesel HGV WTT at roughly a quarter of TTW (0.212 vs. 0.891 kgCO₂e/km), so reporting TTW only leaves out a meaningful share of the true total. This is one of the most common scope boundary errors in practice.
They are invisible in the output. A calculation that applies ISO 14083 allocation incorrectly to every shipment produces precise-looking totals that are systematically biased. The error is only visible when the methodology is examined specifically, which is exactly what CSRD assurance teams, and increasingly large customers checking their own suppliers, do.
No. Most shippers and forwarders fall outside CSRD's scope after the Omnibus revision, which limited mandatory reporting to companies over 1,000 employees and €450 million in turnover. But a large, CSRD-obligated customer, or a bank, can still ask a smaller supplier for shipment-level freight emissions data, and the EU's value chain cap only limits what they can require, not what they can request. A clean, ISO 14083-consistent number is a practical requirement for working smoothly with bigger customers, whether or not the supplying company has a legal mandate of its own.
Yes, at the default level; no, at the primary-data level. CO2Path can produce a fully ISO 14083/GLEC-compliant calculation on its own using GLEC default values, that's a standard, legitimate part of the methodology, not a fallback to apologise for. What it can't do is upgrade a default to primary data on its own: a truck's actual payload, the real corridor driven, the full manifest of a shared vehicle, or a missing pre-carriage or last-mile leg all depend on the client or their carriers actually supplying that better input. What a platform can do is apply primary data automatically the moment it's supplied, and keep it clear which figures are still defaults and which are primary, rather than blurring the two.