A North American mid-market forwarder won six RFPs in a row after losing three for missing a carbon line

What they had before

The customer is a North American mid-market freight forwarder running approximately $140 million of annual revenue across roughly 800 shipments per week. Their flow is mostly trans-Pacific ocean — Shanghai, Ningbo, and Busan into Los Angeles and Long Beach — plus a significant trans-border road book between US Midwest manufacturing and Mexican maquiladoras. They run a homegrown quote engine on top of CargoWise for the operational side.

The trigger was an RFP cycle. In 2024 they entered four RFPs from large CSRD-bound EU shippers and CDP-respondent US importers. All four required per-shipment carbon on quote responses. They won one — the relationship was strong enough to absorb a "we will follow up with carbon numbers in the next cycle" answer. They lost the other three, and in each case the debrief from the procurement counterpart specifically called out the missing carbon line as a tiebreaker against a forwarder that did provide it.

The internal conversation through Q4 2024 was about whether to build versus buy. The customer's engineering lead initially scoped a build — a factor library plus a distance-lookup service plus a GLEC mapping layer — at roughly a quarter of engineering work. The sustainability lead pointed out that the resulting numbers would still need third-party methodology defence in front of CSRD auditors. That was the moment they reached out to us.

What changed

We integrated the EcoFreight API into the customer's quote engine at the point where the quote response is assembled. The shape is the simplest possible: when a spot quote completes the carrier-and-route step, a server-side service calls POST /api/v1/calculate with the GLEC v3.2 factor set, the WTW scope, and the load-factor assumption the customer's lane analytics team agreed on per corridor. The returned co2e_kg and data_quality_tier are attached to the quote response template inside the existing 24-hour SLA.

On the trans-Pacific lanes we configured the request to use the container-ship factor segmented by TEU class, because the customer's carrier panel runs everything from Post-Panamax to ULCV. On the trans-border road book we use the GLEC v3.2 articulated truck above 32-tonne factor with a 72% load factor from the customer's lane analytics, lower than the European 3PL case because cross-border return-leg utilisation in this corridor is genuinely worse.

The integration landed alongside a wider quote-engine rebuild the customer had already scoped for 2025. That co-incidence is important context — and a complication for ROI attribution we will get to. The wider rebuild also introduced a real-time rate sheet for ocean and a much better carrier-of-choice selector. Either of those by itself would plausibly have lifted quote conversion. The carbon line did not land alone.

The customer's procurement-facing template was already mature enough to accept new fields without much template work. The CO₂ line appears underneath the freight rate line on every quote PDF, with the methodology footer the customer's sustainability lead drafted with us: "WTW CO₂e per shipment, GLEC Framework v3.2, ISO 14083 data quality tier as shown." Two RFP procurement teams flagged that footer in their debriefs as the reason they accepted the number without follow-up questions.

Where the numbers landed

The 12% conversion lift comparison is intra-account, not cross-account. We compared quotes inside the same shippers that received a mix of carbon-line and non-carbon-line quotes during the rollout. That removes account selection bias. It does not remove the broader quote-engine rebuild effect, which is the next caveat.

What we would do differently

Two things, with the benefit of hindsight.

We would have insisted on the A/B holdout. The carbon line landing inside a bigger quote-engine rebuild made the post-hoc attribution argument harder than it had to be. Running half the quotes with the carbon line and half without for a defined window would have settled the question and made the case study cleaner. The customer would probably have agreed at the time; we did not push hard enough.

And we would have flagged load-factor onboarding earlier. The 72% load-factor figure on trans-border road was negotiated between us and the customer's lane analytics team in week two of the integration, after the quote engine was already wired to call our API. That negotiation could have happened in week zero and saved two days of regenerating earlier quote tests. Same lesson as the European 3PL: the load-factor source goes into the first integration sprint.