Calculating Carbon Costs Isn’t a Sustainability Project - It’s a Data Problem
This blog post was written by Robin Spickers, Co-Founder and Managing Director of carbmee, for JAGGAER. It originally appeared on JAGGAER.com.
When I talk to procurement teams in manufacturing industry, I often hear the same statement: “We’re currently working on our CBAM strategy.” Fine. But when I then ask what data this strategy is based on, there’s usually silence.
The real problem isn’t understanding CBAM or ETS. The real problem is that most companies simply don’t know what they’re buying at the level of detail needed to calculate carbon costs reliably.
This isn’t a criticism. It’s a structural reality in the industrial supply chain. And it’s the reason carbmee exists.
Why carbon is a major data problem
Imagine a typical manufacturing company: 500,000 to 700,000 unique part numbers in direct procurement. 10,000 to 30,000 suppliers. Hundreds of locations worldwide. Thousands of end-product configurations.
For each of these components, you need to know: What material is it made of? How much of it is there? Where is the supplier located? What emission factor applies to this material in this region? And how does this play out given the current ETS price?
No one can manage that in Excel. And no one should even try.
The problem isn’t a lack of interest—it’s a lack of granular data. Most companies have procurement data in their ERP systems, bills of materials in their PLM systems, and supplier information somewhere in between. But this data is rarely linked in a way that allows for a CO₂ calculation at the part-number level.
This connection is precisely our core mission.
How AI can solve this problem
The first step is always the same: we connect to your core systems—JAGGAER, SAP, PLM, Catena X, Snowflake, or whatever else you have. No months-long implementation. Just one day.
What follows is the real work, and this is where AI comes into play.
A practical example: many companies are unable to link the weights in their bills of materials to their procurement data. Without weights, there is no basis for activity-based carbon calculations. We therefore developed an AI workflow that estimates weights from natural-language product descriptions. At first glance, this may sound straightforward, but it makes it possible to perform activity-based calculations from the outset instead of relying on coarse spend-based estimates.
The result: a data model is created for every part in your supply chain. With an emission factor. With a CO₂ cost estimate. And decisions can be made based on this—immediately, without waiting for perfect data.
From data point to decision
Data alone isn’t enough. What matters is what you do with it.
That’s why we didn’t build carbmee merely as a reporting tool, but rather as a decision-making system. Specifically, this means:
Sourcing decisions with a carbon dimension. For every request for proposal (RFP) or request for quote (RFQ), the procurement team can see right away: what is Supplier A’s carbon footprint compared to Supplier B’s? Carbon thus becomes a genuine evaluation criterion, not just a required field on an ESG questionnaire.
CO₂ in the cost breakdown. Anyone who takes total cost of ownership (TCO) seriously must factor in CO₂ costs. We feed the data directly into the JAGGAER cost breakdown. In this way, CO₂ is included in the calculation and isn’t limited to sustainability reporting alone.
Forecasting, not rear-view mirror. We can create a projection based on your actual purchasing data: what will your current supplier choices cost you under the 2027 ETS prices? Not as a scenario, but as a specific figure per category, per supplier, and per end product.
What Tier-n visibility has to do with CBAM
Many companies know their Tier 1 suppliers well. Tier 2 is more of a challenge. Tier 3 is a blind spot for most.
Until now, this has been a theoretical problem. With CBAM, it will become a regulatory one.
That’s because CBAM doesn’t require you to know your direct suppliers; it requires you to know the carbon footprint of the imported product. But that footprint is often generated several steps earlier in the value chain: at the steel producer, the aluminum foundry, or your supplier’s supplier.
Catena X and similar standards are a step in the right direction. But as long as this data isn’t widely available, we need AI-powered estimation models that derive reliable emissions values from existing data, such as material composition, country of origin, and product category.
This isn’t a compromise. It’s a pragmatic way to take action today.
