For a long time, sustainability was scored on intention: a certificate, an annual report, a statement in the tender about CO2 reduction. The underlying data — energy consumption per process, origin of materials, transport kilometers, waste streams — was collected in fits and starts, usually once a year, often with months of delay. Whoever had the best story first won the comparison, even if the story was just a little ahead of reality.
That asymmetry between claim and proof shifts as soon as the collecting and tracing of data is itself largely done by AI. Data that comes from systems — energy meters, logistics planning, purchase orders, production schedules — can be continuously aggregated and measured against standards without anyone having to compile a quarterly report. What remains as human work is interpretation: is this reduction the result of a structural change or a favorable month, and which exceptions in the figures call for an explanation that no one provides automatically.
Behind a sustainability claim lie three kinds of work, and they do not shift at the same pace.
Collecting and consolidating figures from measurement systems, ERP and supplier data is largely something AI can take over: it is repetitive, structured and verifiable against a standard. Companies that had already linked their systems see this part shift the fastest.
Weighing conflicting signals — a supplier that meets the standard on paper but deviates in practice, a reduction figure that does not match the rise in production — is work requiring human oversight: AI flags the deviation, a person judges whether it is relevant and why. That oversight is not decorative; it determines whether a report holds up the moment an auditor or client asks further questions.
Formulating a sustainability strategy, choosing which chain to make more sustainable and at what pace, and weighing costs against reputation remains human work. That is a governance choice, not a calculation.
The customer does not notice the difference through a higher sustainability figure, but through the speed and granularity of the answer. A provider who has their data continuously consolidated can immediately give a figure on a specific question — the CO2 footprint of this one delivery, not the whole year — with a trace to the source. A provider who still adds this up manually once a year gives an estimate or refers to the latest report.
In a tender process, that difference weighs more heavily than the average sustainability figure itself. A buyer who must demonstrate that their own chain complies with regulations would rather have a supplier who can prove per shipment what was emitted than a supplier with a nicer annual average and no underlying data.
The shift goes faster at companies where the source systems are already structured: connected energy meters, digitized supplier chains, unambiguous product codes. At companies where that data still sits in spreadsheets and separate forms, AI cannot consolidate what has not been supplied in structured form — there, the work remains largely manual, with the same delay as before.
This means that two competitors with comparable sustainability ambitions can nonetheless show a very different level of proof, not because one is less sustainable, but because one's data foundation is suitable for automated consolidation and the other's is not. For a buyer comparing on this dimension, that distinction is often more important than the reported figure itself.
Sustainability rarely stands on its own in a tender. It is weighed alongside certification, alongside the question of whether a supplier takes over guarantees and risks via warranty and risk transfer, and alongside practical matters such as ease of doing business. A provider who comes with proof on sustainability but with a claim on those other dimensions does not automatically win the comparison — the score only counts once each dimension separately holds up.
Whether that oversight of figures and the interpretation that follows already works this way in your own organization, or still happens largely manually, is a question about which work in this company can genuinely be taken over by AI — the work scan from FTE TO AI maps that out per task. Anyone who wants to link that to a concrete comparison with competitors will find the approach for that in how you score a competitor without assumptions and in how you benchmark your company against competitors.
The question of whether AI plays a role in personnel decisions falls outside what is addressed here: separate legal requirements apply to that, apart from this benchmark.
What can be answered here: where you believe you win on sustainability, and whether that claim holds up against a competitor. The free dimension check lets you name where you believe you win and shows which of those claims can be defended with proof. The full benchmark, with the comparison across all dimensions at once, is under construction.
Vraag maar waarop er in uw markt gewonnen wordt. Ik vergelijk liever dan dat ik uitleg.
Answers come from this site’s knowledge base. Not tailored advice, and not a scan of your company.