Supplier ESG Monitoring at Scale, Without the Self-Reporting

The company's supplier sustainability program ran on questionnaires that reached only a fraction of a ~69,000-supplier base. Veridion reproduced its four ESG frameworks from public evidence in the company knowledge graph, turning a periodic, self-reported survey into a continuous, auditable monitoring baseline across the whole base.

1. 1
   Replaced supplier self-reporting

   questionnaires reached only a fraction of the base, slow and unverifiable

2. 2
   Matched the full base

   ~69,000 suppliers resolved in Veridion's company knowledge graph

3. 3
   Scored from public evidence

   disclosures, filings, news and certifications, with no supplier input

4. 4
   Rebuilt four ESG frameworks

   ~100–150 datapoints per supplier, every flag sourced

5. 5
   Held a 0.95 precision floor

   on deterministic data (≥0.80 AI-derived), every datapoint traceable

6. 6
   Monitors continuously

   a periodic survey becomes an always-on, auditable baseline across the whole base

A survey can't reach 69,000 suppliers

The company runs a supplier sustainability program, assessing suppliers against four frameworks: its Supply Chain Standards, a Social Responsibility Scorecard, a set of Carbon Data Points, and a consolidated Supplier Assessment Questionnaire, all collected by sending each supplier the forms to complete.

Self-reporting has a low ceiling at this scale. A returned questionnaire records what a supplier says about itself rather than what it does; the answers are slow to collect, inconsistent between suppliers, and hard to verify. And a survey you have to send, chase, and check can only ever reach a fraction of 69,000 suppliers at any useful frequency.

The company could assess a sample periodically. It could not monitor the base. The brief was to reproduce the frameworks' coverage from public evidence, so assessment no longer waited on a returned form and could run as continuous, low-touch monitoring across the whole base.

Four frameworks, rebuilt from public evidence

Veridion rebuilt the program on its company knowledge graph, the continuously refreshed record of what companies publicly disclose and do. Every supplier on the list resolved to its entity in the graph; from there, Veridion rebuilt each of the four frameworks as a custom scoring model and returned roughly 100 to 150 datapoints per supplier, scored from public sources rather than a supplier's response.

Scoring mirrors the company's own logic: a four-step compliance flag (compliant, mixed, no-evidence, non-compliant) for the Social Responsibility, Supply Chain and questionnaire frameworks, and a three-tier carbon model (Least, Preferred, Most Preferred) across setting targets, sharing progress, and reducing emissions.

Every datapoint carries its evidence: a source URL, detection path, timestamp and confidence value, with a false-positive budget holding precision at a minimum of 0.95 on deterministic data and 0.80 on AI-derived, all from public web content only.

A worked example shows the nuance the method keeps: a commercial-landscaping supplier that pledged carbon neutrality by 2035, with a fleet electrification programme but no formal science-based target, scored Least Preferred on targets yet Most Preferred on both progress and emissions, each flag traced to the supplier's own disclosures.

The pilot scored 500 suppliers across every datapoint, took 25 of them (plus 9 priority suppliers) end-to-end against the full questionnaire for a like-for-like comparison with the self-reported answers, and quantified reach across the ~69,000-supplier base in a coverage report.

A periodic survey becomes continuous monitoring

The company can monitor its entire supplier base on the public record, continuously, rather than surveying a slice of it periodically.

Every answer traces to a source and can sit beside the supplier's self-disclosed answer to show exactly where the two diverge.

Because the underlying graph refreshes continuously, a periodic self-reported snapshot becomes an always-on monitoring baseline, one that flags emerging risk early and points to where a deeper assessment is actually warranted, supporting the existing evaluation process rather than replacing it.