How Walk-The-Talk Score Is Calculated
How Walk-The-Talk Score™ is calculated. Clear principles you can understand. Proprietary scoring you can trust.
Last reviewed: March 2026
Core Principle
Walk-The-Talk Score measures credibility, not disclosure quality. A company that makes modest commitments and delivers on them will score higher than a company that makes ambitious promises without evidence of follow-through. We reward execution over aspiration.
How It Works
Four steps from raw data to Walk-The-Talk Score.
Commitment Extraction
We extract ESG commitments from public company disclosures.
- •Annual reports, sustainability reports, and integrated reports
- •Press releases and official company announcements
- •SEC filings (10-K, DEF 14A) and equivalent international filings
- •Official company website statements and policies
Natural language processing identifies specific commitments with measurable targets and timelines. Generic statements without specificity are flagged but weighted lower.
Evidence Collection
We gather third-party evidence to verify each commitment.
- •Regulatory filings and government databases
- •NGO reports and watchdog publications
- •Litigation records and court filings
- •Academic research and industry studies
- •Investigative journalism and verified news sources
Evidence is matched to commitments using entity resolution and topic modeling. Each piece of evidence is scored for relevance and credibility.
Controversy Monitoring
We monitor real-time signals that contradict stated commitments.
- •Global news feeds (200+ countries, 65+ languages)
- •Regulatory enforcement actions
- •Lawsuits and legal proceedings
- •NGO campaigns and whistleblower reports
- •Social media signals (verified sources only)
Controversies are severity-scored based on source credibility, financial impact, and relevance to stated commitments. Proprietary algorithms weight recency and resolution status.
Score Calculation
We combine the three pillars into a single credibility score.
- •Commitments and Evidence are weighted roughly equally
- •Controversies act as a credibility modifier
- •Proprietary algorithms combine inputs into final score
The final score is 0-100. The specific weighting, reconciliation logic, and scoring thresholds are proprietary. Scores are benchmarked against industry peers.
Update Frequency
Walk-The-Talk Scores are not static. Different components update at different frequencies.
| Component | Frequency | Trigger |
|---|---|---|
| Commitment refresh | Regular | After major disclosure events |
| Evidence validation | Periodic | Scheduled revalidation cycles |
| Controversy monitoring | Continuous | Event-driven detection |
| Score recalculation | Ongoing | As new data is incorporated |
Limitations & Scope
No methodology is perfect. Here's what you should know about ours. See the ESG glossary for definitions of key terms.
Coverage gaps
Companies with minimal public disclosures may have incomplete scores. We explicitly flag data quality issues.
Language limitations
While we process 65+ languages, some regional sources may have lower coverage than English-language sources.
Time lag
Evidence collection has inherent delays. We aim for 24-48 hour latency on major controversies, but some sources update less frequently.
Subjectivity in severity
Controversy severity scoring involves judgment calls. We document our reasoning and allow users to adjust weights.
Want More Detail?
Explore our complete data source inventory or see Walk-The-Talk Score in action.