The client delivers environmental, social, and governance ratings, risk assessments, and benchmarking intelligence to institutional investors, asset managers, and other key stakeholders through its end-to-end SaaS tool. Their proprietary model evaluates corporate ESG performance across all three pillars for publicly listed companies worldwide.
Our partnership with this client began in 2019 — a "first-of-its-kind" project for SunTec India that marked our strategic entry into the ESG research and data services sector. At the time, the client’s internal team managed the rating platform and handled approximately 500 companies. We were brought on to provide ESG research and data management support. We assembled specialized resources in the sustainability domain, created processes, and delivered organized datasets that the client approved and appreciated. This success fueled rapid scaling: our initial team of 5 members grew into a dedicated 35-person team by the third year of our association. By early 2023, our collaboration had helped the client grow its coverage from 500 to approximately 1,500 companies—tripling its original footprint and establishing a proven research infrastructure across all three ESG pillars.
Due to growing investor demand for broader ESG coverage, competitive pressure from rival rating agencies expanding their coverage, and regulatory developments mandating comprehensive ESG disclosures, the client decided in late 2023 to undertake a significantly more ambitious expansion — scaling from 1,500 companies to 8,000+, including Fortune 500 firms, mid-cap corporations, and regional players across 18+ industries. To achieve this, they needed assistance with environmental, social, and governance data collection, as well as cross-pillar verification, metric calculation, and multi-framework mapping. The client also required comprehensive source documentation for every data point to maintain the defensibility of their published ratings.
The client required comprehensive data collection, with datasets formatted for direct integration (Excel, CSV, JSON) into their proprietary ESG rating platform. Additionally, our research processes, from metric definitions and scope boundaries to calculation methodologies and quality thresholds, had to conform to the client's proprietary scoring taxonomy and data schema.
The scope of the ESG data research project included:
Data research, collection, cleansing, and organization across 8000+ companies, with 95+ metrics per company — spanning environmental, social, and governance data research services with distinct data characteristics, disclosure norms, and framework requirements — presented compounding challenges that went beyond the sum of their parts. Additionally, the contextual, cross-referential, and judgment-intensive nature of this project meant that automation could accelerate the research, but human verification was non-negotiable.
ESG data for a single company often tells conflicting stories across pillars. In one case, a manufacturing firm's governance report emphasized "robust environmental oversight" and referenced a board-level sustainability committee. In contrast, its environmental disclosures revealed no Scope 3 reporting and rising absolute emissions. Similarly, a technology company's social report highlighted a "zero-tolerance" human rights policy, but its governance filings showed no board-level human rights oversight and no supply chain auditing mechanisms. Detecting these cross-pillar contradictions required analysts with working knowledge of all three ESG domains.
ESG data is scattered across sustainability reports, CDP responses, proxy statements, regulatory filings, annual reports, corporate ethics codes, and company websites. Environmental metrics use different units (metric tons vs. short tons, MWh vs. GJ). Social metrics lack universal definitions — one company reported "employee turnover" as voluntary exits only, while a peer reported it as voluntary exits plus involuntary separations. Governance disclosures vary by jurisdiction: "board independence" has distinct legal definitions under the UK Corporate Governance Code, NYSE/NASDAQ listing standards, and EU directives. These inconsistencies affected nearly 30% of the initially collected data points, requiring manual reconciliation at scale.
A single data point may require simultaneous tagging across multiple frameworks with different taxonomies. For example, an emissions figure needed mapping to the applicable ISSB (IFRS S2) metrics and targets disclosure, the applicable GRI 305 scope-specific disclosure, the relevant CDP climate change questionnaire module, ESRS E1-6, and the applicable SASB standard. Mapping more than 95 metrics across 8000+ corporations to this framework spread demanded granular familiarity with each standard and meticulous .
Companies report metrics using different units, base years, methodologies, and jurisdictional standards. Data standardization and normalization for 800,000+ data points required consistent application of conversion standards, expert judgment on scope boundaries, and documented methodologies for every derived metric. The need for calculation-based derivation ranged from approximately 10% for large-cap firms to nearly 30% for mid-cap and regional companies with limited disclosure practices.
Meeting large-scale ESG research requirements across all three pillars required a purpose-built workflow. Our solution paired AI-assisted research for speed with subject-matter expertise for accuracy — and adapted continuously as we identified operational roadblocks.
We restructured our existing 35-person team into pillar-specific sub-teams, with a larger environmental contingent reflecting the metric volume, and smaller social and governance units — supported by a dedicated QA Lead with ESG data validation experience. Over a 6-week onboarding period, we trained the team on the client's proprietary scoring taxonomy, the 95+ metric framework, pillar-specific disclosure requirements (CSRD/ESRS, including ESRS 2 cross-cutting governance disclosures; GRI; SASB; ISSB; CDP; UNGP; SFDR), and cross-pillar verification protocols.
Operational Roadblock: During the training period, our inaccuracy rate was 27% — due to inconsistent framework mapping across pillars, missed unit conversions in environmental data, and misaligned social-metric definitions. We addressed this with a comprehensive internal reference guide covering all three pillars, daily calibration sessions led by the QA Lead, and a peer-review pairing structure where analysts cross-checked each other's work before submission. By the end of the training period, the pilot batch met the client's accuracy threshold across all validated data points. Inconsistent data — where companies reported conflicting figures across documents — were flagged and routed to the client for final review.
Our environmental sub-team gathered data from sustainability reports, CDP climate and water responses, annual reports, third-party databases, and more — covering Scope 1, 2, and 3 emissions, energy consumption, water withdrawal, waste generation, air pollutants (NOx, SOx, VOCs, PM), and corporate environmental policies.
Our social sub-team sourced data from CSR reports, publicly disclosed HR policies, codes of conduct, and open-access databases — capturing workforce diversity, employee health and safety, labor rights, human rights commitments, and supply chain ethics.
Our governance sub-team extracted data from proxy statements, annual reports, board charters, remuneration reports, and regulatory filings — covering board composition, executive compensation, shareholder rights, risk and compliance, and governance-linked social policies.
We used AI research tools (Claude, Perplexity, and ChatGPT) for faster document identification, preliminary data extraction, and locating references across corporate disclosures. Every data point was verified against the original source before entry into the final dataset.
Each data point was cross-validated within its pillar. Beyond this, our analysts performed cross-pillar consistency checks—comparing governance commitments with environmental and social disclosures to identify contradictions that single-pillar analysis would miss.
Example: A European energy company publicly declared climate change a "board-level strategic priority" with a published net-zero commitment. Cross-pillar verification contradicted this directly. Environmental data showed that 80% of its total emissions (Scope 3 emissions) were excluded from the net-zero target. 91% of its capital expenditure went toward fossil fuel projects, contradicting the transition to clean energy. The company increased its workforce in fossil fuel divisions by 34%, but there was little to no growth in its renewable energy sectors. Only the company's governance bodies claimed climate leadership.
We documented these cross-pillar findings with comprehensive source evidence, revealing a clear disconnect between the company's public commitments and its actual business practices.
To maintain accuracy beyond the training phase, we formalized a structured quality control workflow. Conflicting data — where a company reported different figures across its sustainability report, CDP response, and annual filings — was flagged, documented with all source variants, and escalated to the client with a recommended resolution. The peer-review pairing structure introduced during onboarding became a permanent QC step: every analyst's output was cross-checked by a second analyst before submission, and the QA Lead conducted randomized spot checks on 10–15% of completed datasets per cycle. Anomalies and edge cases — such as dual-class share structures, partial Scope 3 reporting, or jurisdictions with limited social disclosure mandates — were escalated with documented context.
Final datasets were delivered in client-specified formats (Excel, CSV, JSON) with source citations, calculation notes, and audit trails for direct integration into the client's proprietary ESG rating platform.
Where companies did not disclose required indicators, our team derived them using documented methodologies, such as emissions intensity and energy efficiency ratios (environmental), total injury rate and gender diversity ratio at the management level (social), and CEO-to-median pay ratio, board independence percentage, and variable-to-fixed compensation split (governance). This capability was particularly critical for mid-cap and regional companies, where up to 30% of required metrics were not directly disclosed.
We developed a tiered tagging system — "Complete," "Partial," and "Not Disclosed" — enabling the client's platform to appropriately weight data quality across their scoring models. All calculated values were flagged to distinguish them from directly reported figures, and complete methodology documentation was provided for every derived metric.
Our team tagged each data point to applicable frameworks across the regulatory and voluntary disclosure landscape — CSRD/ESRS (E1–E5, S1–S4, G1, and ESRS 2 GOV-1 through GOV-5), GRI, SASB, ISSB (IFRS S1/S2), CDP, UN Global Compact, UNGP, and SFDR— with EU Taxonomy alignment assessments where applicable. Tagging structures were also designed to accommodate evolving U.S. state-level climate disclosure requirements, including California's SB 253, as the regulatory landscape continues to develop. Where boundary definitions differed across frameworks, we maintained independent scope documentation. This allowed the same data point to have different scope parameters depending on the standard, without forcing one interpretation. This was especially important for metrics such as emissions boundaries, where different approaches (e.g., GHG Protocol, ESRS, ISSB) can yield significantly different figures from the same data.
We delivered high-quality, compliant ESG data that consistently met accuracy benchmarks and client timelines. Building on a dedicated team of 35 resources established for the prior ESG data management collaboration, we restructured and expanded the team to 95 members — integrating pillar-specific sub-teams, a dedicated QA Lead, and AI-assisted research workflows to absorb the coverage expansion without proportional cost scaling. As of now, the client continues to rely on our team for ongoing ESG data research and data collection support.
500 to 8000+ Companies over 7 Years ESG ratings coverage grew 16x over the course of the seven-year partnership.
800K+ Data Points Collected Total 95+ ESG metrics per company across environmental, social, and governance pillars.
Source-Documented ESG Data Delivery 100% datasets delivered with cited sources, calculation notes, and complete audit trails.
From environmental research services covering emissions, energy, and waste, to social research services for workforce diversity, labor rights, and health and safety, and governance research services for board composition, executive compensation, and policy compliance — SunTec India delivers audit-ready datasets for ESG rating agencies, investment consulting firms, and finance/tech consultants worldwide.
Our ESG research services are scalable and can be tailored to your framework alignment, coverage scope, and delivery timelines. Reach out to our team for a consultation or request a pilot engagement.