In today's world, practically every strategic decision is made on the basis of data. Financial models, credit risk, investment analysis, stress testing: all of them rely on one simple assumption: the underlying information is reliable enough to support conclusions.

ESG is no exception. Moreover, it is precisely here that the problem of data quality becomes particularly acute.

Unlike most traditional financial tasks, where an analyst works with a relatively standardized set of indicators, ESG requires the analysis of a huge amount of heterogeneous information. Even for making a relatively small decision, it is necessary to collect data from corporate reports, research, regulatory publications, industry initiatives, news, academic papers, and information from suppliers.

Naturally, not all of these sources are publicly available. Many high quality databases are commercial, while some information cannot be obtained without direct interaction with the company. Therefore, a significant part of the market is forced to work with what is freely available.

And this is precisely where the most serious problem begins.

A significant portion of publicly available ESG information does not undergo independent review or audit and is not accompanied by independent verification. It finds its way into analytical reviews, research, presentations, and companies' internal models. These materials then themselves become new sources of information, which other analysts and organizations begin to reference. 

For example, in a 2024 Deloitte study, 57% of executives surveyed identified ESG data quality as the main challenge for their organization, while 88% included data quality among their top three challenges. At the same time, 81% identified data documentation and the data approval process as one of the key challenges.

This creates a closed cycle.

Incorrect information can gradually turn into accepted knowledge simply because it appears across numerous sources that seem independent at first glance.

The situation is further complicated by the rapid development of generative artificial intelligence. Modern language models are trained on the existing information ecosystem and reproduce knowledge that has already been published. At the same time, the share of content created or substantially reworked using AI continues to grow rapidly.

As a result, we are gradually finding ourselves inside a closed information ecosystem, where new materials are created on the basis of older publications that, in turn, may already have been based on inaccurate or unverified data. Each new iteration can gradually move the final result further away from the original source of the data.

If in most areas such information noise is more of an inconvenience, in ESG the consequences are significantly more serious. ESG data becomes one of the factors in investment decisions, climate risk assessments, the development of decarbonization strategies, and the allocation of financing.

The problem is not only the amount of data, but also its comparability. A study by Berg, Kölbel, and Rigobon, published in Review of Finance, found significant divergence between the ESG ratings of six major rating agencies. The authors attribute these differences primarily to variations in the measurement and scope of ESG factors: differences in measurement accounted for 56% of the identified divergence, differences in scope for 38%, and differences in the weighting of indicators for 6%.

This is precisely why bad ESG data is often more dangerous than a complete lack of data. The absence of information forces us to acknowledge uncertainty and seek additional confirmation. Poor data, on the other hand, creates an illusion of knowledge. And false confidence is almost always more dangerous than honestly acknowledging uncertainty.
The problem is that in a corporate environment, this false confidence rarely remains simply an error in a database. ESG data is used further: it becomes an input parameter for models, a basis for reporting, and one of the factors in management decision making. An error that may initially have seemed insignificant can therefore have very real consequences.

It is at this point that the problem of data quality ceases to be exclusively a technical problem for the ESG team and becomes a business risk.

When Bad Data Becomes a Business Risk

When such information enters decision making processes, the consequences become very concrete.

Many companies still operate on the principle that "providing some data is better than providing none at all." In practice, this is far from always being the case.

When ESG data is incomplete, inconsistent, or unreliable, it stops being simply information and becomes a source of real business risks.

First and foremost, the quality of decision making suffers. Management uses ESG metrics in investment planning, risk management, supplier selection, and the development of sustainability strategies. Any analytical model is merely a function of its input data. If the data itself contains a systematic error, no level of model complexity can correct the result. The Garbage In, Garbage Out principle remains unchanged regardless of whether the tools used are spreadsheets, sophisticated statistical models, or modern AI tools.

At the same time, high quality ESG data can create competitive advantages. It can help identify opportunities to improve energy efficiency, optimize supply chains, reduce operational risks, and obtain more favorable financing conditions.

Trust is no less important. In an environment of increased attention to greenwashing, even unintentional errors can cast doubt on the credibility of a company's entire ESG reporting. Investors, banks, and regulators are increasingly assessing not only the metrics themselves, but also the maturity of the underlying data management processes.

Finally, regulatory pressure is also increasing. Requirements such as CSRD and ESRS in the European Union are gradually making ESG data comparable in importance to financial reporting. Companies must be prepared not only to provide metrics, but also to explain the origin of the data, the calculation methodology, and the possibility of independent verification.

This is already reflected in the architecture of the regulatory framework itself. Under CSRD, companies subject to sustainability reporting requirements must obtain an assurance conclusion on their sustainability statement. At the initial stage, this involves limited assurance, including an assessment of compliance with the ESRS requirements and the process used by the company to determine the information to be disclosed.

Where ESG Data Problems Come From

In most cases, the problem is not the lack of information, but the lack of a unified data management system.

Different departments use different definitions for the same indicators. Some information is still collected manually in spreadsheets. Many metrics have no designated owner, while the processes for updating them remain insufficiently formalized.

In a 2024 EY study, 96% of financial executives surveyed reported experiencing some form of problem with non-financial data. The issues identified included data quality, completeness, consistency, and timeliness. EY also notes that the complexity of ESG topics and the large number of required data sources make it difficult to produce reporting that is comparable in reliability to financial reporting.

As ESG requirements grow, this fragmentation begins to directly affect the quality of corporate reporting.

This is particularly visible when calculating Scope 3 emissions.

For many organizations, Scope 3 represents the largest share of their carbon footprint. However, this data is rarely measured directly. Companies receive information from suppliers, suppliers receive information from their suppliers, and when actual data is unavailable, industry factors, proxy indicators, and expert estimates are used.

The GHG Protocol explicitly provides for the use of different approaches to calculating Scope 3 emissions depending on data availability, including supplier data and other estimation methods. The standard covers 15 categories of emissions across the entire value chain, including both upstream and downstream activities. The GHG Protocol also emphasizes the importance of consistency in methodologies and data used to ensure comparability of results.

Practical evidence from the implementation of ESRS demonstrates how challenging this process remains. In EFRAG's 2024 observations, around 80% of companies participating in the study identified data availability as one of the main challenges. The value chain area was identified as the least developed, while working with data beyond direct business partners remained a significant challenge. 

Each additional link in this chain increases uncertainty. A single initial error can spread throughout the entire supply chain, gradually becoming an "accepted fact" that is then used in new models, reports, and research.

This is precisely why ESG data quality cannot be ensured exclusively at the reporting stage. It must be established much earlier, during the collection, verification, and management of information.

In a 2024 Deloitte study, 92% of executives surveyed indicated a need for additional investment in technology to ensure consistent and reliable measurement, reporting, and disclosure of ESG information. At the same time, 89% reported that they were likely to strengthen their ESG control environment, while three quarters of respondents planned to obtain assurance over ESG disclosures in the following reporting cycle.

How to Build a Reliable ESG Data Framework

Improving ESG data quality requires a systematic approach.

Companies that successfully integrate ESG into risk management and decision making processes usually begin by establishing a unified data management framework. This includes clearly assigning responsibility for metrics, consistent calculation methodologies, automation of data collection, and the implementation of internal control mechanisms.

It is equally important to work with suppliers. ESG requirements are gradually becoming part of contractual relationships, while suppliers need consistent reporting standards, training, and methodological support. The use of industry benchmarks and external data sources provides an additional way to assess the plausibility of the reported figures.

As regulatory requirements grow, ESG data quality becomes not simply a matter of compliance, but a factor directly affecting business resilience, the cost of capital, and the quality of strategic decisions.

ESG Data as a Strategic Asset

ESG has long ceased to be solely a matter of corporate social responsibility.

Today, it is an integral part of risk management, financial planning, and corporate strategy. Just like financial data, ESG information must be accurate, consistent, documented, and suitable for verification.

Bad data creates false confidence. Good data enables informed decisions.

This is why investing in ESG data quality is not simply preparation for new regulatory requirements. It is an investment in the quality of future management decisions.

How RiskSphere Can Help

At RiskSphere, we help organizations move from fragmented ESG initiatives to an integrated system of data and risk management.

We support clients at every stage of developing an ESG Data Framework, from assessing current process maturity to preparing for CSRD requirements and implementing automated solutions.

Our services include:

• ESG Data Management maturity assessment;

• ESG Data Governance Framework development;

• preparation for CSRD, ESRS, and other reporting standards;

• automation of ESG data collection and control processes;

• implementation of data quality management systems;

• independent verification, documentation, and support for ESG reporting. 

High quality ESG data is not only a matter of compliance. It is the foundation for more accurate models, better informed decisions, and more resilient businesses.