The progressive expansion of Corporate Social Responsibility (CSR) has been accompanied by an increasing interest from regulators and market analysts. Governments and supra-national organisations have issued guidance rules on CSR, while market analysts have created a set of gatekeepers focused on its evaluation, publishing rankings and comparative reports. The UN Global Compact and the sustainability indexes are two relevant examples. The complexity and some of the functions of this CSR infrastructure have common features with the financial system. Information is at the core of both. The distinction between information and noise is central for building up efficient financial markets. The aim of this paper is to analyse how information can be separated from noise in CSR. To this end, we develop a qualitative model that centres on the following variables: the CSR features of the project under consideration, its financial features, its relationship with corporate strategy, the performance metrics for its analysis, the different kinds of risk it involves, and its impact on value creation. This model relies on two common functions that we identify in the CSR infrastructure and the financial system: the defining function and the performance information function. The model is applied to Adidas’ CSR policy