How Can Data Support Your Sustainability Strategy?

The rising expectation for companies to take responsibility for their impact on people and the planet is driving the need to implement and operate a sustainability strategy.

Optimally, this sustainability strategy would be interwoven with a company’s overall strategy. However, the reality is that for many companies, integration of sustainability throughout the business is still a longer-term goal. Whether this strategy is standalone or integrated, companies are facing increasing pressure for transparency around their sustainability policies, actions and results – not least in response to the various supply chain sustainability due diligence requirements coming into force across the globe.  

To remain competitive in this climate - where sustainability factors increasingly influence procurement, investment, employment and other key decisions - companies need to adapt at an ever-faster pace, leveraging high-quality data, translated into a digestible form. Companies must have a robust approach to identifying, collecting, processing, and collating quality, relevant non-financial data from across a range of business functions.

Data is Key, but Challenging

Data enables measurement, performance analysis, risk management, and informed decision-making, as such, it should be integral to any sustainability strategy. There are many data points and indicators that support companies to monitor and improve performance on sustainability-related issues and compliance with current and upcoming legislative developments. Below are some examples of relevant data points and indicators:

While data points and indicators, such as those above, now frequently feature in sustainability reporting, collecting, processing, and collating this data are often fraught with challenges and hidden barriers that need to be considered to fully unlock their value:

1. Data identification: Defining the relevant sustainability data points for a particular company or industry, and establishing the right data requirements can be very complicated given the multiplicity of the data points and the absence of standardized metrics and reporting frameworks for ESG data.
2. Data collection: collecting data for purposes of a company's sustainability reporting obligations can be very challenging. Data comes from disparate sources (internal, suppliers, third parties, etc), and is distributed across multiple systems heterogeneously. Companies can achieve substantial benefits and innovative solutions where they embrace intra and inter organisational collaboration to solve these challenges.
3. Data quality: sustainability data may be compromised, incomplete, incorrect, and/or outdated, especially if extracted from systems that have not been primary sources in the past, where less rigorous controls are in place, or where the data does not lend itself to all of the necessary quality and integrity checks. Where data is found to be of a low quality, it will have very limited value. This is known as the ‘Garbage in, Garbage out’ principle, and may result in inaccurate reporting. Data analytics experts will be pivotal in establishing the required rigour checks to support the capture of high quality, accurate and complete data.
4. Data modelling and interpretation: Quantifying certain sustainability measures and metrics can be very challenging. In some cases, such as calculating the carbon footprint of an organisation, it can require multitudes of data sources to feed into complex models and calculations. These models might also be based on assumptions and have limitations that need to be considered carefully when using the model’s outputs.  In other cases, not all the data points will have been measured historically, and some still pose challenges for companies to determine a sound approach to measurement, particularly in relation to social data (e.g. ethical behaviour).
5. Data efficiency: companies will want to take advantage of existing data used for other purposes and avoid duplication wherever possible. While this is generally good data management practice, it is also an important consideration if companies are to avoid unnecessarily contributing to the increasing levels of emissions generated from the volumes of data we are creating and storing across the globe. Environmentally conscious choices should be made throughout the data storage lifecycle, from data acquisition to disposal (Data decompression, deduplication, energy efficient hardware, Cloud Computing, etc).

Leveraging AI

There have been a lot of conversations around AI and its capabilities over the past years, including ethical considerations and the risks of inadvertently introducing bias into datasets. But AI has real potential to achieve efficiencies, fill the gaps in sustainability datasets and even solve many of the challenges above. AI-based data collection procedures support real-time data extraction and updates, as well as automated cleaning processes including missing data imputation, for both structured and unstructured data.

AI also facilitates data standardization and harmonization as collected data can be classified, labelled, and converted into a usable structured format.

Advances in Natural Language Processing (NLP) are enabling the automation of various tasks that require significant human effort, by analysing data from billions of online sources like social media, blogs, and forums, etc; and applying techniques to evaluate conversational tone and sentiment, potential negative news and scandals related to human rights and environmental violations for example. The importance of such tools has increased significantly as many interactions have shifted to social media.

In addition, AI opens a wide range of potentially exciting opportunities for companies.

• Predictive models can be used to establish environmental, social and governance ("ESG”) ratings and scoring models aiming to evaluate the sustainability performance of a business. These ratings are used to support investment decision-making, improvement remediations, and risk evaluation and management.
• AI can optimize energy consumption, resource allocation, and waste management in various sectors.
• AI can enhance supply chain transparency and sustainability to detect any potential risks related to labour practices, environmental impact, or human rights violations.
• AI can analyse climate data, to predict and understand climate change impacts and put in place effective climate risk management and resilience planning.

Responsible Use of Data

A sustainability data strategy should be responsible, transparent, and ethical. Companies need to think critically about all potential consequences of how the data is collected, used, and shared.

• Companies need to adopt a ‘compliance by design’ approach and responsible data practices by safeguarding privacy, promoting trust, and mitigating the potential risks associated with data usage and AI.
• Only necessary data should be collected, in a transparent manner, and used only for legitimate and specified purposes.
• Any collected data should be managed with robust security measures to protect it from unauthorized access, breaches, or misuse.
• Personal data should be anonymized or removed from the collected samples. Organizations must adhere to privacy regulations and implement appropriate measures to safeguard personal data.
• To the extent possible, data should be cleaned to limit the introduction of bias (systematic and unfair inaccuracies or prejudices), and measures like identifying potential sources of bias, using human oversight, and removing sensitive data points - should be put in place to help mitigate bias and promote fairness in the algorithms used to process the data.
• When using data for developing and deploying AI, algorithms should be designed in a responsible and accountable way under strict governance.
• AI processes are power intensive and therefore can have negative impacts on the environment. The process of training and operating models requires a lot of energy, resulting in air pollution, water usage, and carbon emissions. Limiting the environmental impact of AI can be performed by choosing energy efficient hardware, using renewable energy sources, and establishing a sustainable AI strategy.

To Conclude

As sustainable operating practices are adopted even more widely by companies, we expect to see greater integration of sustainability into business strategies, increased transparency, and a focus on measurable impact, especially with the increased enforcement activity related to the topic. Sustainability considerations have become essential for long-term business resilience, attracting investment, and maintaining a positive reputation in an increasingly socially and environmentally-conscious world. It’s only by embracing data and finding innovative solutions to the data challenges faced in this context that companies can truly operationalise sustainability and achieve their commitments to building a more sustainable and equitable future.

Thank you to contributors Jack Simbach and Viktor Josefsson.