The Corporate supply chain sustainability due diligence act (CSDDD) currently advancing through the legislative process, along with similar national laws in Europe (e.g. Germany’s Supply Chain Due Diligence Act; Lieferkettensorgfaltspflichtengesetz or LkSG), underlines the growing importance of businesses monitoring their supply chains to ensure compliance with ESG regulations.
Given the complexity of modern supply chains and the prevalence of outsourcing, keeping tabs on suppliers is a formidable challenge.
To address this, digital platforms have emerged, leveraging data to provide ESG visibility scores for businesses.
One such platform is Prague-based Semantic Visions. The company says it collects and processes millions of documents from open-source intelligence sources on a continual basis, and promises to deliver high-accuracy data on over 120 different ESG-related event types.
However, there are plenty of questions that arise nowadays regarding the accuracy of data and the conclusions that can be drawn from it, particularly in the logistics and freight forwarding industry.
So to learn more about the concept of ESG visibility and what Semantic Visions’ data says about the forwarding sector, we spoke to Sari Serhan, CCO of Semantic Visions, and Lukas Gabor, the company’s ESG subject matter expert.
What is the value of real-time ESG monitoring?
To address this question, we quizzed Semantic Visions about its value proposition and whether there is a genuine need for such a service.
According to Serhan, the rapid evolution of global policies and regulations necessitates that companies stay ahead and maintain visibility across their entire supply chain.
“One of the things we’ve noticed over the last 18 to 24 months is a forced shift towards compliance with regulations that are coming thick and fast,” Serhan told Trans.INFO.
This visibility, Serhan said, is crucial for ensuring compliance with ESG standards and regulations throughout the supply chain and across the multiple-tiers and relationships within a client’s operating eco-system.
“It is very dynamic and from one day to the next, you could find out that your supplier is linked to child labour abuses,” warned Lukas Gabor, Semantic Visions’ ESG expert.
What makes data ‘high-quality’?
The rise of digital logistics and supply chain startups in recent years has sparked extensive debate about data quality and actionability. So, what does Semantic Visions have to say about the way it gathers and presents its data?
Serhan told Trans.INFO that the company uses a “Robust Data Platform Enhanced by Artificial Intelligence and Automation” to continuously monitor and track over 680 global events, covering operational, financial, and environmental impacts, both man-made and natural.
Semantic Visions states that its platform’s ability to precisely correlate these events with identified entities of interest, allows companies to “rely on these signals to stay ahead of and mitigate potential risks”.
Moreover, Semantic Visions’ says its ontology team trains its models using native language speakers who are able to “identify nuanced language within processed data objects that may not be identified through the use of machine translation”.
The company adds that this is “further strengthened by a focus on transforming unstructured data into actionable insights using AI/ML technology, ensuring over 95% precision in event and entity identification and ensuring relevance of results”.
According to Semantic Visions, this approach helps businesses understand the full spectrum of risks they may face. Finally, the company stresses that it does not use self-reported data, which it believes can often be biassed or inaccurate, while a ten-year historical database also provides context for assessing long-term ESG performance.
“From a screening perspective, our top priority is to maintain our database extremely well,” Serhan emphasised.
The aim of this historical data, says Semantic Visions, is to aid in ESG compliance and due diligence, helping companies identify potential risks before they escalate. Currently, the company says it maintains over 3 billion curated and enriched data objects in a 10-year historical archive.
“We collect data in 12 different languages. Without this capability, you can’t identify all potentially relevant events. Our multi-language capability provides clients with localised intelligence, perspectives, and insights from various sources, including published articles, newsletters, blogs, and industry reports,” Serhan told Trans.INFO.
By categorising events into beneficial, neutral, and adverse, and breaking down scores into subcategories, Semantic Visions claims companies can pinpoint specific ESG issues within their supply chains.
“We have a robust core capability and methodology that allows us to collect millions of data objects from open-source intelligence sources and web-based media. This enables us to enrich the data, correlate specific ESG events to entities of interest, and deliver the identified signals through Semantic Visions’ ESG Index or through tailored datasets,” Serhan stated.
What kind of ESG scores are most common?
What does this data say about the state of ESG in general?
“What we see is that the best performing companies have a score of around 80 and above, which indicates they’re performing very well,” Serhan told Trans.INFO.
However, Semantic Visions’ CCO added that when breaking down the overall ESG score, users may see scores well below those numbers for particular categories.
“This would suggest there are negative events affecting the overall score beyond what is seen at the aggregated level. The ESG overall score can be broken down further to see specific events and their impact. For example, we had a client monitoring a long-term ESG score for themselves, and one day the score dropped by 40%. This is a huge drop that’s highly unusual. So we drilled down and identified the specific events that caused this drop, and learned that the company was actually subject to a lawsuit” Gabor explained.
Regional variations in ESG risks
What does Semantic Visions’ data tell us about the most-common ESG risks in different areas of the world?
Serhan told Trans.INFO that there are noticeable differences between regions:
“It is quite region specific. In South East Asia, you tend to find fewer concerns about governance but more risks related to the environmental and sustainability areas,” said Serhan. “ On the other hand, in Western Europe and the States, you tend to find that while the sustainability scores are good, there are question marks over governance. So it depends on the region.”
Sector-specific insights
Which market sectors have the most room for improvement regarding ESG?
According to Semantic Visions, certain sectors, such as Oil & Gas and Automotive & Auto Parts, have more room for improvement in ESG performance.
“According to Semantic Visions’ ESG Index, trends clearly indicate that the Oil & Gas and Automotive & Auto Parts sectors have the most room for improvement,” stated Gabor. “We found that this underperformance may be due to their operations being primarily associated with negative environmental impacts and regulatory challenges.”
The most common ESG risks faced by freight forwarders
For freight forwarders, Semantic Visions’ Index shows that environmental risks, such as emissions and pollution, are significant challenges, but that workplace safety is the most prevalent concern.
“The most common ESG negative events for freight forwarders are workplace incidents, which fall under the ESG Social Factors categorisation and include issues such as worker safety, labour practices, and human rights,” Gabor explained.
What the top ESG-scoring logistics companies are doing particularly well
Semantic Visions’ Index also shows that the top two logistics companies with particularly good ESG index scores are Old Dominion, Freight Line, and Expeditors International of Washington, Inc.
“These companies have higher ESG scores due to an exceptional ratio of positive events relative to the total ESG events associated with them. They outperform other companies in sustainable business practices, sustainable supply chain practices, and emissions best practices, explained Gabor. “Specifically, they have a lower negative impact on both the global and local environment, community, society, and economy while also maintaining progressive environmental and human rights policies.”
The developments that could boost ESG monitoring going forward
Given the advancements in technology we are seeing at the moment, we asked Semantic Visions what technological developments could allow for more comprehensive ESG visibility in the future.
The company believes that the development of Multi-tier Supply Chain Mapping and Large Language Models will particularly help here.
“We’re one of the early and successful adopters of large language modelling. The one current challenge here is that most large language models are best used in the English language, but we cover 11 other languages. So, as the use of large language models becomes broader and more prevalent, I believe we’ll be able to enhance our solution further to provide a much more holistic solution that employs large language modelling across all available languages,” said Serhan. “Additionally, given the demand for greater Supply Chain transparency and resilience, we are currently deploying a multi-tier supply chain mapping solution. I think that’s really where we’re heading in terms of the adoption of the latest technologies that are out there.”