Using sentiment analytics to measure reputational risk

Author: Karim Derrick

This blog was originally published by Insurance Post, May 2022.

While sentiment analytics is not a new concept, its adoption by insurers has not yet been fully realised. Arguably, its biggest potential is in the area of reputational risk.

Sometimes called ‘emotional artificial intelligence’ or ‘opinion mining’, sentiment analytics mines text to identify the meaning behind the words.

When applied to content such as reviews or social media posts, for example, it is an invaluable tool in helping businesses to better understand how consumers feel about their brand, but may also help prepare for future events.

It’s not a new concept, with businesses from hedge funds to brand agencies making use of its insights for a number of years now.

As well as focusing on the polarity of a piece of text for example whether it is positive, negative or neutral, sentiment analysis also detects specific feelings, emotions, urgency and even intentions.

For insurers it allows the potential to glean information about their insureds in real time and at scale. And in a time where environmental, social and governance considerations are increasingly influential in private and public investment, as well as in consumer-spending decisions, the possibilities for using these insights to better understand and manage corporate reputational risks are very exciting.

Yet the intangible nature of a company’s reputation undoubtedly makes it one of the most challenging areas of risk to manage. But, it is a risk that insurers cannot afford to ignore.

According to a joint Lloyd’s of London and KPMG report, the importance of intangible assets has grown to more than 85% of asset value, with reputation and brand identified as the most important.

You only have to look at the huge financial losses suffered by Facebook in 2018 when it was caught up in a massive data breach and privacy scandal that led to its stock plummeting, to know it’s an asset worth protecting.

Data from Allianz suggests that firms that fail to properly prepare for events that may damage their reputation could see their company value slashed by as much as 30%.

To date only a few insurers have dipped their toe in the reputational risk pond, with Beazley’s product perhaps the most well-known. However, as insurers increasingly realise the capabilities of sentiment analytics, and undoubtedly more tools come to market giving them real-time understanding of the organisations they insure and the likelihood of events that could impact their bottom line, we are likely to see a new chapter of reputational risk insurance.

It is expected that there will be a market shift in optimising existing products that add on reputational risk, such as directors’ and officers’ and cyber security; or towards initiatives that offer standalone coverage of reputation as a discreet hazard.

Of course, insurance alone cannot be the solution. Firms must be able to accurately monitor and measure the impact of ESG-related events as part of an effective risk-management strategy.

We need to get to a stage where organisations can model the complex interrelationships between them and their markets in order to measure any reputational risks from third parties. The ideal world scenario is where there is a fusion of systems combined with expert human reasoning to achieve this.

To date the systems for doing this are limited – largely because the data relied upon is unstructured, uncertain and incomplete, but there is a great deal of work taking place in this area, and this is undoubtedly where the next big product development will come and is where insurers will want to keep a keen eye.