Sentiment analysis is invaluable for understanding risks to corporate reputation. How can compnaies better approach this analysis? Co-founder and global MD of corporate risk analysis company Polecat, Bronwyn Kunhardt, talks risk and reputation

How do you define sentiment in this context?

Sentiment analysis emerged with the advent of social media as a way to estimate and diagnose the attitudes and opinions of the population. It provides a binary indication as to the moods being expressed – positive or negative – and business leaders can extrapolate basic information from this. It’s a fairly blunt, black and white categorisation of a complex concept.

Why is sentiment analysis used to make decisions?

Sentiment analysis can sometimes be one input to decision-making. It provides a high level steer on perception, providing a rough estimation of general mood. This can be useful when business leaders require an indication of public opinion, but good decisions typically need to be based on a far more nuanced and thorough analysis than a ‘happy hashtag.’

How is that use flawed?

Sentiment’s flaws become apparent when trying to understand the concept being discussed. For example, we do a lot of work with the NHS, where illness, anxiety and pain of course come up a lot in the discourse, but not necessarily to express a value judgment about the NHS or care received. Sentiment analysis isn’t good at navigating these sorts of nuances or their implications: sad doesn’t always mean bad.

Understanding risk topics – and making important decisions in response – requires more than a binary portrayal of public opinion. It needs leaders to get under the skin of the risks they face – understanding not only more of the what, but also the why.

What are better metrics to understand and mitigate risk?

If you want to mitigate perceived risk, you need to know how different stakeholder groups are engaging with a spectrum of risks that are likely to be important to your organisation and reputation, and you need to know how those groups are influencing each other’s views and ultimately your company’s licence to operate in the court of public opinion.

Unstructured data can be mined for this precise type of intelligence – you can interrogate how different groups or individuals are discussing specific topics related to your brand or business, and how that view may be trending over time and impacting others – all via a single dynamic dashboard.

What kinds of businesses does this apply to?

Almost any business can benefit from close risk analysis. Those for whom the need is most pressing tend to either operate in volatile markets, have complex supply chains or do business in regulated industries.

Recent developments in legislation have shown us that it is no longer sufficient to plead ignorance. Business leaders must be actively monitoring for hazards that could have an impact on their business or, in some cases, are already affecting their business without their knowledge.

How does this impact on corporate reputation?

Monitoring for an assortment of risk-associated topics delivers powerful intelligence to help manage corporate reputation and safeguard licence to operate. Unstructured data can be mined to understand attitudes to the key identified value drivers for a brand and how those attitudes are trending over time. It can also identify specific issues that may be undermining perceptions of those brand values anywhere across value chain. The resulting intelligence can inform a company’s reputation management strategy and also provide a measure of that strategy’s success.

What is the biggest change companies should make when understanding and analysing risk?

Risk is different for each and every company, and no business can define its key risks and material issues without keeping one eye fixed firmly on its internal activity and the other on the rest of the world. Traditionally, companies use some periodic opinion surveys and occasional stakeholder engagement to understand issues and expectations and rank their significance for the business.To navigate this new terrain, companies need to embrace AI solutions that ensure they interlink with the markets, societies and communities on which they depend.