Let’s not be Sentimental about Reputation
For a discipline that is concerned with reputation management, PR takes a surprisingly simplistic view of how media coverage reflects sentiment. To understand reputation, we need to understand not only what audiences think and feel about organisations and brands, but also why they feel that way. Reputation is, ultimately, an asset (or liability) built on emotional responses to behaviour and performance.
That’s why we can’t categorise sentiment in media coverage accurately by measuring positive / favourable versus negative / unfavourable mentions. The same applies to content from online customer reviews and to the responses we gather from traditional forms of market research, including customer surveys or brand tracking. Emotional response is not black and white. To treat it as such is limiting our ability to understand audience motivation and objections.
According to the system developed by psychologist Paul Ekman, there are seven core emotions which are easily discernible in human communication: Happiness, Sadness, Anger, Fear, Surprise, Disgust and Contempt. When used to evaluate response from consumers or media, they take us beyond a polarised view of ‘sentiment’. This provides richer insights about the emotions driving opinions and creating behaviour change in audiences.
When we analyse emotions in media coverage, we should also weight it according to the size of audience it has actually reached. Media evaluation that counts raw volume of negative articles versus positive articles is missing a very important point – some of those articles will have been read by tens of thousands of people and others by a mere handful. A negative article on dailymail.co.uk with millions of daily visitors will almost certainly create Anger, Fear, Disgust or Contempt among a far larger audience than an unfavourable piece in a local print publication. The impact on reputation is therefore very different.
At the recent CX Emotion conference in London, we heard a lot about how AI might be used in future to create an even more detailed picture of the emotions, attitudes and cognitive states reflected in audience response. This will be very powerful if combined with the ability to identify the actual size of the audience and an accurate understanding of the themes and topics that explain why people are responding in this way. Binary sentiment analysis is a very blunt instrument compared to alternatives that already exist, but technology will take this aspect of reputation management into another realm entirely and that’s something we can all be positive Happy / Surprised about.