
Top stories






More news








Marketing & Media
Chicken Licken bravely debones a rare phobia with their latest campaign
Joe Public 2 days








I recently saw a subject line of an email newsletter, stating we should stop putting everything on social media. It was a lightbulb moment for me when I read the subject line because recently I interviewed Brent Lindeque, The Good Things Guy, on The Carmen Murray Show and he said something so profound.
“Every morning when I wake up, I can see the mood of the country by what people comment on social media.” We live in a time where people are so vocal and expressive and they suffer from the “hangovers” of everything that is happening in and outside their lives and projecting their emotions on social media.
Sentiment analysis is nothing new and it’s been around for years, but technology enables free text analysis and all of the energy, emotions, anger, sadness and happiness expressed on social media together with emoji could expose some deep insights about our customers.
I approached Dr Isolde Lubbe and Dr Daniel Maduku from the University of Johannesburg to give us a perspective whether sentiment analysis could be the magic wand to understand the contemporary consumer.
“A moment’s insight is sometimes worth a life’s experience.”– Oliver Wendell Holmes Jr
Market research companies often use traditional research strategies – such as surveys, focus group discussion and interviews in their attempt to gain an in-depth understanding of consumers’ behaviour. However, given that consumers rarely do what they say, the efficacy of research results obtained through these traditional research strategies are often doubtful.
An emerging research strategy might be useful in gaining useful insight into consumer behaviour is sentiment analysis. With the rise of social media, consumers have increasingly adopted the medium to express their opinions about organisations and their products. It is estimated that globally, 2.5 quintillion bytes of data is generated every day!
Furthermore, 80% of these data is unstructured, which include texts, speech, videos or emoji and pictures. Consumers are franker when expressing their opinions about an organisation and its products/services on social media than when they engage in research. Thus, processing social media data using sentiment analysis could provide authentic and valuable insights into customer behaviour and preferences.
Sentiment analysis is, therefore, an automated process of investigating consumers’ opinions about an organisation’s products or services from text, images or spoken language of consumers from social media. This type of analysis is all about understanding the social sentiment of the product, service or brand, while constantly monitoring, for example, online conversations.
@NWSCMD "Social Media is something brands need to embrace as listening posts. I manage my twitter handle personally and read all customer concerns on all NWSC social media pages daily. This helps me get feedback from our customers. We also use Barazas, radio stations etc" @nbstv pic.twitter.com/Jk3Lh8kSAE
— NWSC (@nwscug) August 30, 2019
For marketers, the advantages of using sentiment analysis are enormous.
Some of these include:
Sentiment analysis provides an authentic approach to understanding consumer emotions meaningfully. Prof. Antonio Damasio of neuroscience at the University of Southern California reiterates the importance of emotion as it is a necessary ingredient in how we make choices.
For example, when consumers evaluate brands they primarily use emotions rather than information. Furthermore, the ‘likeability ‘of an advert or content, is more predictive of whether the product will be bought or not. Positive emotions have a far greater influence than trust on consumer loyalty and thus the primary reasons why brand-name products are preferred.
Evaluating the important role emotions play in the buying-behaviour process of products, services and brands, marketers should possibly consider acquiring a taste for sentiment analysis. Although there is definitely still a place for traditional research, do not be too sentimental, cut the mustard and start analysing all those visuals such as the emoji!