The rise of social media such as blogs and social networks has led to a proliferation of reviews, ratings, recommendations and other forms of online expression. Bad publicity can very quickly escalate and potentially ruin a brand reputation. Just look at Toyota and the negative influence the recent product issues had on its reputation. Gone are the days when poor customer service would lead to you losing the disgruntled customer and perhaps a few of his mates he moaned to down the pub. Consumers now have the real power to express their opinions and most importantly influence the behaviour of other consumers. To address this phenomenon a new brand of analytics has recently evolved called sentiment analysis or opinion mining. It aims to determine the attitude of a speaker or a writer with respect to some topic/ product/ organisation. The attitude may be their judgment or evaluation, their affective state (that is to say, the emotional state of the author when writing) or the intended emotional communication (that is to say, the emotional effect the author wishes to have on the reader). Measuring the sentiment behind a comment is by no means simple; relying on counting ‘good’ or ‘bad’ words that appear across an entire text can lead to misleading conclusions. In fact, humans often disagree on the sentiment of a section of text, illustrating just how complex a task it is for software to get this right. We decided to have a look at one of the many free tools on the web to see what consumers were now saying about Toyota. Has consumer sentiment been affected by the recall of 8 million vehicles amid concerns that some safety measures have failed, and will there be a long lasting impact on their buying behaviour? The tool we chose is called Twitrratr which searches Twitter for a keyword and cross-reference the results against lists of positive and negative adjectives.
Of 1147 tweets that referred to Toyota, only 36 were negative as opposed to 60 being positive. Surprisingly perhaps, there are still more positive than negative sentiments expressed towards Toyota. Is this a weakness in technique of sentiment analysis or a real gauge of opinion? We only looked at Twitter, so a wider range of sources would give more representative results. However the results are to some extent backed up by the Consumer Reports 2010 Car Brand Survey, in which 60 percent of current Toyota drivers said they would most likely purchase another Toyota when they come to buy a new car. That’s a drop in 10 percentage points from a December 2009 survey, but certainly not a complete breakdown in consumer confidence. The real value for companies comes in spotting trends in Customer Sentiment – identifying if it is deteriorating and taking appropriate action early. Perhaps Toyota reacted to the negative publicity in time with the product recall! We shouldn’t forget that Consumer Sentiment is just one influencer of customer behaviour. It may also be the case that sentiment expressed in certain forums carries more weight than others. Customers will still be influenced by pricing, marketing and their own personal customer experience. Good retailers will use analytics to understand the relative influence of all of these factors on their own customers to decide upon how sentiment should be used to inform their decision making.