As different social media types – blogs, wikis, communities, discussion boards – become more and more popular, marketers want tools that can capture and measure buzz. Buzz numbers are the most popular metric used in that context, followed by –simply defined – influence. Share of voice expressed in the number of social media articles discussing a brand in given period of time is the measure that is used most often. It has certain advantages: it is easy to obtain using different free tools and it is straight forward to understand and cross compare with other brands. Moreover, as proven in several studies, it is highly correlated to sales for certain product categories, especially FMCG but also consumer electronics. Despite that, share of voice has low explanatory power; it delivers only quantitative information and no insights.

To gain more actionable qualitative information all social media articles can be analyzed and classified in topics they discuss regarding the brand in focus, but also in being positive, negative or neutral towards the brand. Sentiments in social media reveal how people feel about the brands and topics. There are almost always more positive than negative posts (although most posts, up to 90%, are neutral as we discovered in our studies around cars or mobile phones), but one positive post is not equal to another. They can discuss brands in general, referring to customers overall positive experience, or they describe particular products. Especially in the consumer electronic industry, general trust towards a brand is often exchanged for experience with one particular product, because users know there are significant quality differences between two diverse models of mobile phone made by the same manufacturer. Information gained from these positive posts helps to understand why a consumer chose one product over the other, even if the products are very similar.

On the other hand, there are always some negative posts. As Alain Samson notices, “even if it were the case that individual purchasing attitudes are affected more strongly by positive word of mouth, negative comments are likely to survive more easily across time and space. Negative information doesn’t only stick, it sticks longer. At the same time, bad buzz will persist longer if it is communicated to more people”. Simply, people are more likely to remember negative rather than positive messages, and they also give more weight to any negative information that they receive.

Insights that are gained from the negative articles in social media can be used for brand protection and brand reputation management. Some brands – Dell Hell being the most known one - actively monitor the blogosphere for possible negative issues and the number of brands doing so is continuously increasing; some brands could also prevent an growing issue in the past by detecting it on time and taking measures.

Positive and negative word of mouth are very important to gain good, qualitative insights regarding customer feelings, but are they equally important? Does one positive post have the same weight as a negative one? Alain Samson claims in his study, that the weight and importance of positive versus negative buzz depends on the industry. To understand which sentiment will have greater importance for a given industry, the industries must be split into high versus low commitment and high versus low choice industries. Many products, among others mobile networks, financial services or cars, are subscription products, of which mostly only one is used by a consumer for given period of time. These products tend to have fewer competitors and therefore consumer choice is limited. Entertainment industry on the other hand is an area where many products – movies, CD’s – can be used at the same time and also easily exchanged as there are many competitive products people can use.

Negative word of mouth should be more important for high commitment, low choice brands like cars. The automotive industry has high number of consumer advocates for two reasons. On one hand, it is a subscription brand, and this is a category where people – having to use one product for longer time – are compelled to be more convinced about and to which consumers are generally more committed. On the other hand, people stuck with expensive subscription products that they don’t like as much as they expected to tend to rationalize their choice anyway. People are also likely to expend more effort to find the right product if it’s a high cost subscription brand. Negative sentiment will have higher damage within these products.

For industries within which high choices exist, positive word of mouth will have more influence on the final selection. For example there are many movies that are played at the cinema and for most viewers it will be difficult to choose from 3 or more potentially good movies unless they receive recommendations about one of them. If they receive negative information about one of the movies and exclude it from their choice, there are still 2 or more options remaining so that exclusion is not the best selection process within industries with many products.

Whatever sentiment is more important, one thing is certain – sentiments in social media are extremely relevant and cannot be neglected. However, given the amount of articles posted by social media users every day, it may be very time consuming to analyze all articles about a given brand. Automatic sentiment detection is already built into Attentio’s Brand Dashboard, analyzing the sentiments in real time for our customers.

Digg this » Add to del.icio.us »