Much of the user generated content from social media sites has become ubiquitously integrated into the search engine results. Try doing a search on a product or service and you won't just see the product page or home page of the company but a list of comments and postings from social working sites such as Twitter, Facebook or here in China, many postings from BBS's.
It is well researched and documented that the consideration and purchase intent is impacted by comments online. However, I want to increase the awareness and use of the notion of 'toneism' (which is really the sentiment proxy) with comments (negative = 'black buzz' and positive = 'white buzz').
Many Marketers that I speak to have only a basic understanding of the ideas of sentiment/commenting on the purchase intent of online buyers. It would be true to say, many Marketers know negative comments are bad... positive good... but most have incorrectly assumed that a negative sentiment index charted by their online reputation tools has a direct correlation with black buzz (comments). Not true. This post will provide greater insights for the marketer on sentiment and the relationship between this and comments.
There are a number of online reputation monitoring tools that will automate the finding 'discovery' of information online and then apply sentiment measurement, (which can also be known as "toneism" or "tone analysis") to give the Marketer a snapshot about the conversations online about their Brand, Product or Service. There a growing number of companies moving into this expanding sector but there are only a number of vendors I think are doing a better job providing tools for marketers such as Buzzmetrics, Radiant6, BrandsEye, and SinoBuzz). These four products all use analysis to determine an online sentiment. (Note: If you think there is another great product for marketers, then comment below so I can review for future posts)
Sentiment measurement, can also be known as "toneism" or "tone analysis". For those that are not sure how most vendor's implement this, it is a sentiment classification that is keyword-based. In this approach, terms, mainly adjectives (e.g. awesome, awful, good, bad, love, hate) and fixed expressions (e.g. police state, on cloud nine), are used as sentiment indicators. The list of indicators can be prepared manually (the most common approach), composed semi- automatically, or acquired by machine learning algorithms that infer the best indicators from tagged samples in the domain of interest. I won't give away all the secrets of the online reputation management/monitoring vendors but the language taxonomies that are supported with natural language processing (NLP) provide the heart and soul to their solutions.
But the sentiment analysis information that is presented by the online reputation tools only goes part way for a Marketer in understanding the dynamics affecting their Brands image and positioning. For example, does the negative comment (the online black buzz) always have a negative tone? Not necessarily. The white buzz can also have a neutral tone (unlikely to have a negative tone, although possible).
Further, the extent that online comments can impact on a Brand is not simple to determine. There are many factors that contribute to the impact on the Brand such as the context of where those comments where discovered. For example, where the comments found through an online search (i.e., Baidu/Google), within an online review site or within the feedback page of a retailer site? Moreover, the influence of comment information is based on the perceived reliability and intentions of the writer. Again, impacted largely by familiarity/trust one already has with the Brand, product or service.
So for us Marketer's, I would encourage looking beyond the pretty charts, plotting the sentiment over time to overlaying the actual comments (black/white buzz) and look for the inter-relationships. Don't use a simple comment count with a negative sentiment score as the metric for black buzz.
This deeper analysis will set your social media strategy on a steadier foundation and bring much greater value to the brand managers and PR teams trying to leverage/address online commenting.
Link to the original post