The explosion in both quantity and popularity of social media sites has proved something of a double-edged sword for many web-based businesses; on the one hand social media offers everyone an equal playing field in which to promote their brands and business, and gain and retain customers. If executed properly it can reap the greatest of rewards, but, because this platform is a relatively new one in comparison with ecommerce sites, tracking and analysing user information and how it relates to the business in question has so far proved a little tricky.
While various analytics packages offer referring site metrics at a glance, customer sentiment and opinion about a brand has proved somewhat elusive and required the manual searching and trawling for information that most business owners simply don’t have the time for.
In an attempt to address this, IBM this week has launched its new social media monitoring tool, called the SPSS Modeler data mining and text analytics software. This will, it says, allow the monitoring of changes in customer attitudes across social media such as Twitter, Facebook and the blogosphere, uncover deep insights in doing so, and predict factors which will in turn drive a brand’s future campaigns for customer acquisition and retention.
Analytics for social media
The software has the capability to get to grips with the jargon and slang common to social media as it will use Natural Language Processing, meaning it can extract data from slang, industry jargon and it even scans emoticons, giving a better insight into customer sentiment toward commodities and brands by reading meaning for non-textual symbols.
The software cleverly forms its own new semantic networks by analysing trends and insights from social media relating to a particular topic, collecting synonyms that mean it can understand when different terms are referring to the same product or brand; for example the software will recognise the terms ‘floating rate’ and ‘mortgage loan’ as the same thing within its semantic network.
The data gleaned from social media can then be integrated with a company’s own internal analytics data which should provide a comprehensive snapshot of the market and form a strong basis for campaigns.
Understanding the importance and meaning of colloquialisms could play an important part in analytics for social media in the future as the web increasingly moves towards a more sociable, rather than searchable tool, with recommendations from online communities becoming an increasingly pivotal product purchase push.
Related posts: