Social media is a treasure trove of insights and useful information, but most people have realised that the path from hype to reality is hardly straightforward or easy to implement.
At the heart of the matter, information gleaned from social channels is a combination of structured and unstructured data infused with human errors, typos, acronyms, emoticons and other semi-recognisable symbols and expressions.
This variety of formats means that, while organisations may want to treat social media like every other data source in their information management portfolio, gleaning insights from these channels necessitates an entirely new approach.
The key to solving the social media data conundrum is embracing the concept “the whole is greater than the sum of its parts”. Rather than reinventing the wheel or making risky, costly investments, companies can use their existing knowledge and technologies in new combinations to gain valuable insights from social media.
Traditional business intelligence (BI) tools can be used to handle the structured data from social media. Although most social network services only provide access to data for a limited time, companies can leverage their existing storage capabilities to store social media data in a repository to maintain history.
This will enable easy historical trend analysis for any time period for which the data was collected. So while existing BI capabilities make the analysis of structured social data fairly straightforward when it comes to unstructured information, a little clever innovation is acquired.
How can companies analyse social conversations, determine what topics customers, prospects and competitors are most frequently discussing and integrate these findings into existing databases?
By introducing text analysers into search technologies, organisations can quickly find any relevant text, keyword or phrase in unstructured social data.
Business users can then extract the word frequencies and visually depict them in tag clouds or streaming graphs, easily displaying how these words and phrases are mentioned over time, and how these trends may vary based upon company news, new product introductions or other such criteria.
The next step, one that makes the sum of all the parts, is to share all of the data stored in the search index and database in an interactive dashboard. In addition to easy visualisations of tag clouds and stream graphs, business users can develop a faceted navigation based upon both the structured and unstructured content.
Sliders can be utilised to see the most frequently used words – or on the other side of the spectrum – the least frequently used and further analysis can be undertaken from there.
For example, if a word or a combination of words is of interest users can drill down and retrieve all related messages and ascertain the exact meaning in a matter of seconds. While this is not semantic analysis, it is extremely quick and accurate and enables companies to understand what’s happening on social channels, and how it relates to company news and other initiatives.
Users can only expect that organisations will grow more reliant on social media for all facets of their business in the years to come. As these channels become more popular, technology vendors must continue to innovate their solution sets to enable greater analysis of the resulting data.
However, new technologies and enhanced skills are not necessarily required now for companies to start obtaining valuable insights from the conversations happening on these platforms. By re-engineering existing technologies, today’s business users can access the depths of social media data and integrate it within their existing information management portfolio.