Technology Measurement and monitoring (social media monitoring tools) solutions are just tools. Real insight comes from human analysts that understand how business works and know how to use these technology tools.
Technology tools are helpful in solving client problems and for automating some tedious tasks.
Business is an eager user of a wide range of technology tools, but those that put their faith in technology to be the “silver bullet” will end up disappointed.
Natural language processing and text analytics are mature technologies and their benefits are well understood, but there are challenges. Automated sentiment analysis is difficult enough to get within tolerable margins of error (and sometimes even one error is enough to undo credibility in the eyes of an executive).
However, we are still some distance from a “magic algorithm” that can decide whether a website, or a blog post, matters or not. For the ultimate analysis, basic human judgment is still required to analyse the various threads.
It’s still extremely difficult for machine intelligence to identify the most important sites and blogs of conversation on a given topic i.e. A machine cannot understand the difference between ‘My bank sucks’ and ‘MY BANK SUCKS’ or ‘my bank sux’.
The interaction and commentary on a social networking site is often not logical or rational. A lot of slang and incorrect spelling is used, and is therefore beyond the capabilities of a software program to understand.
Furthermore, software can’t distinguish sarcasm, neutral feedback or implied criticism and irony.
Sentiment analysis is extremely difficult for a piece of software to analyse –at best it is 60% – 70 % actuate.