The importance of Social Media Monitoring

”If you make customers unhappy in the physical world, they might each tell 6 friends. If you make customers unhappy on the Internet, they can each tell 6,000 friends" said Jeff Bezos, the CEO of Amazon.com

With the appearance of the so-called Web 2.0 media, the web user has become an active player and can create, organize and broadcast content of his own. Expressing one’s own opinion and relaying it to the widest audience possible is no longer the sole prerogative of journalists and technology buffs. Thanks to the availability of ever simpler and increasingly collaborative tools, every user connected to the net is a potential form of media: users may discuss your company on their blogs, post comments on a social news site (OhMyNews, TPM Café, Digg, Newsvine), take part in a wiki, give their views of your product on a consumer opinion platform (epinions, ConsumerReview), create a file about your services on a social network (Facebook), etc.

That's why "Social Media Monitoring" has become an increasingly field of attention by companies and institutions alike.

Using Semantic Technology

The challenge now is: How can we monitor sources of information containing hundreds of thousands of comments and posts? Can we extract what is really important and commonly mentioned there in order to take action?

At Inbenta we have used our Semantic Search technology, particularly our Semantic Clustering to extract thousands of comments from the Internet, and grouping them by their meaning, creating groups of comments that are similar in meaning, (we call them "semantic clusters"). That allows to monitor which are the actual trends behind tons of comments that we couldn't deal with otherwise.

The case of Ciao and ING Direct 

We have conducted this study with www.Ciao.es concerning ING Direct in Spain, and the results are straightforward:
Advantages of the "Orange Mortgage"

• No commissions (42% of opinions)
• Low rate interest (38%)

Disadvantages:

• Bad customer care (60%)
• No physical offices (20%)

Taking together these advantages and disadvantages our "semantic clustering" algorithm has grouped 80% of all opinions, thus proving a concentration on these kind of opinions over all others.

We are developing now new connectors for other social medial websites, that will provide a wider vision of opinions in the social media.