The 2010 Paris e-marketing event that inbenta attended hosted this year 200 exhibitors, 125 conferences and over 11 500 visitors.

One of the trends that mostly caught our attention was the tendency for WebAnalytics platforms to propose "Social Media Monitoring” (SMM) or “Brand Monitoring” tools. These tools, primarily for companies, are designed to measure and analyze what is said about the brand of the company on social medias like Twitter or Facebook and thus get an idea of the “e-reputation” of the company.

If this analysis is relevant, it allows the company being aware of the main complaints, praises, problems, questions and needs expressed by his target audience. Further analysis could also allow getting to know the positive or negative impact of a new marketing campaign, a crisis situation or the influence of competitors on this same audience.
All this information is very helpful for the decision-making process of the different actors in the company: Customer Service & Response, Marketing Performance Tracking, Crisis Management, Technical service for products or services …

Concretely the SMM tools proposed by the WebAnalytics platforms allow carrying out quantitative/volumetric analysis on what is said about a brand on the Internet. This analysis is done in 4 steps:
1) List categorically the keywords to monitor (brand name, name of a product/service, name of the field of activity, words related to a new marketing campaign)
2) Scan social media in search of these keywords
3) Perform quantitative analysis on the occurrence of these keywords over time and on different types of media (Twitter, Facebook, influent Blogs, forums, specialized websites ..).

Some SMM solutions also offer to carry out the same analysis for the competitors of the client-company to enable the client-company to position itself among the competition.

4) Following the quantitative analysis, WebAnalytics platforms create tag clouds highlighting for a given period the most popular keywords  (Example category "Products": Xbox, PSP3, Sony PSP , Wii).

This analysis is very interesting and can track over time the popularity of different selected keywords. Some SMM solutions are also able to match the peaks of popularity with the launch of new competitor’s product/service or the announcement of a future partnership/takeover, … (by using tools like Google Trends / Insight)

However, in qualitative terms such information is not relevant at all and we do not know anything about the value of the popularity: is it good or bad popularity? What measures should we take to improve/conceal this good/bad popularity?

To provide such qualitative information, which is much more valuable than quantitative information, we have to analyze the context in which each keyword has been mentioned. "I am very happy with my Xbox” does not have the same value than "I am not at all satisfied with my Xbox ". The concept of Natural Language Processing (NLP) enables to interpret the meaning of the sentences and their context. The NLP concept would allow, during the quantitative analysis, to weighting each keyword with a positive or negative value and then provide two tags clouds ; one tag cloud showing the keywords with a positive popularity and another tag cloud showing the keywords with a negative popularity. Each good/bad popularity would be explained by the main complaints, raises, problems, questions and needs expressed by the target audience of the company and related to the selected keyword.
This qualification of information, made possible by the Natural Language Processing, would provide high value-added to existing SMM solutions: This solution is available here.