How to structure a monitoring system in the Dark Social Media era

The first myth to debunk is that online conversations are found only in Social Media.

Since the very beginning of our work in this industry and to this day we insist that a sound monitoring project is a Web and Social Media system.

Forums, bulletin boards, blogs, news channels are abundant in meaningful conversations; probably more meaningful to organizations than the noise on the more broadcast-oriented Social Media Channels.

The issue we see here is in approaching the social web from the audience point of view and not from the contact point of view.

Huge audience produces dispersion, useless interactions and a lot of noise, but it’s easy to understand how unchallenging is to sell vanity metrics in a C-room.

Yet, these vanity metrics hardly convert into a business goal, especially if we are talking about conversion rates.

  foto2           foto3

The three sources of a Web and Social Media Monitoring system

There are three pillars in structuring the sources into a monitoring system


Proprietary channels are, or should be, the places where the conversation is mostly devoted to product support, CRM, one-to-one interaction. The real touch point. Among them we can account for social media channels, messaging system, blog, websites.

Social Media are places where conversations happen between users mostly in an unstructured way and in an unfiltered, non-thematic environment. Threads can be related or not to news, comments, information shared or leaked about a brand and a product. They tend to be rather unpredictable abnd usually have a short life cycle, even though sometimes themes are recurring.

The Web is where conversations take place into structured environment (blogs, forums and BBS have a dominant or exclusive theme e.g. automotive, cruise, food, and usually a moderation system in place). They usually have a longer life cycle (a thread could last months).

The conversations sources that seem to be more easily impacted by the rise of messaging systems are the Social Media ones. They will probably be the first sources of information to run dry: quite an easy prediction, given that these are the less structured places and the closer for both the contact list and the UX to the Social Media Channels.

So what?

We suggest that the structure of the typical monitoring system has to change.


According to the industry and/or the purpose of the client, it’s a must to set up filters and priority index.

We worked on a proprietary one based on a mix of quantity, sentiment, helping staff, internal or client, in catching at a glance the key contents.

Index is on a scale from 0,0 to 1,0 with four grades of color as additional hook.

The idea behind is that a prolific contributor tends to have a more loyal audience and to be more viral than a rather unproductive one, though we have to rank him/her in correlation with the goal of the research and measure his/her overall sentiment as an author or media outlet. Eventually this creates a visual map of the contributors with their qualitative positioning, topics, semantic approach.

It’s shifting towards a one to one model that support organizations in getting a deeper knowledge about authors and contributors able to impact on brand perception, to learn their language for a more productive interaction with them, to act quickly when a news is circulating.

Big data are helpful, small data are valuable.