eCMetrics in IIeX Latam 2016 Conference

During the past years, social media has turned into a worldwide social phenomenon with Latin America leading the world regions with the highest usage within the Internet space. Currently, more than 95% of Internet users have at least one account, with Facebook being at the top.

As a consequence, a large amount of social media “monitoring” technology tools has emerged and the Corporate Areas of Communications, PR and Online Marketing have been the ones using the tools from the beginning to pinpoint specific dissatisfactions, negative publicity through sentiment analysis and metrics to measure social media effectives. The field of Market Research, however, has taken a long time and is still struggling to use these tools or other techniques, such as Big Data, to generate more in-depth insight from consumer generated media.

Usage of Social Listening Tools and Big Data techniques for Consumer Insights

In the past 3 to 5 years our Company has been working both internally and with clients developing and applying tools and techniques to capture process and analyze consumer generated media for consumer insights.

The objective of this paper is to share our experienced applying these tools and techniques in two (2) specific projects for specific clients in Latin America: 1) social media listening and analysis of opt-in panelists with specific profiles using public and private posts generated by them and their friends and followings in Social Media (Twitter and Facebook); 2) listening and analysis of posts, responses and comments from an insights community of TV user of more than 20,000 participants.

In both cases we will compare the procedures, tools, techniques and best practices used to generate insights which include:

– Recruitment and opt-in process for monitoring

– Mining of post in Social Media and large insights communities

– Key Social Media Metrics automation

– Usage of Supervised and Unsupervised machine learning for predictive modeling of the following:

• Predicting language of posts
• Categorization of Posts based on its contents
• Subcategorizations of posts based on contents and subjects
• Sentiment analysis of contents
• Analysis of emotions of contents generated

-Segmentation modeling of users based on category of contents generated

– Usage of BI (Business Intelligence) Tools to generate dashboard reports (online demo)

– Usage of dashboards for data analysis

source: IIeXLA2016