Social media effectiveness (sentiment track)


A leading CPG player

Business objective

  • The client wanted to create new categories in the fashion styling space, by understanding customer preferences in adjacent categories.
  • Our client leveraged social media data from multiple sources. However, each source had a different format. Also, the client had to handle significant unstructured data comprising of relevant customer reviews and feedback.

Benefits & outcomes of our engagement

  • Identified new product concepts in the personal grooming space that had a high sentiment score (tweets, retweets, posts, etc.)
  • Assessed the intent to purchase, perception, imagery and consideration towards a hypothetical set of new product concepts
  • Built an ongoing tracker that consolidates social media data-based user sentiments with online panel- based brand imagery

SGA approach

Step1 - SGA’s web scraping module
  • SGA’s social media analysts used python-based web-scraping algorithms to extract unstructured data such as product reviews, comments, engagement and posts on social media forums
  • Arrived at key social media KPIS using topic analysis, keyword search, taxonomy, word cloud, network analysis, word frequency and association

Step 2 - Online panel setup & recruitment
  • SGA’s research experts set up an online community
  • Identified target segment respondents across relevant markets (Singapore, Thailand, Indonesia)
  • Conducted opinion poll and qualitative surveys for concept testing
  • Assigned activities to online panel participants including uploading style statements, selfies, favorite variant, etc.

Step 3 - BI visualization
  • Integrated data from online panel and social media to report key metrics like brand sentiment, social media reputation score, intent To purchase, brand consideration, etc.
  • Published visualization dashboard on Tableau server with per hour refresh that gave business stakeholders a holistic overview of their customer’s voice