The Kids’ basic cable and satellite television network of a large US-based global mass media company with interests primarily in cinema and cable television.
The client wanted to use Natural Language Processing techniques to understand the customer sentiments from various social media sources namely Twitter, FB, Vine, Tumblr, Youtube, Instagram and Pinterest as well as combine the social media data with TV Ratings. Specifically for Facebook data, there was a need to map Facebook interactions with their shows and ratings as well as extract relevant variables for the selected TV shows.
The client specifically wanted to answer the following business questions:
How does audience engagement in social media correlate with the TV program’s performance?
How does the performance of TV programs relate to respective social media posts from the official account?
Does direct engagement with audiences help increasing ratings?
The client wanted to understand customer sentiments as well as plan their customer outreach strategy and needed help with extracting useful insights from the massive pile of unstructured social media data.
They also wanted to automate the consolidation of all social media data into one master database that was always up-to-date to enable an efficient data analysis.
SG Analytics provided a consolidated environment to capture social media data from all different data sources, either via API’s or by utilizing web scraping techniques. The data collected from all the different sources such as Twitter, FB, Vine, Tumblr, Youtube, Instagram, and Pinterest, was collected daily. According to the specific engagement metrics of the source, the data was appropriately aggregated at a weekly level for storage purposes.
Specifically for Facebook data, the data was extracted, cleaned, given sentiment scores, aggregated and then compared with the TV rating data using correlation analysis and regression models. The output for this model was the right genre classification for the facebook genres which were then reallocated.
SG Analytics also took business inputs from the client regarding the various KPI’s across geography, time, and type of channel as per the business need and created a dashboard which helped the client visualize the real-time data with the help of Tableau and D3 technologies. This helped in week-on-week viewership tracking on social media to understand the major drivers impacting viewership and plan different marketing strategies.
SG Analytics helped the client identify customer retention drivers and predict customer lifetime to assess the optimum length of trial periods, cadence for video release, and discontinuation of non-performing shows.
SG Analytics models helped to build a video recommendation engine that improved the customer retention rate by 25%, leading to a significant revenue growth.
The dashboard framework helped encouraging collaborations to further drive customer engagement.