Text Analytics is a technique used to understand consumer’s opinions regarding products and services. The way people discuss their experience on public platforms, such as social media and blogs, provides insights for businesses on how to adapt their offerings. Feedback in the form of comments, discussions or even complaints can give insight into the areas that need improvement. An extension of Text Analytics is Sentiment-Analysis, that uses the text in context with the data to gauge the overall feelings regarding the topics being discussed.
Text Analytics works by initially extracting raw text from a data source such as Twitter, or customer reviews. The data is then filtered to contain only the text of the message itself, stripping out any meta data or link data. The remaining text data is processed into information, for example if the purpose is to find a unique variety of words, the repeated words are removed. Various analyses can be performed on the information to reach the required result.
Netflix is a popular video streaming service, offering a wide variety of content. and through text analytics we can see that there are some perceptions associated with the brand. These insights can be derived from looking at the publicly available sentiment relating to Netflix on Twitter through our custom–built MATLAB functionality.
The figure above captures the most frequently occurring words in South African tweets containing the keyword: ‘Netflix’. The conversations surrounding movies and documentaries are comparatively limited. Additionally, very little discourse surrounds anime. The predominant discourse is surrounding series and shows, however, further investigation exposed that Netflix is continuously requested to update their show’s seasons as well as introduce more of a variety of shows from other networks, such as The Office US. Netflix viewers are predominantly attracted to the series available on the platform; especially ‘local content’ as previously SABC exclusive local shows are available on Netflix. Keywords that kept showing up in the discourse surrounding Netflix were related to requests such as ‘please’ and ‘bring’.
Looking at the geolocation of the tweets, it is clear that the majority of the tweets come from Gauteng, Western Cape and Kwa-Zulu Natal – areas that contain prominent hubs and city areas. It is apparent that areas without fibre connection may have trouble accessing Netflix, with data being an expensive commodity in South Africa. Netflix promotes local content creators to display their shows on a worldwide platform. However, the data shows that Netflix still loses out on consumers solely using entertainment platforms to watch sports and news. DSTV, with its Showmax partnership, allows for existing DSTV users to have access to streaming content and retain satellite advantages for news broadcasts and sporting events.
The bar graph above gives insights into the content generally discussed when Netflix is mentioned in conjunction with Showmax, Netflix’s most significant streaming-competitor. The discourse has moved from shows and series in the first bar graph, Prominent Words, to movies. Despite ‘Showmax and Netflix’ being discussed in conjunction with each other as a unit, the primary focus is related to recommendations for potential viewing on either platform.
Based on public data available on Twitter and using text analytics tools available to us through MATLAB, we were able to find insights that can help understand the customer experience on Netflix in South Africa. This functionality shows the potential of what can be done using Advanced Business Analytics techniques and how you could gain further insights into your target market and their preferences.
What Can I Do Next?
- Request a trial.
- Visit our Advanced Business Analytics Competency Page.
- Find out more from the team.