Smart Categorization For Smart Businesses

Manas Agrawal, CEO, AffineBusinesses are increasingly relying on analytics for decision-making. This truly is the way forward, if only the decision makers are provided the right analytical support. The situation is further complicated with the fast - evolving environment. India has currently over a billion mobile subscriptions and more than 30 percent smartphone users. The digital penetration across much of the developed and developing world is similar or even higher. There is a paradigm shift in the way products & services are being consumed. Businesses are seeing a need to evolve faster and at times completely reinvent them to keep pace with the changing world. The customers of today are moving to text based systems such as social media, chats and blogs along with continued usage of email to voice their opinion and reach out to companies. To help businesses effectively come to terms with this change from the predominantly voice communication era, there is a need for smarter customer management system.

A combination of text analytics with human intelligence and machine learning algorithms is needed to create business solutions centered on interpreting unstructured text data. There are machine learning modules for classifying data, understanding the inherent topic of conversation along with the context and sentiment. These modules have multiple applications. The two very specific business use cases being:

• Voice of Customer (VOC) helps in understanding what makes your customers happy, discover how they perceive your product and what motivates them to stick to the brand. It also helps in building a competitive analysis by analyzing what your customers mention and say about the brand. This will help the brand to create loyal and a strong base.

• Smart Customer Communication Management: Managing customer communication via text based channels (emails, SMS, chat, social media) is labor intensive and inefficient. AI based automated processing, can significantly reduce response times while increasing efficiencies and ROI. Recently launched, ATOM (Automatic Text Organizing Machine) has already received a positive response and is engaged in talks with few clients for a pilot project.
An Interesting Case of a Telecom Giant
Since the launch of Reliance Jio, there have been countless reports and articles on the web that aim to provide detailed insights of the financial impact on the incumbent telecom players but there aren't many articles that capture the change or shift in consumer discussions. In this digital era, a lot of these discussions are happening on social media sites, especially Twitter, where consumers are freely voicing their opinions and/or expressing their concerns.

There are machine learning modules for classifying data, understanding the inherent topic of conversation along with the context and sentiment

Using text categorization and sentiment analysis, we picked and analyzed thousands of tweets targeted towards major telecom players to see how customers rated their experience since the launch of Jio and what implications that could have going forward. Amongst many others, our initial analysis on customer activity revealed that there was a considerable increase in overall activity on Twitter, with telecom targeted tweets increasing by almost two folds during the Dec'16 quarter alone. Below are some of the more key findings.

• After the launch of Jio, the number of total tweets increased by more than 83 percent in the Dec'16 quarter.
• Customer activity for Airtel and Vodafone increased by 90 percent and 66 percent respectively since the Jio launch.
• Tweets targeted towards Jio have increased by 48 percent since the announcement of Prime Membership.

The analysis also indicated that while Jio was able to garner considerable consumer interest on Twitter during its free period, it has been able to sustain that momentum after announcing paid services.

There are machine learning modules for classifying data, understanding the inherent topic of conversation along with the context and sentiment

For a large, extremely competitive and all pervasive industry such as the Indian Telecom Industry, capturing and analyzing the Voice of Customer (VOC) is an extremely important input for the management to devise new operational strategies especially if the industry is facing a tactical shift after the entry of a new player. We used the AI powered platform to extract the contextual information from the tweets in order to identify the topics of conversations. The data showed that after Jio, there was a considerable shift in consumer expectations from telecom service providers. An increased number of users were tweeting about price and promotions offered as against connectivity.

Summing up, our findings from the Twitter analysis suggest that Jio was able to bring a renewed customer activity on social media platforms by offering audacious free services followed by cheap data and free voice calls for the life. It was also able to considerably shift the rhetoric of consumer discussion away from Connectivity to Price, Promotions and Apps. Not so surprisingly though, large incumbent players like Airtel and Vodafone faired differently on the overall consumer sentiments barometer which can be attributed to the different level of aggression displayed by each of these brands in countering Jio's offers.