March 25, 2022 By: Sathyan Pari
A version of this Blog was first published on the Elets BFSI.
Organizations are always on the quest for information that will help them improve various aspects of their business, like sales, building better brand equity, customer opinion about their brand/product, people’s opinion about the competitors, etc.
A traditional enterprise reporting approach simply indicates what has happened, like, sales for a particular period, or were the weekly/monthly targets achieved, etc. To know why this has happened, organizations need to look for insights beyond enterprise information that will exist in the form of text in various platforms like social media, review websites, customer surveys and even feedbacks/surveys that are hand filled. The digital platforms have bridged the gap between the business and consumers.
Digitization has changed the way people share their opinions about products and brands. Approximately 350K tweets are posted per minute, Facebook accounts for 71% of all social media visits in the US and there are 330M monthly active users on Reddit. People tend to share their opinions, suggestions, or reviews of the product on all these platforms. This huge volume of chatter can be positive/negative, constructive feedback for improvement, or a complaint about a specific product/service. Let’s dive into how organizations can leverage this information to gain meaningful insights that can serve as business drivers.
Text Analytics Beyond Sentiment Analysis
When we talk about deriving insights from publicly available information, then sentiment analysis is the first aspect that comes to our mind. This technique to calculate the sentiment of a particular brand/product is widely popular and many organizations are adopting this to understand the sentiment. However, there are a lot of other insights that can be derived out of the open information. The quest to determine the sentiment has opened doors for analyzing the available information in multiple different ways.
- Organizations can now benchmark their Net promoter score(NPS) by comparing it against their competitors.
- Market research can be done by collecting the chatter around a particular product or service.
- Organizations can unearth some genuine feedback/suggestions from the massive chatter, and it can be redirected to their respective business teams. For example, a tweet for a hotel review can be – “The rooms in the hotel are very comfortable and clean, but the room service is very poor.”
The topics discussed in the above-mentioned example are room quality and room service. Likewise, if we can classify the entire chatter on a particular product/brand then it paves the way for the relevant team to get the information about their business area. This technique of identifying the topics discussed is widely known as Topic modeling in Natural Language Processing (NLP). The Topic modeling ensures the business groups focus on their areas rather than going through the entire information.
- Organizations, that deal with customer complaints and suggestions via social media/chatbots can employ a smart way of picking the keywords used in the discussion to redirect the query to the relevant team in an automated fashion. This technique of picking the keywords is known as “Named Entity extraction”.
These are some of the examples where organizations can leverage the power of Text Analytics to gain insights.
Importance of Accommodating Text Analytics Within Enterprise Reporting
The basic utilization of analytical reports from enterprise information alone needs to be changed. The insights from openly available information must be integrated into enterprise reporting. This will help the organizations remain on top of the things that are happening with their customer front. However, only text analytics shouldn’t be employed, when there is a need to know about the sentiment during a special event. In the digital world, online platforms are the best available medium to hear out the customers which if utilized in a better way will lead to better customer experience and satisfaction.
