Realize business value from big data with Infosys data analytics solutions.

« Navigating your next Customer experience | Main | COVID-19: How 'Live Enterprise' can help Organizations cope better »

Big Data for 360-degree view of the Customer

Posted by Ashish Suratkal (View Profile | View All Posts) | February 8, 2019 8:58 AM

Customer is the King! This saying is turning out to be quite true in the current times. Any Enterprise, be it a Retail giant, CPG cos, Top Bank or any Manufacturer is looking to please the Customer. In order to please the Customer, it is essential to understand the needs & priorities of the Customer. This can be achieved by analyzing the customer behavior at various touch points. An Enterprise needs to capture data about the customer at all its points of interactions like:

·         Customer visit to its Store / Outlet

·         Customer calling the Helpdesk / Contact Center

·         Customer posts, profiles on Social Media

·         Customer visit to its Website


 Traditionally enterprises have been capturing the data in case of Customer visit to its store by means of POS entries in their transactional database systems or Data warehouses. But this information is inadequate to understand the unstated needs of the Customer.

Contact Center Data:

                                The Contact Center represents that list mile in getting finer aspects of the Customer. Be it the choices a customer made in the Interactive Voice Response system or the discussion with the Contact Center Executive, all represent a crucial component of the Customer sentiment and their preferences. This information is in the form of voice which needs to be stored and analyzed for deriving insights. Here Big Data comes to our rescue. This information can be stored in HDFS (Hadoop Distributed File systems), usual Relational Databases cannot store this information since this data is unstructured / semi structured on many occasions. There are several ways to do Real time sentiment analytics on the Voice and provide the outcome to the Contact center executive so that he can adjust his interaction based on the Customer's sentiment and mood at that point of time. Some enterprises prefer to convert Voice to Text and then store it in HDFS. Contact centers operate in several local languages based on the Customer Geographies. Most of the mature Analytics models are in English so it becomes prudent to convert all multi lingual Text to English language Text. This Transcript is sent as an input to the HDFS. Most of the Enterprises prefer to have a Data lake or Data Hub which is a central warehouse of the Enterprise wide data.  This contains Transactional data, Historical data and data from External Sources. This Enterprise Data lake has its storage on HDFS (Hadoop Distributed File System). Data from the existing Data warehouses & other feeds can be ingested in the Data lake via Infosys Information Grid (IIG).  This a complete Data lake management solution from Infosys.


All Customer conversations with the Helpdesk are transcribed, translated and stored in the Data lake. This helps in numerous ways, in case of a Customer complaint the Sales rep who goes to this Customer knows about it already. He can assure the Customer that his complaint will be resolved and how the Company will ensure it will not recur. Contact Center conversations represent a major source of Customer information which if properly harnessed can provide -

·           View of Customer sentiment

·           Accurate information about issues faced by Customer

·           Likelihood of Customer switching to competition

·           Exact needs & requirements of the Customer

·           Likelihood of cross selling success to Customer

·           Suggestions from Customer about product & service improvement

·           Feedback about the Contact center interaction


Data from Social Media:

     To understand the implicit needs of the Customer it is crucial to understand the behavior, preferences, buying patterns of the Customer. There is a treasure trove of information available about the Customer from Social Media like Facebook, Twitter, Instagram, Linkedin, Youtube, Pininterest etc. Combining the Customer profile from Social Media along with his Buying patterns (from Transactional Systems) helps an enterprise to accurately predict buying patterns. This also enables Cross selling and prevents Customer churn. The crucial challenge about Social Media data is about Type of Data & Volumes. An Enterprise Data lake allows to store huge volumes of data at negligible cost and integration of this huge untapped social media data enables an enterprise to derive proper insights about Customer behavior & preferences.


Data from Website Clicks:

    A Customer may visit the website to look at the product offerings, understand the utility of the products, to evaluate competing products or to compare the prices and offers from the company and its competitors. An enterprise can explore cross selling opportunities with the customer by understanding & analyzing the clicks on its website. This data can be streamed to the Data lake by means of Real time streams and the information can be captured in near real time. Infosys Real time streams can help in capturing this data in the enterprise data lake.


Automation Enabler: Big Data provides an ability to store unstructured data at minimal cost which enables an Enterprise to automate several processes. We all remember filing up a large form to open an account in a Bank. Top Banks across the Globe are using Hadoop to store scanned copies of these forms and information extracted from them is used to complete the Customer profile in the Bank's systems.

Cost Effective: The Hadoop distributions come with a negligible cost to an Enterprise and do not need costly servers to host them.

Input to Analytics: An Enterprise Data lake provides crucial data needs to derive Insights about the future. It helps to build Analytical models which predict the future buying patterns and behavior of the customer. Infosys provides Analytics Workbench (AWB) to enable easier application of Analytics by using the data in the enterprise data lake. 



Thanks for the this good information.

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)

Please key in the two words you see in the box to validate your identity as an authentic user and reduce spam.