Information mining in Big Data is the upcoming field that can form a unique selling proposition for a company. Data is the most valuable asset of business that can increase the credibility of the business much beyond the competitors reach. But few organizations have named big data as ‘Dark Data’, a journey inside a black hole which attracts one from outside but at the end takes the business nowhere. According to a study by an open source research firm, the companies expected a return of $3-$4 per dollar invested, but big data has helped them to get only 55 cents per dollar as their return on investment (ROI).
Big data as the name suggests is a collection of huge ‘Volume’ of data with different ‘Varieties’. It is the task of the experts to create a high dollar ‘Value’ of big data. Some organizations have underutilised big data while the others have taken irrelevant paths to access it. The reason behind such shortcomings is the lack of knowledge about the data locator.
The process of locating the business relevant data should be the initial step while implementing Big Data, where it is important to select the desired touch points and discount the rest. These touch points can also be further drilled down to obtain sub points. A concrete analysis of relevant data, out of the ocean of data, can help the organizations improve their returns due to the reduced number of touch points. Moreover, the requirement of reduced infrastructure will further reduce the initial investment cost. The decrease in the touch points will also reduce the complexity of data and will keep the analysts motivated towards the project findings.
To understand further, an FMCG company is considered which wanted to launch a new product. The company should trim down the list of all the available touch points and analyse the relevant five to six touch points that are required to carry forward the market research for new product development (NPD). Moreover, these touch points can be ranked according to the project analysis. This process will reduce the complexity associated with the increasing number of touch points.
Where to look for and where not to?
The process can be divided into 3 phases, viz., ‘Planning Phase’, ‘After the Promotion stage’ and ‘After Launch stage’. In the ‘Planning Phase’ of the product, websites of direct competitors, websites of Indirect competitors and their Facebook and Twitter pages, can be some of the high priority and relevant touch points. In the ‘After the Promotion’ stage, data can be gathered from the media resources, one of the unstructured forms of big data, to analyse the sentiment of the target customer towards the product. Finally, in the ‘After Launch’ stage, several websites like ida.org, fdi.org as well as the product’s social media pages can be referred for their critic comments and product review from different users/ non- users respectively. However, certain touch points like blogs and articles by writers can be avoided due to reduced chances of valuable data.
An equivalent insight can be construed by the use of 10-15 touch points instead of approaching the mountains of data. Hence, the dilemma faced by the organizations today, regarding huge investment and lack of ROI in Big Data, can be solved by finding relevant data locations, which can increase the efficiency of data extraction and decrease the cost and time requirement considerably.
Swati Gupta is currently pursuing PGDM-Finance in Institute of Management Technology, Ghaziabad. She has worked in Cognizant Technology Solutions for a UK based Retail company after pursuing B. Tech from University of West Bengal. She has also interned in Bank of Baroda, Dubai in the field of Risk Management. She has been working on several projects related to Portfolio Management and Risk Analysis. Her primary area of interest lies in Financial Risk Analytics.
Gagandeep Singh is currently pursuing PGDM-Marketing from Institute of Management Technology, Ghaziabad. Prior to MBA he has 2 years of work experience in Accenture where he dealt with Microsoft Dynamics AX for Michelin client. He holds a B.Tech degree from Maharaja Agrasen Institute of Technology, GGSIPU, New Delhi. His keen interests lie in analysing social media to construe business acumen. He has been working on various problems related to Text Analytics.
Try deep learning using MATLAB