Posted in Operations & IT Articles, Total Reads: 1607
, Published on 05 April 2012
Business Analytics is the process of exploration of an organization’s data with the help of statistical analysis. Analytics convert a huge amount of data into highly valuable asset of an organization. Today analytics is not just a support function within a company; it has become a key framework to understand significant pattern which help them in complex decision making and taking insightful action for future growth. The volumes of data generated by the companies are increasing exponentially. The volume itself has become a major problem. Executives worldwide are facing the same problem of not getting the proper data when needed. The data that can make a difference are buried somewhere.
Moreover business struggle to gather real time information even though it’s a must as the companies have shorter time to react to changes. The availability of quality analytics help the organizations to react in real time. Hence we see that with the increase in complexity of the business, data analysis should be done on a regular basis to improve business performance. With the increasing need of incorporating analytics into the business, the analytics techniques have become more sophisticated. To manage the huge volume of data, usage of advanced statistical model is necessary to detect relationships and trends. Today companies also use predictive analytics extensively. The term refers to the process of successfully predicting future events using statistics and data mining. It enables corporations to identify opportunities and risk of a business based on historical data.
Analytics is not too technical to master. It is basically a group of approaches targeted at the business users rather than IT people. These approaches are used in combination of several tools to gather information, analyze them and finally present the findings.
The field of business analytics has shown a great improvement over the past few years. Business users are now able to get better insights from data stored in transactional system. An example could be e-Commerce application which has made data mining a very effective analytical process. Analytics application and business intelligence are now better integrated with transactional system than before. A close loop between operation and analysis has been created. Mined information is now available to be used by a large set of audience which helps to take the advantage of analytics in everyday activity.
Analytics can be used in the following activities:
Credit and Market risk in banks
Fraud detection and financial crimes in banks
Text fraud in public sector
Demand forecasting in manufacturing
Optimization of various activities in Retail sector
Core Performance Areas
Core performance areas of Analytics are:
Economic: - Use of analytics can optimize revenue by predicting future trends. Non value added services or products can be removed. This is especially applicable in challenging economic conditions.
Strategic: - In future there may be changes in business model, customer requirements, regulation requirements, external environment etc. With the application of analytics companies can analyze their position with respect to these changes and can take corrective actions if needed.
Operational: - Companies can understand whether they are using proper operational business model for customers, vendors, regulators and vendors. This will help to optimize the operational efficiency.
Organizational: - Analytics can also help to achieve best management practices. Companies can analyze employee satisfaction level, usefulness of their compensation and reward system, the corporate social responsibility the company is undertaking. This will bring more sustainability to the organization.
BI Vs BA
The focus of Business Intelligence (BI) is improvement of data capture and information delivery. It mainly concentrates on the distribution of known facts. For example BI can say how many transactions took place through credit card on a given day. But it cannot say how many of those transactions are likely to be fraudulent. BA, on the other hand can provide answer to such questions. It provides insight into almost all aspects of a corporation’s value chain by driving innovation. Thus with the help of BA an organization will be able to respond to the market more quickly and hence it can insulate itself from environmental changes.
There are basically two types of data: Structured data and unstructured data. Structured data tells us what customer did. They exist in database. BI is useful for analyzing structured data. Unstructured refers to the data that doesn’t exist in database. Unstructured data have significant business value as it can tell us the reason behind a specific customer action. Today most of the business is carried out on unstructured data. But businesses are unable to derive the right answer because of inadequate technologies. BA can analyze the unstructured data and transform it into actionable intelligence.
Data are becoming more and more complex. There are various types of data – Behavioural, Textual, Geospatial, Graphical, Social etc.
The analytics model is base on 4 steps:
Structured and unstructured data management throughout the lifecycle of the business
Transformation of the data into a consistent source of information
Use of simple and advanced technique to gain extra insight into the behavior of the data
Making the result of the analysis available to the process
The business analytics revolves around statistical methods and for this reason it starts with information logic designed for statistical data processing. The complexity lies in the representation of the objects to be considered for data modelling. This is mainly because of the huge volume of data available today. For this reason there is a gap between desired information and the outcome we’re trying to see. The following issues need to be taken care of in order to reduce this gap:
The overall cycle time of collection of data, analysis of data and the presentation of the finding must be reduced.
Within this cycle, the time required to analyze data should be reduced.
Business goals should be clearly defined. Unclear goals and metrics can misguide the effort.
Quantitative findings should properly be translated into language so that it can be distributed to a broader range of business users.
Older technologies find it difficult to keep pace with analytic advancements. Data warehousing technologies are very slow. Sometime it takes days to getting data. Cost of deployment and maintenance is also very high. Data appliances resolve these problems to some extent but it still is unable to support advanced analytics. Data appliance is useful as long as data is flat and the workload is small. Analytic DB also can’t run complex workloads and advanced analytic functions well. Hence a platform environment is the need of the hour. A proper Analytic platform should be able to integrate multiple data types, run complex workloads and advanced analytic functions. The platform should also be able to empower users to perform various functionalities. In short, the role of Analytic platform should be as follows:
It should take less time to display the result
Increase the accuracy of the result
Reduce the operational cost to give a better ROI
Offer new levels of innovation
Future of Analytics
In the upcoming years it is expected to have more business investments in the field of analytics. There are so many things to be done about Analytics. By now it’s clear how an ideal analytic platform should be. But all the objectives are not completed yet. We can expect more open source platforms for advanced analytics and implementation of predictive analytics into core products. At the same time it might become cheaper and as the vendor offerings will mature. Most importantly analytics will diversify the career options as it will become a distinctive field in its own right. Hence, Analytics is surely going to be a hot cake of this decade.
This article has been authored by Sayak Gupta from SIMSR.