Predictive analytics is the use of data, statistical algorithms, and machine learning to predict the likelihood of upcoming/future predictions based on historical data.
To get benefited in business, these algorithms uses past transactional data to identify risks and opportunities in the market. Predictive Analysis is used in various fields, including marketing, telecommunications, retail, travel, healthcare, pharmaceuticals, insurance and others.
As we are increasing our computing power through predictive analysis as well as both the quantity and quality of the data we access, it is to be expected that our ability to predict future outcomes will improve. Predictive analysis is becoming more acceptable due to factors such as the growing volumes and availability of data, less expensive and less complicated computers and software, and growing competition amid tougher economic conditions.
What Predictive Analytics Can Do?

• Identify Trends
• Understand customer behavior
• Help to improve business performance
• Helpful to make strategic decision making

As predictive analysis becomes easier for more companies and organizations to access, the possibilities will be higher for extensive use of predictive analysis.For example Hong Kong is using predictive analytics on large quantity of unstructured data to predict and address issues related to public complaints before they happen. Utilities are using predictive analytics to determine the most profitable times to sell excess electricity.
Healthcare providers are using predictive analytics in a number of interesting ways, including predicting the effectiveness of new test procedures and medications; improving outcomes by improving patient care; and saving money by identifying patients who are not likely to adhere to their prescribed treatments once they go home.
As vast quantities of high quality data become more and more accessible to companies of all sizes, it will become more common for companies to calculate a lifetime value of their customers, and to predict which products and offers will be most appealing to which customers moving forward.
Limiting Factors
Of course, predictive analytics is not all knowing. The approach is limited by certain factors, including the following:

• Lack of good data : If the data quality is poor or limited availability of quantity of historical data are the biggest barrier to companies that seek to adopt predictive analysis. Even the smartest algorithms can’t be useful for poor and small quantity data.
• Poor statistical work : Statistics can be notoriously fuzzy when considering which variables are correlated to an action. If your math or correlations are fuzzy at all, your predictions will be less accurate.
• Big assumptions The biggest assumption in predictive analytics is that the future will continue to be like the past. But if you have been alive for more than a handful of years, you know that’s not true. Companies that don’t continuously question and monitor their assumptions become very poor.

As a manager (and not an analyst), it’s a good idea to talk to your analyst about these limiting factors to understand how much you can rely on your predictive analytics. What are the sources of your data and are they representative of the overall target group of customers? What kinds of outliers are you dealing with when creating the statistical models? And what assumptions are the predictions based on?

Understanding these basics, you should be able to rely on your predictive analysis.