Types of Data analytics


There are 4 types of analytics: • Descriptive analytics  • Diagnostic analytics • Predictive analytics  • Prescriptive analytics The more complex an analysis  is, the more value it brings. Descriptive analytics  Descriptive analytics answers the question  of what happened. For instance, a healthcare   provider will learn how many patients were  hospitalized last month; a retailer - the   average weekly sales volume; a manufacturer – a  rate of the products returned for a past month,   etc. Descriptive analytics juggles raw data  from multiple data sources to give valuable   insights into the past. However, these findings  simply signal that something is wrong or right,   without explaining why. For this reason, highly  data-driven companies do not content themselves   with descriptive analytics only, and prefer  combining it with other types of data analytics. Diagnostic analytics At this stage,   historical data can be measured against other data  to answer the question of why something happened.   Thanks to diagnostic analytics,  there is a possibility to drill down,   to find out dependencies and to identify  patterns. Companies go for diagnostic analytics,   as it gives in-depth insights into  a particular problem. For example,   in the healthcare industry, analysis of customer  segmentation coupled with several filters applied   such as diagnoses and prescribed medications,  allowed measuring the risk of hospitalization. Predictive analytics  Predictive analytics tells  what is likely to happen.   It uses the findings of descriptive and diagnostic  analytics to detect tendencies, clusters   and exceptions, and to predict future trends,  which makes it a valuable tool for forecasting.   Despite numerous advantages that predictive  analytics brings, it is essential to understand   that forecasting is just an estimate, the accuracy  of which highly depends on data quality and   stability of the situation, so it requires a  careful treatment and continuous optimization.

Thanks to predictive analytics and the proactive  approach it enables, a telecom company,   for instance, can identify the subscribers  who are most likely to reduce their spend,   and trigger targeted marketing activities to  remediate; a management team can weigh the   risks of investing in their company’s expansion  based on cash flow analysis and forecasting. Prescriptive analytics The purpose of prescriptive analytics   is to literally prescribe what action to take to  eliminate a future problem or take full advantage   of a promising trend. This state-of-the-art  type of data analytics requires not only   historical data, but also external information  due to the nature of statistical algorithms.   Besides, prescriptive analytics uses sophisticated  tools and technologies, like machine learning,   business rules and algorithms, which makes  it sophisticated to implement and manage.   That is why, before deciding to  adopt prescriptive analytics,   a company should compare required  efforts vs. an expected added value. What types of data analytics do companies choose? For the Global Data and Analytics Survey: Big  Decisions, PwC asked more than 2,000 executives   to choose a category that describes their  company’s decision-making process best. Further,   C-suite was questioned with what type of analytics  they rely on most. The results were the following:   descriptive analytics dominates in the  “Rarely data-driven decision-making” category;   diagnostic analytics tops the list   in the “Somewhat data-driven” category,   while it is closely followed by descriptive  and prescriptive analytics;   predictive analytics leads in  the “Highly data-driven” category. This survey proves that at different  stages of a company’s development,   there appears a need for one  or the other type of analytics.   In fact, the companies that strive for  informed decision-making, find descriptive   analytics insufficient, and add up diagnostics  analytics or even go as far as predictive one.

The same survey reveals another trend.   Executives want decision-making to be faster  and more sophisticated. This means that   more businesses will strive to gradually  enlarge the share of predictive analytics.   Another survey of business intelligence trends  for 2017 carried by BARC proves this hypothesis:   2,800 executives confirmed the growing importance  of predictive analytics and data mining. In summary, with various types of analytics,  companies are free to choose how deep they need   to dive in data analysis to satisfy their business  needs best. While descriptive and diagnostic   analytics offers a reactive approach, predictive  and prescriptive analytics makes users proactive.   Meanwhile, current trends show that  more and more companies come to   the situation when they need advanced  data analysis, and choose to adopt it.