To address various business problems, there are four different types of analytics commonly practiced: descriptive, diagnostic, predictive, and prescriptive.
The breakdown of how businesses use these four types of analytics is summarized in the image below.
Continue reading below for more details on each type of analytics.
Descriptive Analytics
Descriptive analytics is one of the most commonly practiced analyses of data. It does not draw inferences or make predictions from its findings. Descriptive analytics uses current and historical data to identify trends and relationships.
Descriptive analytics includes common math and statistics tools, such as measures of center, averages, and percent changes. It also includes visualizations, such as line graphs, pie charts, and bar charts, that can be easily understood by a wide range of audiences. This type of analysis is most frequently used during day-to-day operations, such as company reports on workflow, revenue, and sales, or the inventory of goods and services, in order to provide snapshots of company operations. Snapshots are time and energy savers for the stakeholders who need to understand what is going on in the data.
Example: The Walt Disney Company releases financial statements to share information about its current business performance. Company investors use this information to determine whether or not to continue investing. This type of analytics is called descriptive analytics because the analysis is used to share past and present business performance with investors in an easily understandable way.
In a
financial summary about Walt Disney Co.
, you can find statements like “For the six months ending 30 March 2019, Walt Disney Co. revenues increased 1% to $30.23B.” This percent change is a business-performance metric that describes underlying data collected and is thus considered a type of descriptive analytics.
Diagnostic Analytics
Diagnostic analytics takes descriptive analytics one step further by diagnosing why something is happening in the data. This could be understanding why a trend arises or why a problem has occurred. For example, a marketing team may be curious why their ad campaign performs poorly with a certain demographic group. With diagnostic analytics, the analyst can help identify why that demographic group is less engaged with the advertisement. Part of diagnostic analysis involves separating the cause of an issue from other variables that could mislead the interpretation of the data. This might mean separating patterns, finding a specific reason for a pattern or trend, or investigating an underlying issue with an abnormality.
Example: Consulting firms, such as McKinsey & Company, help manufacturing companies by applying analytics to identify the main cost drivers in manufacturing processes. Clients use this information to reduce production costs. This type of analytics is called diagnostic analytics because the analysis is used to answer why the costs were higher in a given manufacturing process.
In a report on the declining labor share of income in the United States
, McKinsey makes statements such as “Rapidly rising commodity prices in the energy and minerals sector tend to increase profits (and investment) and reduce labor’s share of income.” This diagnosis is one of the factors contributing to the declining labor share of income. This is an example of diagnostic analytics.
Predictive Analytics
Predictive analytics uses past information along with the underlying background information of the data in order to predict what may occur in the future. In predictive analytics, the focus is on forecasting future outcomes that could happen and identifying their likelihood. With this information, businesses can organize and plan better, set realistic goals, and avoid unnecessary risks. In addition, it can address the “what ifs” that businesses have about certain changes they are considering.
Example: A business analyst may be in charge of creating accurate projections of next year’s spending for the finance team of a business. They could be responsible for predicting demand for certain goods depending on the region or time of year for the operations team. Banks, such as JPMorgan Chase & Co., predict the probability that a person or business will fail to repay a loan. Banks use this information to determine the interest rate for a loan.
In this article analysts at JPMorgan Chase & Co. explain what they think will happen with the markets in the short term: “Today’s news could lead to elevated market volatility in the coming week, but it has been widely expected — as bearish positions are unwound, some temporary disruption can be expected, but the turmoil will likely pass quickly.” This is an example of predictive analytics.
Prescriptive Analytics
Prescriptive analytics focuses on what should be done or what can be done to make something happen. This takes predictions one step further in uncovering truly meaningful business insights. It also means interpreting how to solve the problem at hand or realizing a valuable new opportunity for a company. Goal-oriented actions, like maximizing, minimizing, optimizing, and promoting, are often the result of prescriptive analytics.
Example: UPS uses analytics to create their delivery plans by optimizing the transportation of multiple customers’ packages. In doing so, they make sure packages are delivered on time and at the most affordable cost for the business. This type of analytics is called prescriptive analytics because analytics is used to minimize cost, which is a previously identified problem. Prescriptive analytics takes predictive analytics one step further by offering specific remedies for how to solve the issues based on the information discovered.
In this article from Harvard Business School, you can learn how UPS used a system called ORION to optimize its deliveries: “As of 2016, UPS estimates it has optimized 55,000 of its routes, achieving fuel savings of 10 million gallons, reducing CO2 emissions by 100,000 tons and saving the company between $300-400 million in costs. UPS calculates elimination of one mile, one driver per day could generate $50 million in savings per year.” This analysis resulted in prescriptive practices like preventive maintenance for the vehicles.
You now have knowledge and examples about each type of analytics for business analysts.
