Understanding the Differences Between Descriptive, Predictive, and Prescriptive Analytics

Descriptive, predictive, and prescriptive analytics are all advanced data analysis methods that help companies and organizations understand what is happening, why things happened, what might happen in the future, and what smart strategies they can implement next. Organizations collect an overwhelming amount of data, but harnessing it and using it to make smarter, faster business decisions can improve company performance. 

While these three forms of data analytics are not the only methods you can choose from, they are commonly used to help businesses reflect on their processes to make improvements, decide where to direct the company’s future, and create a more engaging story from the data they collect. Data professionals in industries like finance, retail, government, utilities, health care, human resources, and manufacturing use advanced analytics to better understand their organization and drive decision-making. 

Discover the differences between descriptive, predictive, and prescriptive analytics and how organizations use them to uncover insights needed to make data-driven decisions. 

Descriptive, predictive, and prescriptive analytics

Descriptive, predictive, and prescriptive analytics are all categories of advanced analytics that provide a different type of insight for companies and organizations. Compare how these types of analytics differ with a simple lemonade stand analogy:  

• Descriptive: This type of analytics describes what happened in the past and gives you data to use in other forms of analysis. For example, descriptive analytics can tell you that your lemonade stand sold 100 cups last week, 18 on Monday, 24 on Tuesday, 20 on Wednesday, 37 on Thursday, and only three on Friday. 

• Predictive: These calculations help you predict future events. Using data from the past several weeks of descriptive sales data, you might use predictive data to estimate that you won’t sell more than 10 cups of lemonade next Friday. 

• Prescriptive: This type of analytics builds off predictive analytics to help you understand the actions that you should take. If you don’t expect to sell very much lemonade on Friday, you may be able to save on the cost of your ingredients or—if you were a lemonade tycoon—the cost of staffing your store. 

A fourth common type of advanced analytics, diagnostic analytics, helps your organization understand why things happen. In the lemonade example, diagnostic analytics might reveal a smoothie shop open from Friday to Sunday a block away that draws away your customers. 

These scenarios offer a simplified view of the range of insight a company and organization can achieve using descriptive, predictive, or prescriptive analytics. Explore these advanced data analysis techniques in detail, including their benefits, potential drawbacks, and everyday use cases. 

Descriptive analytics

Descriptive analytics is the most common form of data analytics and describes past events. This data serves as a foundation for other kinds of data to build from because it establishes a baseline of factual information. 

What are some uses of descriptive analytics?

Descriptive analytics help you measure key performance indicators using your organization's vast and varied amounts of data. A few use cases for descriptive analytics include: 

• Revenue and expense reports

• Accounts receivable and payable

• Price-to-earning ratios

• Social media engagement

• Survey results

You can use your descriptive data to compare current and past performances to gain insights and notice trends. You can also use your descriptive data to conduct other kinds of analysis. 

Pros and cons of descriptive analytics

One of the biggest advantages of descriptive analytics is that you can use it to create engaging charts or visuals that help demonstrate key metrics to stakeholders, potentially to demonstrate growth or overall performance. Descriptive analytics offers an easy way to make complex data easier to digest. As the simplest and most common form of analytics, you don’t need to have data analytics skills to work with descriptive analytics. Once you have descriptive analytics, you can build on your analysis and apply predictive, prescriptive, and diagnostic analytics to understand how to reach your company’s goals on a deeper level. 

The main drawback to descriptive analytics is that it doesn’t do more to tell you why things happened or what you should do to course correct or take advantage of the opportunity. You need to apply your data to another form of analytics to answer those questions. 

Predictive analytics

Using your descriptive analytics, you can use predictive analytics to estimate what your organization can expect in the future on your current path. Diagnostic analytics, which offers insight into why your descriptive analytics occurred, are also important for predictive analytics. Knowing what happened and why helps you plan for the future. 

What are some uses of predictive analytics?

Predictive analytics can help organizations forecast what might happen in the future in a number of ways: 

• Forecasting cash flow: You can use predictive analytics to estimate your cash flow in the future. 

• Set staffing requirements: Predictive analytics can drive improved decision-making for the number of people to hire and schedule per shift. 

• Prevent machine malfunction: Predictive analytics can help you maintain machines before they reach a point of disrepair that they malfunction. 

• Patient care: Health care professionals use predictive analytics to predict patients who might be at risk of developing illness or need care in other ways. 

• Detecting fraud: Financial institutions, cybersecurity professionals, and other safety professionals can use predictive analytics to detect abnormal transactions and potentially fraudulent behavior. 

• Personalized marketing: Companies can use predictive analytics to understand users' online shopping behavior and target them with advertising at the points of the sales funnel when the customer is most likely to respond. 

Pros and cons of predictive analytics

Predictive analytics allows you to prepare for the future using data to inform your strategy. While predictive analytics can offer you a great deal of insight in a variety of ways, you should also be aware of the downsides. For example, predictive analytics are only as good as your data and require a lot of data to make accurate predictions. A second downside to predictive analytics is that it can only offer insight into situations that have happened in the past and doesn’t have the flexibility for random events or new information. This is where the next category of analytics, prescriptive analytics, shines. 

Prescriptive analytics

Prescriptive analytics build on the data compiled in descriptive, diagnostic, and predictive analytics, offering insight into what might happen in the future based on potential business strategies and outside factors. Prescriptive analytics can help you compare scenarios to determine what might happen if you take one business strategy over another or how a decision in one area could impact other areas. 

What is prescriptive analytics used for?

Prescriptive analytics can help you gain insight into the best possible outcomes in scenarios such as: 

• A/B testing: Prescriptive analytics can help you determine the most effective strategy by analyzing both and comparing the results. 

• Market research: Prescriptive analytics can help you compare what features or services your customers want to purchase or what offerings could bring in the most new customers. 

• Medical treatment: Health care professionals can use prescriptive analytics to estimate the best course of treatment while considering the patient’s unique health care history and symptoms. 

Pros and cons of prescriptive analytics

Prescriptive analytics is the most advanced form of analytics and can offer your company a great deal of insight, but you should also consider the drawbacks of the analysis. Prescriptive analytics can help you understand what might happen in the long term, but it’s still best used for short-term planning. Prescriptive analytics can consider many different factors in calculations, but those factors change over time, leaving your analytics outdated. To overcome this, you can revisit your prescriptive analytics often to update changes and reevaluate your company’s progress towards goals. 

Learn more about data analytics on Coursera. 

Descriptive, predictive, and prescriptive analytics are all important components of a data analysis strategy to help your organization make data-driven decisions. Consider an online course to learn more about data analysis. To begin, you could enroll in Introduction to Data Analytics offered by IBM. You could also gain hands-on skills through the Introduction to Data Analysis using Microsoft Excel Guided Project on Coursera. Or, prepare for a career as a data analyst with IBM’s Data Analyst Professional Certificate.