3 Levels of Analytics: Descriptive, Predictive and Prescriptive
Analytics––descriptive, predictive and prescriptive is a rapidly evolving field that gives companies the knowledge to make smarter business decisions. It provides valuable insight into past performance and future prediction and decision guidance. According to Gartner, the number of companies using prescriptive analytics tools may increase from ~10% in 2016 to over 35% by the end of 2020. Gurobi Optimizer is one of these tools: a math programming solver. While prescriptive analytics is growing, analytics overall is still dominated by descriptive (what happened in the past) and predictive (what is likely to happen in the future) tools. As a result, it’s important to see how prescriptive analytics fits into the broader analytics landscape.
- Answers the Question: What happened and why?
- Primary Tools: Data aggregation and data mining
- Limitation: A snapshot of the past may have limited ability to guide future decisions
- Best Use: Summarize results for all or part of your business
Descriptive Analytics gives insight into the past and current state of your business through the use of business intelligence tools. These tools can help you obtain a range of insights into your business, such as:
- How much of a given product you sold over a certain time-period
- Your current product inventory levels distribution centers
Most business functions in your company are already using descriptive analytics in the form of recurring or custom reports. With analytics tools, you can dive deeper into the past and retrieve meaningful results. You can also conduct Diagnostic Analytics which focuses on understanding why something happened, as part of your descriptive analytics activities.
- Answers the Question: What might happen?
- Primary Tools: Machine learning, statistical models, and simulation
- Limitations: Estimation of the future
- Best use: Backing of an educated guess for the results of low complexity decisions
Predictive Analytics seeks to provide insight into what the future may hold for your business. It takes existing data and applies statistical techniques often using machine learning. Results may be coarse-grained (e.g., expected industry growth or raw material pricing), company-centric (e.g., revenue or profit growth), or operational (e.g., expected changes in demand by product line).
- Answers the Question: What should we do?
- Primary Tools: Optimization and heuristics
- Limitations: Most effective where you have some control over what is being modeled
- Best use: When you need to make important, interdependent, complex or time-sensitive decisions
Prescriptive Analytics applies computational sciences, typically through math programming models, to optimize a set of decisions for directing a given business situation. Utilizing this tool in support of achieving desired business outcomes will bring specific direction and high value to your business. By building a decision model and using a math programming solver, you can:
- Explore an astronomical number of possible combinations and options and find the proven best option.
- Apply a range of option constraints to maximize or minimize your objectives
- Reduce decision-making risk
- Free up time for higher-value efforts such as performing scenario analysis or considering larger strategic questions
For example, Prescriptive Analytics could answer these business decision questions:
- In which order should you produce what products?
- In which manufacturing facilities?
- On what product lines?
- In what quantities?
Subject to a range of constraints:
- Minimum production of a given product
- Required manufacturing time and cost of a particular machine
- Raw material inventory
- Finished goods inventory capacity
To maximize or minimize your objectives
The result is improved decisions, increased profitability and saved time for your business.
Math Programming Solvers: The Primary Tool in Prescriptive Analytics
Used as the primary tool in prescriptive analytics, mathematical programming solvers, such as the Gurobi Optimizer, help transform data and models into smarter business decisions.
How it works
Leveraging a broad range of programming languages, users can:
- State their toughest business problems as mathematical models, then call a solver to automatically consider trillions or more possibilities and find the best one.
- Use the Gurobi Optimizer solver as a decision-making assistant, helping guide the choices of a skilled expert, or as a fully automated tool, making decisions without human intervention.
- Solvers rapidly consider large numbers of business constraints and decision variables within minutes, far exceeding the choices a human brain could consider over the course of many years.
- Solvers support companies’ needs to refine the way they currently make decisions and enable them to efficiently and effectively take a wider array of factors and options into consideration than ever before. The end result – decisions that drive better business results.