Prescriptive Analytics

Prescriptive analytics recommends what action to take. It builds on descriptive analytics, which shows what happened, and predictive analytics, which estimates what is likely to happen. It compares possible decisions against a goal, such as the lowest cost or the highest margin, and against limits, such as budget, capacity or delivery dates.

Methods include optimisation, which searches for the best combination of choices, simulation, which tests decisions against many possible futures, and business rules. It is the least common form of analytics, because it needs good predictions and a clear statement of goals and limits.

For example, a manufacturer decides how much of each product to make at each plant next month, given forecast demand, machine capacity and shipping costs.