What is Prescriptive Analytics? - Data Science Wednesday

Summary

Prescriptive analytics is the final step in a broader analytics process. It turns data and forecasts into recommended actions that help guide business decisions.

The speaker distinguishes four stages:

• Descriptive analytics: Shows what is happening in the business, such as how products are selling.
• Diagnostic analytics: Explores why something is happening, such as what is causing revenue to rise or fall.
• Predictive analytics: Uses a model to estimate what is likely to happen in the future.
• Prescriptive analytics: Combines predictions with business information to determine what action to take.

Grocery-Store Example

A store uses product sales patterns and inventory data to plan stock:

• A predictive model forecasts that a blue product’s sales will rise in summer, while a red product’s sales will fall.
• Inventory data shows the blue product is running low and the red product is plentiful.
• For a specific cereal, week, and store, the forecast is 56 units sold, but only 40 are on hand.
• Because the business wants to avoid running out of stock, the prescriptive recommendation is to ship 18 more units, providing a buffer.
• When prescriptive recommendations are integrated into business processes, actions like these can happen seamlessly, freeing staff to focus on other decisions.

Core lesson: Prescriptive analytics does more than describe, explain, or forecast. It uses available data to recommend practical actions.

Speaker and Source

Tessa Jones is a data scientist at Decisive Data and the presenter of Data Science Wednesday.