What is Prescriptive Analytics?

Prescriptive analytics is about what we should do. 

In other words, how can we take the set of constraints that we have, the drivers of performance we’ve identified, and the goal that we have, to produce a single view about the range of outcomes we can have. Prescriptive analytics is the toolset and thought process required to move from a single prediction or question to reviewing a range of possible outcomes for the best one.


  • A retail analyst can holistically evaluate shelf space optimization, and think about how increasing or decreasing how every product on the shelf features. The analyst might use a scenario evaluator, we can take a model for sales performance, and try a variety of inputs and gather all outputs. 
  • A banking analyst might consider how new savings account customer acquisition varies according to both the interest rates they give out, and the interest rates their competitors might set. Determining this model allows them to chart a plan forward in a changing competitive environment. 

Building an Action Plan for Analysis

Descriptive, predictive, and prescriptive analyses frequently occur in sequence, but this does not imply one step is more important than the rest. Instead, the stepping describes the order that analysts typically move through an analysis. 

Without robust and well-understood descriptive analytics, organizations might not identify the right areas to focus on, and have perfect predictive models that constantly waste time changing what isn’t actually broken. But without meaningful and well-understood predictive drivers, “prescriptive” tools start looking more like inscrutable black boxes. 

Each step might provide a specific and necessary contribution to the overall goal of changed decision making. In our example above, an analyst employed all three on the way to delivering a recommendation about how to optimize the shelf: 

  • Descriptive Analytics: Identified the opportunity, and created the prompt to solve. 
  • Predictive Analytics: Identified the set of key drivers for sales performance. 
  • Prescriptive Analytics: Reduced a list of “what-if’s” into a single easy to interpretable output. 

Einblick’s Approach

We help you take action by combining descriptive, predictive, and descriptive analytics on a single collaborative data science canvas. Explore data in an intuitive way, build models with the AutoML engine, and then explore what-if scenarios all in one place. Traditional platforms which separate the three steps of analysis create friction, while Einblick to helps you understand the importance and impact of different variables. Our fully collaborative platform helps you to rapidly align the entire organization to new objectives.