Prescriptive Analytics: New Concept or Buzzword?

I often heard about Prescriptive Analytics recently, as the new type of analytics (in addition to Descriptive and Predictive Analytics). According to websites and whitepapers, Prescriptive Analytics is the next step after descriptive/predictive analytics. Once you get predictions, you need to prescribe actions to leverage these predictions.


I’m now wondering if this topic is really new or if it’s just a new name for something we have been doing for years. First, I wonder since people have been working on Decision Management Systems and Decision Support for years now (for example experts such as James Taylor). I remember discussing about the step “from analytics to action” years ago. Is Prescriptive Analytics the new term to name the step from analytics to knowledge in the data to knowledge pyramid?

If it’s the case, then it’s fine, but it’s not new. Who would make a prediction without any prescriptions in a data mining project (at least in industry)? In online advertising, for example, you score each visitor for each ad model. At the end, you prescribe an action (which ad to show). Without this prescription, you wouldn’t make any money at the end of the day.

In my view, most analytics projects already have a “Prescriptive Analytics” phase in order for the project to be valuable for the company. What do you think about Prescriptive Analytics? Buzzword or real new emerging concept? Feel free to share your opinion.

Two articles about Prescriptive Analytics:

Big Data Analytics: Descriptive Vs. Predictive Vs. Prescriptive
The Analytics Journey


Recommended Reading

Comments Icon3 comments found on “Prescriptive Analytics: New Concept or Buzzword?

  1. I think it’s nothing really new but at least it may help create a buzz to further push analytics into the right direction: it must be at the core of intelligent action in companies and organizations at large! We used to call it “actionable intelligence” a few years ago; it’s the same idea although I think this term speaks much more for itself as “prescriptive analytics” does.
    And just picking up your phrase above: “Once you get predictions, you need to prescribe actions to leverage these predictions.” You should not even be working on predictions without knowing what to do with them later. This is how Data Mining works in my sense: it always has a business mission to accomplish, it’s not a pure R&D function. The prescription comes embedded….

Comments are closed.