Big data is old news by now. What is new, however, are the ways in which companies are putting their data into use to improve decision making and business strategy. Prescriptive analytics has emerged as a model for taking data to the next level, offering more than just a model of what might happen in the future.
The Evolution Of Prescriptive Analytics
Over time, big data analytics have evolved significantly. Let’s examine how we got from data collection to data action:
This type of analytic answers the question, “What has happened in the past?” The answers help establish relationships between customers and the products and services they purchase, with the goal of improving future outcomes. Historic sales, finance, and operations data are descriptive. Netflix uses descriptive analytics to study a user’s viewing history to provide them with movie recommendations they are likely to enjoy.
By looking at historical data and identifying patterns, analysts are able to go a step beyond descriptive analytics and determine the likelihood of a future outcome. Predictive analytics answers the question, “What is going to happen next?” Loan-application scoring is a simple example of predictive analytics in action. By analyzing an applicant’s past behavior, a bank can determine the likelihood that a borrower will repay a debt on time.
Building upon both descriptive and predictive models, prescriptive analytics attempt to determine the effect that future decisions will have on the business. Based upon those predictions, analysts can then model outcomes. A retailer may use descriptive analytics to identify customer segments. Predictive analytics allows that same retailer to identify which of those segments are likely to attrite within a three month time frame. Prescriptive analytics will help the retailer craft the precise deals and offers that will entice those customers to take action. If that customer subset is not particularly tech-savvy, prescriptive analytics might identify direct mail as the most effective medium for a personalized offer.
How Prescriptive Analytics Improve Business Outcomes
Tracing its roots back to 2003, prescriptive analytics is still considered a relatively new field. Companies that have adopted this type of modeling use it to optimize pricing, production, inventory, supply chain, sales leads, and even hiring. Prescriptive analytics looks at historical patterns, predicts future outcomes and then prescribes a “treatment” for those future outcomes. Let’s say a company determines that a subset of its customer base is five times more likely to attrite than all other segments. Prescriptive analytics says, “Here is what you can do to reach those customers, based upon their unique data sets.”
So what does prescriptive analytics look like in action? Best Buy famously used prescriptive analytics to determine that 43 percent of its total sales were coming from less than ten percent of its customer base. Using that data, they segmented their customers into archetypes. Based upon those data-driven character models Best Buy redesigned its stores to reflect the spending habits and preferences of their best and most profitable customers. However, due to unforeseen economical hardship, their efforts were slightly deflated.
Prescriptive analytics can be used to drive competitive deals and offerings, strengthening customer loyalty and improving the bottom line. By moving beyond simple historical data and future predictions, prescriptive analytics can guide strategy decisions that will have a positive impact on customer retention, profitability and long-term sustainability.