Companies are familiar with the idea of descriptive analytics—using data about past decisions to improve business processes—and the promise of predictive analytics, which focuses on determining likely customer action and response based on currently-available information. In the middle is an emerging force: Prescriptive analytics, which a recent Information Age article calls “the next stage of analytics maturity in which you begin making decisions based not only on individual predictions,” but also using an aggregated view of predicted relationships.
The result? Recommendations about the best decision to make to maximize future opportunity. The challenge? As with any prescription: Finding the right dose.
According to CIO, predictive tools are the next step in analytics. But there’s already room to grow in this market, especially when it comes to speed. With data scientists spending three-quarters of their time prepping data sets and just one quarter running analysis, it’s easy to get bogged down as the sheer volume of information ramps up. And since the goal of predictive solutions is to give insight on both long-term and immediate business opportunities, there’s real need for better infrastructure and integration spending: Data must be collected, verified and analyzed in near real-time to have prescriptive value.
A February 10th article from Forbes, meanwhile, discusses the newest frontier for predictive analytics: Automation. In fact, industry expert Tom Davenport has suggested that automated analytics is the natural extension of a prescriptive model, potentially eliminating the need for human interaction. Here, finding the right dose means giving automated processes the freedom to handle basic analytics tasks, while reserving higher-level decision making for human experts. It’s a trial-and-error process to some extent, since companies won’t know where the ideal balance lies until they’ve tested a few data sets and implemented the results.
The final key to proper dosage? Choosing to build in-house or opt for expert help. According to Gartner analyst Lisa Kart, “when people are getting started in new areas, it makes a lot of sense to call in experts in that area.” She notes that the most important factor in success is working closely with decision-makers, but that trying to do everything in-house may be a daunting task.
Bottom line? Prescriptive analytics are the next logical step in leveraging data. By investing in speed, striking a balance between human oversight and automation along with in-house versus outsourced, it’s possible for organizations to tap a new kind of business intelligence.