Prescriptive method of analytics boasts broad benefits

Prescriptive analytics has broad potential, but according to SearchCIO's #CIOChat participants, CIOs should proceed with caution when it comes to pursuing this particular branch of business analytics (BA) lest the application -- and thus, the projected outcomes -- be less wide-ranging than they'd hoped.

The prescriptive method of analytics is closely related to both descriptive and predictive analytics; however, while descriptive analytics provides insight into past events and predictive analytics forecasts future events, prescriptive analytics seeks the best course of action to obtain a particular result.

Opinions about the applicability of prescriptive analytics vary. In a recent SearchCIO Trailblazer profile, Erick Brethenoux, director of BA and decision management strategy at IBM, explained that certain types of prescriptive analytics can be effective for any company looking to improve its business processes. "Even with a very small amount of data, you can do an amazing amount of analytics to save money right away," he said.

In our #CIOChat, SearchCIO Managing Editor Rachel Lebeaux was first to weigh in on whether prescriptive analytics applies to all industries or just those with closed systems and limited parameters:

@RachelatTT @searchCIO The caution is to take care with prescription. It can be overly restrictive. #CIOchat

— Tim Crawford (@tcrawford) April 30, 2014

@tcrawford @searchCIO A1 Right: Don't want to hem self into a corner based on data that may (or may not) be forward-looking enough. #CIOchat

— RachelTT (@RachelatTT) April 30, 2014

Though applying a prescriptive method might be simpler in industries with straightforward metrics, several basic benefits of employing prescriptive analytics outweigh any corporate hesitations:

1. Data is valuable

2. Analytics equals insights

3. Perspective mitigates risks

A1 Hard to imagine an industry where data doesn't matter-If there is data & decisions to be made, prescriptive analytics has a role #CIOChat

— Tom Doub (@tomdoub) April 30, 2014

The bottom line is that analytics provide insight that businesses can make decisions on. There is a balance to keep on metrics. #CIOChat

— Tim Crawford (@tcrawford) April 30, 2014

@ITCompliance So true. Mitigating risk perhaps the most important use of prescriptive/predictive analytics along w/ biz development #CIOchat

— RachelTT (@RachelatTT) April 30, 2014

More explicitly, what specific industries can derive top value from prescriptive analytics models? Tom Doub, CEO of the Centerstone Research Institute (CRI) and this month's SearchCIO tweet jam expert, suggests that healthcare IT tops the list, followed by customer-facing functions and others:

A1 Healthcare another example. Love to see what IBM & Sloan Kettering to integrate analytics into cancer care #CIOChat

— Tom Doub (@tomdoub) April 30, 2014

Our expert is no stranger to the healthcare field: His organization, CRI, is one of the nation's largest nonprofit providers of community-based care to individuals with mental-health issues. At CRI, Doub is currently overhauling the prediction model by analyzing patients' treatment and progress over time to pinpoint precisely when a patient starts to get better.

While the prescriptive benefit to health care is clear, tweet jam regular Andi Mann had to ask, "Where else?"

.@AndiMann Customer churn/customer service. What actionss should sales & service people make based on predicted customer behavior. #CIOChat

— Tom Doub (@tomdoub) April 30, 2014

@tomdoub Makes sense. I've read a lot lately about web analytics re: cart abandons, A-B tests, time-to-sale, etc. Rich datasource. #CIOChat

— Andi Mann (@AndiMann) April 30, 2014

@tcrawford @tomdoub That is really unfortunate. Even as a layman in medical, I see enormous potential. #CIOChat

— Andi Mann (@AndiMann) April 30, 2014

What does it take to get a prescriptive analytics plan rolling at your organization? SearchCIO Executive Editor Linda Tucci asked #CIOChat-ters, "Does IT have to adjust its rules of data management to get something like this going?" Some responses:

@LTucci IT first needs to understand the business they’re in. Then understand the data needed to characterize it. More than tech. #CIOChat

— Tim Crawford (@tcrawford) April 30, 2014

@LTucci Almost sure, esp. in larger orgs. Data is so often silo'd and protected; but correlation often better with larger dataset. #CIOChat

— Andi Mann (@AndiMann) April 30, 2014

@tomdoub @LTucci However, in order to do so, IT needs to clearly understand the business (i.e.: how their biz makes money). #CIOChat

— Tim Crawford (@tcrawford) April 30, 2014

@tcrawford @tomdoub #CIOChat Yes, the key is to start with the business problem with analytics, meaning you have to know the business.

— Linda Tucci (@LTucci) April 30, 2014

In what industries do you think the prescriptive method of analytics is most useful? What does it take to get the ball rolling? To read more from this #CIOChat conversation, head over to Twitter and follow @SearchCIO. Our next tweet jam will be Wednesday, May 28, at 3 p.m. EST (subject matter TBA).