As the wave of big data and next generation business analytics continues to evolve, technologists and business users are under a storm of industry buzzwords surrounding big-data, analytics, and data science. Driving this industry buzz are the possibilities of a new approach to using data far beyond traditional business intelligence. New analytics tools and approaches have evolved from the traditional reports, static metrics, and established vendor tools into a desire and expectation for a dynamically driven data toolset that does more than deliver standardized reports and metrics.
However, when we drive to work we use more than the traffic reports or GPS instructions to get to our destination. We are also using our cognitive resources to adapt to what is happening right before us as we are driving.
Standardized traffic reports are helpful, but they don’t warn you that a mattress is going to fall off of the moving truck in front of you or that the businessman next to you is to focused on his call to notice he is drifting into your lane.
The current evolution of analytics and data practices seeks to extend our analysis capabilities beyond standardized business intelligence reporting and analysis into technologies and approaches that place decision makers in the driver’s seat; enabling them to instantly react to changes and make informed decisions. To understand the new direction of business analytics it may be easier to set aside the current round of buzzwords and focus on the desired analytics tools, predictive and prescriptive analytics, and the technologist position needed for make the most of these new tools, the data scientist. By setting ourselves apart from industry buzzwords and focusing on these key terms and the current phase of business analytics we will be able to establish a clear view of the road ahead.
Business Analytics 2.0, Predicting and Prescribing for the FutureThe business analytics lifecycle can be broken into phases. The majority of business analytics today falls into the realm of traditional business intelligence. Referred to as descriptive or diagnostic analytics. This approach uses structured, historical data to provide the standardized post-mortem reporting many of us are familiar with, our regularly updated morning traffic report.
|Gartner Symposium, Orlando, October 2012|
The next phase in the analytics lifecycle is predictive analytics. Predictive analytics combines structured historical data with unstructured data and rules and algorithms to understand the likelihood of a possible outcome. In a recent interview with InformationWeek, Atana Basu briefly describes predictive analytics, “I tell you what will happen and I leave it up to you to figure out what to do with it.” Predictive analytics moves us further down the road but does not complete the picture of what you should do “now”. Returning to our Monday morning traffic report; we learn that an accident has just occurred in front of us but the report does not advise us of a specific alternate route to take that is left up to us to decide.
The final phase of analytics uses structured and unstructured data, leverages the rules and algorithms developed from predictive analytics, and then suggests actions to implement from the predictions. Prescriptive analytics answers a broader set of questions. “It answers what will happen, when it will happen and why it will happen,” Basu says. “And then (it tells you) how to take advantage of this predictive future.” In our driving analogy we move closer to an automated commute where the car is driving based upon prescriptive analytics, think of the Google Car.
We have Big Data Analytics now what about the people? Enter the Data ScientistHowever our driving analogy begins to break apart when we look at applying prescriptive analytics to improving businesses. In his “Five Pillars of Prescriptive Analytics Success”, Basu shares that our dream of the automated commute is called prescriptive automation. For a business to both implement and harness the power of prescriptive analytics it still requires human assistance to build the analytics and then carry out the prescriptions. This need for human knowledge and action has developed an emerging technologist role referred to as the data scientist. Just as the analytics lifecycle is evolving so is the role of the business or data analyst in to the role of the data scientist.
A data scientist has many roles to fill in this next phase of business analytics. Whether it’s in helping to shape and build the predictive and prescriptive algorithms, identifying the right business problems to address, or providing informed conclusions and recommendations based upon prescriptive analysis and analytics. The data scientist is the required human element of prescriptive analytics. The data scientist helps us apply the newest analytics tools to avoid hitting the mattress falling of the truck in front of us or to avoid the distracted driver.
Getting down to Analytics, how do we learn more?With a clear vision of what the next phase of analytics looks like it’s important to know where to start learning more. Even though the next wave of analytics has great potential predictive analytics is just now being adopted widely and according to Gartner prescrpitive analytic is used by only 3% of organizations today.
The first step in adopting any new technology is understanding the basics, the vendors, and staff you will need to leverage the technology. Hopefully, this article has cut through the buzz to provide you the basics to look at predictive and prescriptive analytics vendors.
A leading vendor in the predictive analytics field is IBM. The IBM Predictive Analytics site has a wealth of information to review and provides detailed use cases on predictive analytics. For prescriptive analytics Atanu Basu’s company site for AYATA is also worth a review for more detail on AYATA’s approach to prescriptive analytics.
Universities are also beginning to offer graduate degrees in big data analytics to help train data scientists to fill the talent gap that the industry is experiencing. InformationWeek detailed 20 top big data analytics programs in January of 2013 as part of their special coverage series on big data.
Every new technology movement is filled with buzzwords and industry-speak that distract from the core use cases and vision of what makes the movement important. Hopefully this brief post has helped to shed away some of the mysticism of big-data analytics and the emerging role of the data scientist for you.
 Basu, Atanu. “Five Pillars of Prescriptive Analytics Success.” Analytics-Magazine.org, March/April 2013. http://www.analytics-magazine.org/march-april-2013/755-executive-edge-five-pillars-of-prescriptive-analytics-success, Accessed on 4/25/2013
 http://en.wikipedia.org/wiki/Prescriptive_analytics. Accessed 4/29/2013
 Bertolucci, Jeff. “Prescriptive Analytics and Big Data: Next Big Thing?.” InformationWeek, April 15, 2013. Accessed 4/29/2013.