Gartner Business Intelligence & Analytics Summit
March 14 – 16, 2016 at Grapevine, Texas, United States
Solution Provider Session Report
Ryohei Fujimaki, Ph.D., Research Fellow at NEC Corporation
Prescriptive Analytics: The Marriage of Your Business and Data Science
1) Prescriptive analytics starts with defining a clear business challenge.
2) The winning combination: Accurate prediction + optimization technology.
3) Incorporate measures to counteract limitations in data-derived decisions.
4) Advanced analytics is vital for the IoT era. Don't wait. Get started now!
Advanced Big Data analysis, and the investment flows to fund it, are moving from what happened and why, to more outcome-oriented questions such as what will happen, or how to make it happen. In today's increasingly prolific IoT era, more sophisticated business-centric "prescriptive analytics" is vital to business success.
Prescriptive analytics can help derive actionable decisions that directly solve business challenges and achieve a good return on investment. But the key for businesses is to first revert their way of thinking. Instead of starting from the data, they need to start from the business challenge, because they can only leverage the true power of data science once they have a feasible business challenge, process and operation in place.
These uses cases illustrate three key best practices for newcomers to prescriptive analytics:
1) Define business-centric analytics processes
Evidence-based proactive training – SMRT public bus operator, Singapore
- Business challenge: Reduce traffic accidents and ensure safety of 1 million daily commuters.
- Business process: Use full cabin simulator to train company's 2,600 drivers.
- Leverage data science: Use NEC's data analysis technology on a range of inputs to profile driver behavior and target drivers for proactive training.
- Result: 30% reduction in risk index, financial savings, brand image protection.
2) Maximize business results using optimization technology
Profit-oriented campaign optimization – Large telecom carrier
- Business challenge: Maintain profitable customer base.
- Business process: Proactively offer campaigns to retain customers before they lose loyalty.
- Leverage data science: Use NEC's optimization technology to transform proactive campaigns to profit-orientated campaigns.
- Result: 2.5 x higher profitability with optimization v. simple prediction technology.
3) Prescribe business practices for adjusting analytics results
New product production optimization – Asahi Breweries
- Business Challenge: Decide accurate production volumes for new products.
- Business process: Forecast demand for new weekly products and adjust production.
- Leverage data science: Use NEC's new product demand predicting, and develop business practices to prevent unexpected losses from anomalous results or additional factors.
- Result: Minimized over and under production.
NEC's 600 data and research scientists worldwide define business-centric prescriptive analytics solutions using the company's Heterogeneous Mixture Learning technology to automate highly accurate, large-scale predictions from multi-source, dynamically changing data, and Predictive Robust Optimization Framework technology to form advanced strategies and plans conventionally carried out by humans. NEC also offers end-to-end delivery from consultation and data science to managed services, cloud and outsourcing.
NEC is developing cutting-edge prescriptive analytics solutions from product pricing for retail outlets to optimum water distribution for entire cities. Representatives of this summit have suggested that, by 2020, 40% of net investment into advanced analytics will be into predictive and prescriptive analytics. Businesses need to pluck up the courage and propose right now, to unite highly accurate prescriptive analytics with optimized business results.