Analytics May/June 2017 by Lionheart Publishing Inc.

H T T P : / / W W W. A N A LY T I C S - M A G A Z I N E . O R G

DRIVING BETTER BUSINESS DECISIONS

M AY / J UNE 2017

B2B

e-Commerce How to avoid costly mistakes

ALSO INSIDE:

Executive Edge Heine Krog Iversen, CEO of TimeXtender, on why data professionals must move up corporate ladder

• New paradigm: Service as a software • Big data & better business decisions • How to transition from CSP to DSP • Healthcare analytics: What’s next? • Storytelling skills: The write stuff • Corporate profile: Praxair model

INS IDE STO RY

Rx for healthcare debate Five months have come and gone since the U.S. presidential election and three months have come and gone since the inauguration, and yet the country remains evenly divided over almost everything even remotely related to politics. No, I didn’t expect partisan passions to magically dissipate after Election Day, but I did expect some sort of compromised agreement on how to best fix a U.S. healthcare and associated insurance system that all parties agree is flawed, needs fixing and is in everyone’s interest to do so. Turns out members of the majority House party initially couldn’t agree among themselves on how best to move forward, let alone bring members of the minority party into the discussion. Politics aside, is there any more critical issue facing the potentially uninsured or underinsured than healthcare, knowing that at any day a serious illness can potentially cost them their life savings if not their lives? Sure, the economy, the environment, terrorism, North Korea, Russia, China, Syria, the Middle East in general, etcetera, etcetera are all major concerns, but as the old saying goes, there’s nothing more important than your health. Thankfully, analytics has and will continue to shape U.S. healthcare in positive and meaningful ways. The Affordable 2

A N A LY T I C S - M A G A Z I N E . O R G

Care Act, despite its many flaws, mandated value over volume in patient care, and ushered in an era of accountable healthcare. Yale University professor and INFORMS Past President Ed Kaplan, over the course of his career, has shown that data-driven research can convince skeptical politicians to change public health policies, from their reluctance to a needle-exchange program to halt the spread of HIV/AIDS in Hartford, Conn., to their resistance of blood donations from Palestinians in Israel. In this issue of Analytics magazine, several contributors weigh in on the healthcare industry in the post-election era. In his “Viewpoint” column titled “Healthier use of healthcare data,” Tim Spaeth says, “While the deluge of data has become a challenge that impacts every aspect of the healthcare industry, it also represents a tremendous opportunity for IT and Finance to collaborate and create measurable value.” Meanwhile, Rajib Ghosh discusses the “Evolution of data organizations in healthcare” in his regular “Healthcare Analytics” column. For the latest in optimizing health service operations and outcomes, see the preview of the 2017 INFORMS Healthcare Conference. ❙

– PETER HORNER, EDITOR peter.horner@ mail.informs.org W W W. I N F O R M S . O R G

Ready for more robust analytics and optimization software? Boost productivity and collaboration with a technology that’s easy to customize, update and deploy. Find out how JBS, a global leader in beef, lamb and poultry processing, increased their margin by 25% by trading spreadsheets for AIMMS-based Prescriptive Analytics applications.

“

Thanks to the flexibility of AIMMS, the development of new optimization modules can be done with the agility comparable to the dynamism of our company and its market.” - João Batista Rocha de Souza, Integrated Planning Manager at JBS

Read the case study The AIMMS Prescriptive Analytics Platform helps you evaluate and identify the best options to tackle your most pressing challenges with sophisticated analytics that leverage mathematical modeling and scenarios while pulling from multiple data sources. You can immediately gauge, not just what is likely to happen, but what you should do about it for the best possible outcome. That’s why teams at JBS, Shell, GE and Heineken and many more fire up AIMMS every day.

C O N T E N T S

DRIVING BETTER BUSINESS DECISIONS

MAY/JUNE 2017 Brought to you by

FEATURES 32

NEW PARADIGM: SERVICE AS A SOFTWARE Leveraging the interconnectedness of business problems to accelerate better decision-making.

By Deepinder Dhingra

32 36

JOURNEY FROM CSP TO DSP Text mining will play a pivotal role in the transition from communications service providers to digital service providers.

By Somnath De, Saibal Samaddar and Upasana Mukherjee 44

LESSONS LEARNED: DECADE OF B2B ECOMMERCE Four common points of failure in B2B eCommerce initiatives, including ways to avoid re-platforming and other costly misfires.

By Steve Shaffer 50

LEVERAGING BIG DATA FOR BETTER BUSINESS DECISIONS To truly take advantage of the value of big data, the full process – from sourcing data to deploying models – must be supported.

By Dave Oswill

CORPORATE PROFILE: PRAXAIR Making our planet more productive: Fortune 300 company at the leading edge of the digitalization of manufacturing.

By Larry Megan and Kristin Bruton 64

STORYTELLING: THE WRITE STUFF New book explains how to present analytics to a non-technical audience. Hint: Shorter and clearer is better.

By Doug Samuelson

50 4

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

Welcome to Analytic Solver ® Cloud-based Optimization and Simulation that Integrates with Excel

Everything in Predictive and Prescriptive Analytics Everywhere You Want, from Concept to Deployment.

functions; easy multiple parameterized simulations, decision trees, and a wide array of charts and graphs.

The Analytic Solver® suite makes the world’s best optimization software and the fastest Monte Carlo simulation and risk analysis software available in your web browser (cloud-based software as a service), and in Microsoft Excel. And you can easily create models in our RASON® language for server, web and mobile apps.

Forecasting, Data Mining and Text Mining.

Linear Programming to Stochastic Optimization. It’s all point-and-click: Fast, large-scale linear, quadratic and mixed-integer programming, conic, nonlinear, non-smooth and global optimization. Easily incorporate uncertainty and solve with simulation optimization, stochastic programming, and robust optimization.

Comprehensive Risk and Decision Analysis. Use a point-and-click Distribution Wizard, 50 probability distributions, automatic distribution fitting, compound distributions, rank-order correlation and three types of copulas; 50 statistics, risk measures and Six Sigma

Analytic Solver is also a full-power, point-and-click tool for predictive analytics, from time series methods to classification and regression trees, neural networks, and access to SQL databases and Spark Big Data clusters.

Find Out More, Start Your Free Trial Now. In your browser, in Excel, or in Visual Studio, Analytic Solver comes with everything you need: Wizards, Help, User Guides, 90 examples, even online training courses. Visit www.solver.com to learn more or ask questions, and visit analyticsolver.com to register and start a free trial – in the cloud, on your desktop, or both!

Tel 775 831 0300 • Fax 775 831 0314 • info@solver.com

DRIVING BETTER BUSINESS DECISIONS

DEPARTMENTS

2 8 12 16 20 24 70 72 74 76

Inside Story Executive Edge Analyze This! Healthcare Analytics Viewpoint Newsmakers Conference preview: Healthcare Conference preview: Marketing Science Five-Minute Analyst Thinking Analytically

President Brian Denton, University of Michigan President-Elect Nicholas Hall, Ohio State University Past President Edward H. Kaplan, Yale University Secretary Pinar Keskinocak, Georgia Tech Treasurer Michael Fu, University of Maryland Vice President-Meetings Ronald G. Askin, Arizona State University Vice President-Publications Jonathan F. Bard, University of Texas at Austin Vice President Sections and Societies Esma Gel, Arizona State University Vice President Information Technology Marco Lübbecke, RWTH Aachen University Vice President-Practice Activities Jonathan Owen, CAP, General Motors Vice President-International Activities Grace Lin, Asia University Vice President-Membership and Professional Recognition Susan E. Martonosi, Harvey Mudd College Vice President-Education Jill Hardin Wilson, Northwestern University Vice President-Marketing, Communications and Outreach Laura Albert, University of Wisconsin-Madison Vice President-Chapters/Fora Michael Johnson, University of Massachusetts-Boston INFORMS OFFICES www.informs.org • Tel: 1-800-4INFORMS Executive Director Melissa Moore Director, Public Relations & Marketing Jeffrey M. Cohen Headquarters INFORMS (Maryland) 5521 Research Park Drive, Suite 200 Catonsville, MD 21228 Tel.: 443.757.3500 E-mail: informs@informs.org ANALYTICS EDITORIAL AND ADVERTISING

Analytics (ISSN 1938-1697) is published six times a year by the Institute for Operations Research and the Management Sciences (INFORMS), the largest membership society in the world dedicated to the analytics profession. For a free subscription, register at http://analytics.informs.org. Address other correspondence to the editor, Peter Horner, peter.horner@mail.informs.org. The opinions expressed in Analytics are those of the authors, and do not necessarily reflect the opinions of INFORMS, its officers, Lionheart Publishing Inc. or the editorial staff of Analytics. Analytics copyright ©2017 by the Institute for Operations Research and the Management Sciences. All rights reserved.

A N A LY T I C S - M AGA Z I N E . O RG

Lionheart Publishing Inc., 506 Roswell Street, Suite 220, Marietta, GA 30060 USA Tel.: 770.431.0867 • Fax: 770.432.6969

President & Advertising Sales John Llewellyn john.llewellyn@mail.informs.org Tel.: 770.431.0867, ext. 209 Editor Peter R. Horner peter.horner@mail.informs.org Tel.: 770.587.3172 Assistant Editor Donna Brooks donna.brooks@mail.informs.org Art Director Alan Brubaker alan.brubaker@mail.informs.org Tel.: 770.431.0867, ext. 218 Advertising Sales Aileen Kronke aileen@lionhrtpub.com Tel.: 678.293.5201

Welcome to Analytic Solver ® Cloud-based Data and Text Mining that Integrates with Excel

Everything in Predictive and Prescriptive Analytics Everywhere You Want, from Concept to Deployment. The Analytic Solver® suite makes powerful forecasting, data mining and text mining software available in your web browser (cloud-based software as a service), and in Microsoft Excel. And you can easily create models in our RASON® language for server, web and mobile apps.

Full-Power Data Mining and Predictive Analytics. It’s all point-and-click: Text mining, latent semantic analysis, feature selection, principal components and clustering; exponential smoothing and ARIMA for forecasting; multiple regression, logistic regression, k-nearest neighbors, discriminant analysis, naïve Bayes, and ensembles of trees and neural networks for prediction; and association rules for affinity analysis.

distributions, 50 statistics and risk measures, rankorder and copula correlation, distribution fitting, and charts and graphs. And it has full-power, point-and-click optimization, with large-scale linear and mixed-integer programming, nonlinear and simulation optimization, stochastic programming and robust optimization.

Simulation/Risk Analysis, Powerful Optimization. Analytic Solver is also a full-power, point-and-click tool for Monte Carlo simulation and risk analysis, with 50

Tel 775 831 0300 • Fax 775 831 0314 • info@solver.com

EXE CU TIVE E D G E

Hungry for data, thirsty for time Awash in data and opportunities: Why data professionals must move up the corporate food chain.

BY HEINE KROG IVERSEN

Organizations of all sizes and types are awash in data possibilities, yet most of them cannot capitalize on the potential for a variety of reasons. The good news, however, is that with the right decisions and focus, these possibilities can turn quickly into realized opportunities. Business users in these organizations are hungry for comprehensive, contextual and timely data on which to make decisions. The process of “idea creation, data gathering, decision” is a motivational force for economic development in all organizations. The key to effectively implementing this is the smooth democratization of access to data. Can, in one fell swoop, an organization satisfy its newly enabled users’ data hunger while simultaneously quenching information technology’s (IT) thirst for time? That’s the challenge.

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

As organizations strive for data agility, breaking down internal barriers to data democratization is essential. To stretch the metaphor, dynamic organizations must be able to simultaneously eat and drink. Data democratization and the ensuing benefits are not an episodic or ephemeral phenomenon for the organization; instead, they are fundamental and constant elements of future success. Data infrastructure technology has come of age and is ready to be deployed and exploited in the service of data agility. It is imperative for all organizations to create a plan and chart a clear course of action. Itâ&#x20AC;&#x2122;s time to banish both hunger and thirst. Most C-level executives by now have recognized that becoming a data-driven company is a priority to their success and that it plays an instrumental role in increasing profit and growth. As technology continues to push the speed of change, we can be assured that this speed of change will continue to increase even faster, and will put even greater pressure on companies to adjust. The amount of IT systems in place is growing, and the average lifecycle of each system is being lowered dramatically. For instance, more and more organizations are trying to keep up by implementing scrum and agile approaches to their implementations. A NA L Y T I C S

Another way to look at this is that business organizations are putting pressure on IT, and getting more work accomplished in less time demands having data as needed, when needed. Businesses also mandate that new systems are put into the landscape, forcing IT to know and understand these systems at the speed of light. If IT canâ&#x20AC;&#x2122;t deliver, companies just bypass IT and turn to cloud offerings. All these new systems and constant change of systems puts pressure on the way organizations build and maintain a governed platform for data analysis. When you add in big data, IoT and all the social data that flows around the business, we can now see that companies are looking for help to provide organization, quality and cost reduction around these systems. The need for speed and the current increasing workload forces data professionals to get out of their comfort zone. Companies that continue to operate as they have over the last 20 years will not survive. DEMOCRATIZING DATA ANALYSIS AND DECISION-MAKING Big data and social data, along with other data systems and applications, are driving the intelligent enterprise movement. For example, Twitter, Facebook and M A Y / J U N E 2 017

EXE CU TIVE E D G E

LinkedIn are now an integrated part of almost every organization’s marketing system. Organizations need to monitor what people are saying about them (the unstructured part), and they need to monitor their followers and how this varies based on what they are doing. It is also important to connect all this data to the financials to measure the cost per follower or what is called “the structured part.” Data analysis is now becoming a top priority in selecting go-to-market strategies, as well as in the big picture strategy to obtain market leadership. Therefore, companies expected to climb to the top are those that combine internal and external data. Going forward, IT budgets are expected to be taken over by the chief marketing officer or the chief financial officer. This will give IT less impact regarding decisions on what IT will be used, yet IT will have to run and support all these systems. Meanwhile, numerous recent published reports outline in detail the emergence of data discovery, and why organizations must now consider a governed data discovery hub as their central platform for democratizing data analysis and decision-making. Current research also shows that speed in delivering new IT systems or new platforms for data analysis is becoming more important than trying to 10

A N A LY T I C S - M A G A Z I N E . O R G

cut costs. If you believe the proposition that speed is paramount for survival and growth (and I do, by the way), then the only way forward is to empower business users and liberate IT, which can easily be accomplished with data democratization (as long as governance is ensured). Numerous companies are now leveraging data democratization platforms to improve operational performance by saving time, making better decisions and redeploying staff for more strategic, benefit-oriented initiatives. Further, companies are seeing better data-driven decisions by improving the line of business with big data and social data initiatives. Having governed and localized data discovery at the fingertips of all business users plays an essential role for organizations trying to lower costs, improve efficiency and enhance data quality. Today, with the deluge of data and the advent of big data and social data, it is imperative to re-engineer the enterprise in this fashion. Companies that do have a much greater chance at gaining a competitive advantage and even survival, while those that don’t risk dying of hunger or thirst. ❙ Heine Krog Iversen (hki@timextender.com) is the CEO of TimeXtender, a software leader dedicated to democratizing access to corporate data through Discovery Hub, and the largest provider of data warehouse automation software for the Microsoft SQL Server.

W W W. I N F O R M S . O R G

BIG DATA OPTIMIZED ODHeuristics is designed with massive models in mind. It’s created by Optimization Direct’s experts – global leaders innovators in the field of optimization. It runs under IBM’s CPLEX and takes advantage of the multi-core processing power of modern computing by breaking models into multiple threads. The combined innovation of Optimization Direct with the power of CPLEX, gives effective results on packing problems, supply chain and telecoms as well as scheduling applications. On large scale MIPs it provide solutions that are often beyond the reach of traditional optimization methods. Find out more and challenge the Optimization experts with your large matrices, find us at optimizationdirect.com

The IBM logo and the IBM Member Business Partner mark are trademarks of International Business Machines Corporation, registered in many jurisdictions worldwide. *IBM ILOG CPLEX Optimization Studio is trademark of International Business Machines Corporation and used with permission.

ANALY ZE TH I S !

Problem-solving: Keep it real with gemba When it comes to mathematical modeling, “the real place” is usually messy.

BY VIJAY MEHROTRA

Some years ago, I got a call from “Frank,” a finance director at a start-up company with a cloudbased software solution. Its platform was hosted by one of the large public cloud providers, and that was why he was calling. “The bills for hosting have been outrageous,” said Frank, who had called me based on a strong referral from a former consulting colleague. “Charges are rising at a ferocious rate, much faster than revenue. As far as data, we do get a bill from our provider with a ton of detail in it, but we are having a hard time getting our arms around what’s driving the peaks in the traffic loads. We are really worried about managing these costs as we grow.” As it happens, I have a background in queueing and experience with load forecasting in many different industries, so this project seemed to be a good fit with my background. Wrapping up this initial conversation, Frank suggested that the right next step was a

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

follow-up conversation with “Oscar,” the director of operations from the company’s engineering organization. Later that day, Frank sent an introductory email to the two of us, strongly suggesting that Oscar make some time to bring me up to speed on the issues they were facing. On Friday afternoon of that week, I spoke with Oscar for the first time. For much of our conversation he seemed to be speaking a foreign language. He talked about customer pods, network latency, CPU densities, API calls and user counts. He quickly explained the structure of their data warehouse, with obtuse references to fields with cryptic names and unclear definitions. Overall, his tone suggested that he doubted that an accurate cost prediction model could be developed by anyone, much less an outsider with no experience with cloud computing architecture. Over the next couple of weeks, there were several follow-up conversations with Frank’s finance folks, Oscar’s ops team and people from other parts of the company. Sadly, within a month or so, the project was abandoned. My sense was that it was basically my fault for not knowing enough about the domain. As data sets continue to grow larger and larger, there are more and more specialized individuals and teams dedicated A NA L Y T I C S

to searching for patterns that can be exploited for one business purpose or another. If this is what your current job or project is like, perhaps you think that understanding the business context is not all that important. My perspective is different. After losing out on the project opportunity with Frank and Oscar, I am now an even bigger advocate of going to “gemba,” a term that is very familiar to practitioners of various lean and quality management methodologies. In Japanese, the literal meaning of gemba is “the real place,” in contrast to the simplified models of other people’s reality with which most of us are more comfortable. Gemba is usually messy. The late Gene Woolsey, professor of operations research at the Colorado School of Mines, believed that this trip to gemba should be more than a stop at the drive-thru window [1]. Each of his graduate students was required to learn to do the job(s) associated with a particular process or system before they would be allowed to develop a model that purported to improve it. Woolsey saw this as essential to any successful project, both for the domain knowledge that would be acquired in the process of learning these roles and for the credibility and trust that need to be established for most solutions to M A Y / J U N E 2 017

ANALY ZE TH I S !

be accepted. When I first heard about Woolsey, the requirements he placed on his students seemed extreme and unnecessary to me. It took me years to understand that this type of firsthand engagement was usually incredibly valuable. Gemba, Part 1: The (real) problem. Most of the time, the description of the business problem that you first receive at the start of the project is either incorrect, incomplete or both. This is not because the people who provided the original problem statement to you are fools and/or scoundrels, but rather because any snapshot that is taken is influenced by the position, time and scope at which the picture is taken. While the available data may offer some insights, the investigation of the business context, particularly through structured face-to-face conversations with key participants early in the process, can provide vital clues. A friend of mine once received a call from a manufacturer asking us to look at their historical data to develop an improved model for forecasting market prices for key inputs. Through initial discussions, however, what he discovered was that the company’s real need was for a model to support its 14

A N A LY T I C S - M A G A Z I N E . O R G

sourcing group to negotiate contracts with suppliers. This discovery led to a successful project, albeit one that was quite different than what had first been described by the project sponsor. Gemba, Part 2: The (real) context for the data. With most analytics projects, there is usually no way to get to know the data without taking the time to really understand the underlying business context. Without taking the time to understand the processes that drive the data capture, you can easily be confused or worse, misled, by unclear field definitions, underlying capture logic and time and/or space dimensionality or many other foundational concepts. Without understanding the broader context, you can often incorrectly identify data points as outliers. And without some understanding of the context, there may be an infinite number of initial visualizations that you can create, but a much smaller number that can help you to develop good hypotheses to investigate and/or to make good decisions about what to include or exclude from your model. Gemba, Part 3: Credibility and Trust (the real “real place”). Whether you are a consultant, a new hire or a project team from another part of the organization, it is possible that those whose “problem” you W W W. I N F O R M S . O R G

have been asked to investigate will view you with some suspicion. Most likely the decision to hire you was made by someone else and thrust upon them, an odd kind of arranged marriage. From the perspective of the people you have been sent to “help,” you do not know them nor do you know their business. Your project is sure to disrupt their lives – and you surely will not be around to clean up the mess you have left them.

Ivan B. Class of ‘18 Oil and Gas

REFERENCE 1. For much more of the wisdom of Woolsey, see http://www.lionhrtpub.com/books/wp1.html.

How to deal with this particularly vexing challenge? We will dig into that next time. ❙ Vijay Mehrotra (vmehrotra@usfca.edu) is a professor in the Department of Business Analytics and Information Systems at the University of San Francisco’s School of Management and a longtime member of INFORMS.

“This program has helped me develop some great tools for my analytics belt.” -Spring 2016 Exit Survey

MORE INFORMATION: ANALYTICS.STAT.TAMU.EDU

Informs ad.indd 1

A NA L Y T I C S

10/17/16 2:57 PM

M A Y / J U N E 2 017

HEALT H CARE A N A LY T I C S

Evolution of data organizations in healthcare The uproar about healthcare made us confront the age-old question: Is healthcare a right or an entitlement?

BY RAJIB GHOSH

The big news for the Affordable Care Act (ACA) in March was the failure of the alternative healthcare bill on the House floor. For the first time since the election campaign season began, we heard real conversations between politicians and their constituents about healthcare. The uproar about healthcare in recent times made us confront the age-old question in the United States: Is healthcare a right or an entitlement? Despite the healthcare dialogue during and after the campaign, weâ&#x20AC;&#x2122;re still struggling to find an answer. The ACA didnâ&#x20AC;&#x2122;t address this directly, although it reduced the percentage of uninsured people nationwide. Surely it increased the cost initially for the insurers, but the overall healthcare cost showed a slower rate of increase during the last few years. The expectation was that over time, the total cost would be lowered as high-utilizers of healthcare resources would get access to their healthcare at lower-cost delivery centers such as primary care office visits rather than at higher-cost hospital emergency rooms. Many times, those

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

higher-cost episodes were also uncompensated, causing an increase in cost for paying customers and eventually to the insurers. Insurers, in turn, increased their premiums for all customers, individual or group purchasers, to recoup their losses. EMERGENCE OF THE CHIEF DATA OFFICER ROLE Regardless of the policy and politics, data has become the king in all industries, including healthcare, as organizations worldwide have added the role of a chief data officer (CDO) at a rapid rate the last few years. Gartner predicts that this trend will continue. However, the definition of the role is still not well articulated in healthcare. Some organizations have a CDO and a separate chief analytics officer (CAO). Others have fused them into one. The nascent CDO position faces challenges due to role ambiguity. To understand the landscape better I recently attended a summit for CDOs. Event organizers invited leaders in data and analytics from various industries to an informal setting to discuss their challenges and successes. Attendees from different healthcare organizations, which are at different levels of maturity in terms of using data for their business, participated in several roundtable discussions on a range of issues. In this article, I will underscore some of those core topics A NA L Y T I C S

that participants described as critical to achieve sustained competitive advantage for their data organizations. WHERE DOES DATA FIT IN ORGANIZATIONAL STRUCTURE? As companies continue their pursuit of transforming themselves, boardrooms are embracing the concept of data as a strategic asset. However, most industries struggle to identify the appropriate reporting structure for a data organization. Some companies place it under Marketing, some under IT. Others choose Operations or Strategy. Even within the same industry there is no common pattern. Typically, healthcare organizations build data and analytics departments under the supervision of Clinical or IT depending on the organizationâ&#x20AC;&#x2122;s perception of data. If it perceives data and analytics as a primary tool for clinical quality improvement, a Clinical silo is the best fit. If it is considered as another technology, then IT is the chosen silo. None of these structures, however, does justice to the role of data and analytics. In my opinion, data organization in healthcare should directly report to the CEO. Operations, Finance, Marketing, Strategy and Clinical â&#x20AC;&#x201C; all silos of a healthcare organization â&#x20AC;&#x201C; benefit from using data and analytics as a transformative tool. Some organizations M A Y / J U N E 2 017

HEALT H CARE A N A LY T I C S

Figure 1: Organization structures. Which one is best?

have started to create a new structure where the data and analytics department is part of the Office of Transformation, which reports to the CEO. That makes more sense to me. After all, organizations can transform themselves using data to become customer-driven, operationally excellent and more profitable with a sustained effort to do the right things at the right time. In many cases that also includes correctly predicting their future. This canâ&#x20AC;&#x2122;t be achieved without the data organization having a direct link to the CEO (Figure 1). CENTRALIZED VS. FEDERATED MODEL FOR DATA AND ANALYTICS Most organizations, especially large ones, operate in a hub and spoke model. Regional entities, satellite offices or

A N A LY T I C S - M A G A Z I N E . O R G

sister organizations all prefer to stay independent especially when it comes to managing their own data and building analytics for their business units. This is also true in large health systems. However, having centralized governed data sources and analytics programs has its benefits. So how do we balance the desire for a loosely coupled federated structure with the advantage of economy of scale of a centralized model? There is no easy answer to this question. Organizations have taken different approaches to resolve such conflicts. Some keep the data stewardship function decentralized, while higher-cost functions such as data science or data visualization are maintained centrally. Some organizations have adopted a hybrid approach, where most functional roles are duplicated while keeping the scope of the work of different units from overlapping. Healthcare organizations have adopted all the above. IS DATA GOVERNANCE REQUIRED IN THE BIG DATA WORLD? Some organizations that deal with big data raise questions about the importance of the data governance functions such as data quality analysis or data

W W W. I N F O R M S . O R G

stewardship. With big data, the quality of individual data elements does not matter as much it does for “small data” environments. In healthcare, most use cases are based on small data sets and, therefore, the importance of data governance and data quality management is high. Regardless of the size of the data, some key functions of data governance such as data validation and data access policies and procedures are important for any organization, especially healthcare. This is becoming even more important as healthcare delivery organizations are becoming larger in size with ongoing merger and acquisition activities. “Region and corporate,” or the “hub-andspoke” model, is quite common for health systems. Large organizations such as Kaiser Permanente and Sutter Health are prime examples. Leaders of both organizations consider data governance a foundational building block for their analytics programs. ACA dodged the bullet in March, but the political firing will continue as the

A NA L Y T I C S

congressional reform effort rolls on. As a result, the healthcare industry, both payers and providers, will experience uncertainty. In the end, I am sure that all parties will accept the fact that healthcare in the United States is too complex to dramatically change. Many questions remain unanswered. Despite that, the CDO summit showed me that the enthusiasm to deal with the organizational challenges of data and analytics remains as high in healthcare as it is in other industries that are not facing such uncertainties. That is reassuring. Our data journey is very much on – rain or shine. ❙ Rajib Ghosh (rghosh@hotmail.com) is an independent consultant and business advisor with 20 years of technology experience in various industry verticals where he had senior-level management roles in software engineering, program management, product management and business and strategy development. Ghosh spent a decade in the U.S. healthcare industry as part of a global ecosystem of medical device manufacturers, medical software companies and telehealth and telemedicine solution providers. He’s held senior positions at Hill-Rom, Solta Medical and Bosch Healthcare. His recent work interest includes public health and the field of IT-enabled sustainable healthcare delivery in the United States as well as emerging nations.

M A Y / J U N E 2 017

VIE W POIN T

Healthier use of healthcare data What keeps the CFO up at night is how to run the company more effectively to generate revenue, lower costs and improve productivity.

BY TIM SPAETH

Back in the days before the data deluge, there were clear and distinct roles regarding IT and Finance in the healthcare industry. When patient and payer medical information was captured on multipage paper documents attached to a clipboard, it was easy to manage the data, however ineffectively. But those days are long gone. Healthcare organizations now need to take a serious look at the potential business value of developing a strategic partnership with IT in order to create analytic solutions that deliver quantifiable results. Traditionally, the IT team has focused on maintaining data security, reducing risk and managing infrastructure uptime. But what keeps the CFO up at night is how to run the company more effectively to generate revenue, lower costs and improve productivity. Today, the opportunity exists to create a dynamic partnership that helps both organizations achieve success and drive positive business results. To the IT team, data is something that needs to be stored and managed and protected. They tend to view it from a very transactional perspective, focusing on proper formatting and whether it has been organized into the right categories. They have little or no big picture perspective on how it can be used to improve billing accuracy or reduce costs and increase profitability. Thatâ&#x20AC;&#x2122;s not their job.

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

On the other hand, the CFO wants to know where there are potential sources of revenue and ways to drive cost savings. Current analytics tools have the ability to answer these and many other types of data-related questions, but solving the problem requires aligning two historically separate agendas. The IT team wants to run analytics in as risk-free and Healthcare organizations need to take a serious look at the potential business value of efficient a manner as pos- developing a strategic partnership with IT. Photo Courtesy of 123rf.com | ÂŠ Takashi Honma sible and often without the proper subject matter expertise regardmany IT departments, they feel that their ing what data to capture and measure. job stops there. Finance, on the other hand, doesnâ&#x20AC;&#x2122;t However, from the perspective of the clearly understand the capabilities IT has Finance team, accurate and consistent that could drive new insights based on data submission is the expectation, the tathe accessible data and, as a result, finds ble stakes. What they really want to know itself at the mercy of IT to develop and is whether all the data made it through, build models to answer these questions. were there any errors, and if yes, what Take for example the process surwere those errors? And ultimately, who rounding claim submissions to the governcan fix those errors? Very often, itâ&#x20AC;&#x2122;s not ment that eventually results in payment in fact IT since these kinds of errors are to a health plan. A typical plan sends milrelated to other stakeholders and include lions if not billions of transactions each things such as enrollment information, year to the government. Viewed from a provider-specific data or diagnosis codes. strictly IT perspective, the primary goal But with access to this kind of data, is to get the data into the proper format the potential for cost savings and and submit it. This is, of course, required improved productivity are enormous. in every situation and unfortunately for Using analytics, Finance can quickly A NA L Y T I C S

M A Y / J U N E 2 017

VIE W POIN T

identify and prioritize errors, ignoring the unimportant ones while focusing on those that have the largest business impact. These include ones related to providers with submission issues or outliers with potential compliance problems. This kind of insight leads directly to identifying potential fraud, waste and abuse areas, all things that IT is not typically concerned with. While the deluge of data has become a challenge that impacts every aspect of the healthcare industry, it also

represents a tremendous opportunity for IT and Finance to collaborate and create measurable value. Companies who take advantage of this new partnership will quickly move ahead of the competition and see significant business benefit. ❙ Tim Spaeth, senior vice president, payer solutions, leads Episource’s RAPS/EDPS submissions and analytics services, directs the company’s risk adjustment consulting practice, and provides strategy around analytics for both Medicare and ACA. He has more than 10 years of experience leading risk adjustment projects and revenue analyses in the healthcare industry, as well as 20+ years of experience in finance.

Healthcare 2017 OPTIMIZING OPERATIONS & OUTCOMES

INFORMS Healthcare 2017 brings together academic researchers in “healthcare analytics” & industry stakeholders who are applying & sharing research to improve the delivery of effective healthcare.

Keynote Speakers

Dimitris Bertsimas Operations Research Center, Massachusetts Institute of Technology Brian Denton Department of Industrial & Operations Engineering, University of Michigan Dr. Eric de Roodenbeke CEO, International Hospital Federation

Early Discount Rate Deadline is Monday, June 5

http://meetings.informs.org/healthcare2017

A N A LY T I C S - M A G A Z I N E . O R G

HEALTHCARE 20 7 Rotterdam, Netherlands | July 26–28, 2017

W W W. I N F O R M S . O R G

Want to deploy more applications in 2017? Now, you can build and deploy ANY optimization or analytic application 80% faster with FICO® Optimization Modeler (powered by Xpress). This platform delivers scalable and high-performance algorithms, a flexible modeling environment, and unmatched rapid application and reporting capabilities.

FICO is a proud sponsor of the 2017 INFORMS Analytics Conference Email us at (+%1#OGTKECU"ƒEQ EQO to schedule a demo, or visit Booth 32 in the Exhibit Hall for your chance to win a mini drone, and other great prizes. © 2017 Fair Isaac Corporation. All rights reserved.

NE W S M AK E R S

Edelman Award, INFORMS Prize and other honors Holiday Retirement wins Edelman by improving the pricing model for its more than 300 senior living communities.

HOLIDAY RETIREMENT EARNS EDELMAN AWARD Holiday Retirement won the 2017 Franz Edelman Award for Achievement in Operations Research and the Management Sciences from INFORMS for its use of operations research (O.R.) to improve the pricing model for its more than 300 senior living communities across the United States. The prestigious Edelman Award, considered the “Super Bowl” of O.R. practice, was presented at an Oscar-like gala held in conjunction with the INFORMS Conference on Business Analytics & Operations Research in April in Las Vegas following a nearly year-long process that included vetting and site visits, as well as judged presentations by finalists at the Las Vegas conference. With approximately $1 billion in annual revenue, Holiday Retirement is the largest private owner and operator of independent senior living communities in the United States. The company partnered with Prorize LLC, an Atlanta-based revenue management firm that leveraged O.R. to develop its senior living rent optimizer (SLRO), the first revenue management system in the industry. The SLRO system enables a consistent and proactive

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

Scott Shanaberger (fourth from left), chief operating officer at Holiday Retirement, and the Edelman Award-winning team celebrate at the Edelman Gala. Photo Courtesy of INFORMS

pricing process across Holiday, while simultaneously providing optimal pricing recommendations for each unit in every one of their communities. “We are thrilled to be presented with the Franz Edelman Award, which recognizes the greatest achievement and most significant impact in operations research and analytics. This is an incredible accomplishment and honor,” said Scott Shanaberger, chief operating officer at Holiday Retirement. “This recognition is the realization of the hard work and collaboration of many people at both Holiday and Prorize in harnessing the power of operations research to transform our senior living operations and business model.” A NA L Y T I C S

This year’s other Edelman Award finalists included: • American Red Cross for “Analyticsbased Blood Collection Methods” • Barco for “Platform-based Product Development” • BHP Billiton for “Detailed Integrated Capacity Estimate (DICE) Model” • General Electric (GE) for “RailConnect 360” • New York City Department of Transportation for “Off-Hours Delivery (OHD) Program” First awarded in 1972, the Franz Edelman Award recognizes and rewards outstanding contributions of analytics and O.R. in the for-profit and non-profit M A Y / J U N E 2 017

NE W S M AK E R S

sectors around the globe. Each year, INFORMS honors finalist teams that have improved organizational efficiency, increased profits, brought better products to consumers, helped foster peace negotiations, and saved lives. Since its inception, the cumulative dollar benefits from Edelman finalist projects have surpassed $240 billion.

on Business Analytics & Operations Research in April in Las Vegas. The INFORMS Prize honors effective integration of operations research into organizational decision-making. The award is given to organizations that have repeatedly applied the principles of O.R. in pioneering, carried, novel and lasting ways. The U.S. Air Force launched its operations research program in 1942, which U.S. AIR FORCE AND DISNEY RECEIVE THE 2017 INFORMS PRIZE has since grown to include approximately The U.S. Air Force and The Walt 539 military and 849 civilian operations reDisney Company both received the 2017 search analysts serving in nearly 300 Air INFORMS Prize for their pioneering Force organizations and agencies. The diand enduring integration of operations verse applications of O.R. span operational research (O.R.) and analytics programs effectiveness, acquisition of new systems, into their organizations. The prizes were cost analysis, manpower and personnel, presented at the INFORMS Conference along with logistics and infrastructure. Air Force analysts provide timely, credible and defendable analyses that have been and continue to be essential to informing decisions. “The Air Force’s commitment to operations research was forged 75 years ago in the urgency of World War II and continues today,” says Kevin Williams, director of Air Force studies, analyses and assessments. “We conThe U.S Air Force (left side) and The Walt Disney Company (right side) were co-recipients of tinue to drive analytical inthe 2017 INFORMS Prize. Photo Courtesy of INFORMS novation and excellence as 26

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

we address the challenges of the future in nearly every facet of the planning and execution mission of the United States Air Force.” From improving the consumer and guest experience, to creative content development and operational excellence, The Walt Disney Company uniquely leverages diverse analytic applications to create innovation and value across nearly every system and organization within the company. In addition, Disney brings together analytics professionals and industry leaders to connect and share best practices to advance the field as part of an annual conference. “Driven by our desire to create, innovate, learn and inspire, we use analytics and data to enhance our business in unique ways,” says Mark W. Shafer, senior vice president, revenue & profit management, Walt Disney Parks and Resorts. “At Disney, we’re storytellers, and analytics have become part of our story, and will ensure we continue to create magic for years to come.” AIR FORCE ACADEMY’S O.R. PROGRAM SALUTED The U.S. Air Force Academy won the 2017 UPS George D. Smith Prize for its operations research (O.R.) program, which prepares graduates to become frontline O.R. practitioners as analysts A NA L Y T I C S

in the Air Force. The program exposes more than 50 percent of cadets to at least one O.R. course and provides cadets the opportunity to graduate with a Bachelor of Science degree in O.R. A year-long applied senior capstone serves as the apex of the program, during which teams of cadets consult for military, corporate, local government and nonprofit organizations to address real-world problems. Organizations that have teamed with the USAF program in the past include DARPA, the Missile Defense Agency, Lockheed Martin, Walmart and the Healing Warriors Program. Named in honor of the late UPS Chief Executive Officer – a champion of operations researchers at a leading Fortune 500 corporation – the UPS George D. Smith Prize is created in the spirit of strengthening ties between industry and the schools of higher education that graduate young practitioners of operations research. The prize is awarded to an academic department or program for effective and innovative preparation of students to be good practitioners of operations research or analytics. “By teaming with a variety of organizations to provide its students access to real-world data and problemsolving opportunities, the Air Force Academy stands out among other academic institutions in its dedication to M A Y / J U N E 2 017

NE W S M AK E R S

the University of Warwick, Nottingham University and Cardiff University in the United Kingdom developed a new method to produce betterperforming soybean strains that could lead to a worldwide increase in food production. Their solution was developed as part of the first annual O.R. and Analytics Student Team Competition. The new Jack Levis of UPS (far left) and Prize Committee Chair Robin Lougee (far right) congratulate competition, organized by the USAF Academy. Photo Courtesy of INFORMS INFORMS and sponsored this preparing graduates for success,” said year by Syngenta, challenges students Melissa Moore, executive director of to apply their considerable talent to INFORMS. “As the demand for O.R. and develop solutions to some of the biggest analytics professionals continues to rise, challenges facing our world today. it is more important than ever for young Eight teams comprised of graduate professionals to stand out as the top talent. and undergraduate O.R. and analytics The Air Force Academy’s program makes students from around the world comsure they do.” peted in this year’s inaugural completion. The award was presented at the 2017 Each team used the same data sets and INFORMS Conference on Business software options to develop solutions to Analytics & Operations Research. provide unique new analysis of data on For more information about the UPS soybean varieties bred for commercialGeorge D. Smith Prize, click here. ization, and propose solutions for developing better performing plants. “The O.R. and Analytics Student STUDENT TEAM COMPETITION TARGETS FOOD PRODUCTION Team Competition not only shines a With the backdrop of a fastspotlight on the incredible young talent growing global population and looming that is preparing to enter the field of O.R. challenges to produce enough food to and analytics, but provides these student feed the world, a group of students from teams the opportunity to hone the skill 28

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

sets that will place them on the path to success as they embark on their careers,” said Melissa Moore, executive director of INFORMS. “From problem-solving and teamwork, to documenting and communicating their findings, the competition prepares students for the demands of a real-world workplace experience.” Aurelie Thiele presents the U.K.-based team of Peter Riley, Anna Scholes and Adam Green “We are honored to have (l-r) as the winner of the O.R. and Analytics Student Team Competition. Photo Courtesy of INFORMS won the first O.R. and Analytics Student Team Competition,” said methodology selection, data use, model Peter Riley, student team leader. “Not building, and quantitative analysis. only did we have the chance to test “Syngenta is proud to support innoour skills against brilliant students from vative thinking among emerging young around the world, but we were able to do leaders in analytics,” said Joseph Byrum, so in way that could help make a lasting senior R&D strategic marketing executive difference for countless people. It makes with Syngenta and Syngenta lead for the for a truly rewarding and fulfilling experiStudent Competition organizing commitence, and we look forward to building on tee. “This competition served as an exwhat we’ve learned and accomplished cellent entry point for students interested as we move forward in our studies and in learning more about how analytics can careers.” help solve the challenges facing agriculThe winning team was announced ture today.” at the 2017 INFORMS Business AnalytIn addition to the winning team from ics Conference in Las Vegas, following the United Kingdom, seven other univera final oral presentation by each finalist sities were finalists in the 2017 competiteam. Finalists were selected by a panel tion including Drexel University, National of industry and academic experts based University of Singapore, Özyeğin Univeron each teams’ use of the full analytics sity (Turkey), Université catholique de process, from framing the problem to Louvain Team 1 (Belgium), Université A NA L Y T I C S

M A Y / J U N E 2 017

NE W S M AK E R S

catholique de Louvain Team 2 (Belgium), University of Cincinnati and University of North Carolina at Chapel Hill. For more information about the competition, click here. STUDY: SHARING SOCIAL RESPONSIBILITY PRODUCES SURPRISING RESULTS Firms sharing social responsibility for the social good with customers is generally seen as a win-win – more patronage from socially responsible customers and larger benefits to society. A forthcoming study in the INFORMS journal of Marketing Science, a leading academic marketing journal, however, questions the premise. The study finds that when a firm shares social responsibility with customers by asking them to “pay what you want,” promising a certain percentage of revenues to be donated to charity, consumers respond to whether firms give, but very little to how much they give. A firm only needs to donate very little for customers to open their wallet – a win for firms, but not for charities and society. The study, “Signaling Virtue: Charitable Behavior under Consumer Elective Pricing,” authored by Minah Jung (NYU), Leif Nelson (University of California, Berkeley) and Uri and Ayelet Gneezy (University of California, San Diego), examines consumer behavior 30

A N A LY T I C S - M A G A Z I N E . O R G

under the broad umbrella of “shared social responsibility” – where firms and consumers take joint responsibility for the social good. They operationalize shared social responsibility creatively as a variant of “pay what you want pricing,” at a major supermarket retailer in which more than 2,700 customers were offered the option of how much they would pay for the retailer’s reusable shopping bag, when a certain portion of their payment goes to a charity. The surprising finding that customers are very sensitive to whether a portion of their payment goes to charity, but seemingly insensitive to how much goes to charity, has critical implications for the design of shared social responsibility programs. In the field experiment, customers paid more than twice as much for a reusable shopping bag when told that 1 percent of their payment would go to charity relative to when nothing would be offered to charity. But they did not pay much more when donations increased to 50 percent, 99 percent or even 100 percent. A little charity goes a long way; a lot more does not go any further. Digging deeper into this surprising behavior, the researchers found that consumers felt the same level of “warm glow” – the emotional happiness from having done a good deed – irrespective of how much of their money went to a W W W. I N F O R M S . O R G

Apply for 2017

charity. “Consumers feel about the same whether 1 percent or 99 percent of their payment went to charity,” said lead researcher Minah Jung. The authors dub the pattern that consumers are sensitive to whether firms give to charity, but not how much as “scope insensitivity.” Scope sensitivity sounds anodyne, but it is by no means innocuous. Notes Ayelet Gneezy, “It gives firms perverse incentives in how they frame their corporate social responsibility programs. Offering a minimal

contribution can increase profit dramatically. But as the charitable contribution increases consumers don’t give more, so profits go down. The most profitable strategy for the firm is to give to charity, but the smallest possible amount.” Nelson cautions, “Sharing social responsibility with one’s customers sounds like a sure multiplier for the social good. Not so fast. When all customers care is for the warm glow of giving, sharing responsibility with them may not be the promised manna for the social good.” ❙

Apply to win this prestigious practice prize that rewards professionals who devise innovative analytical methods, utilize those methods in a verifiably successful O.R./analytics project, and describe their work in a clear, well-written paper. Two-page abstract is due by May 1, 2017. This top INFORMS practice prize spans all O.R. and analytics disciplines and application fields. Any work presented in an INFORMS section or society practice-oriented competition is eligible as long as the work did not result in a published paper. The Wagner Prize competition is high-profile, with its own track at the INFORMS Annual Meeting. Presentations are widely distributed via streaming video. Finalist papers are published as a special issue in INFORMS respected practice journal Interfaces. Last year’s competition was held at the INFORMS Annual Meeting, November 13-16, 2016, in Nashville, Tennessee. The first-place prize will be awarded to Mikael Rönnqvist, Gunnar Svenson, Patrik Flisberg, and Lars-Erik Jönsson at the Edelman Gala during the April 2017 Conference on Business Analytics and O.R. in Las Vegas, Nevada. Don’t miss your chance to win this illustrious award for 2017.

Daniel H. Wagner

www.informs.org/wagnerprize

A NA L Y T I C S

M A Y / J U N E 2 017

DATA- D RIVE N DE C I S I O N S

Service as a Software: A new paradigm Leveraging the interconnectedness of business problems to accelerate better decision-making.

BY DEEPINDER DHINGRA

the short span of a few decades, the world of data-driven decisions has gone through a significant transformation at a bewildering speed. The cause can be linked to the hastening change in the business environment, as well as the rise and proliferation of connected devices, which continue to yield enormous amounts of data.

A N A LY T I C S - M A G A Z I N E . O R G

Software has been a boon to enabling and scaling analytics for decision support in large organizations. But, the two main paradigms of software development â&#x20AC;&#x201C; packaged software that enables scale for repeatable, well-defined problems and the traditional services model where we develop custom solutions to solve specific business problems â&#x20AC;&#x201C; have limitations; that is, a lack of flexibility in

W W W. I N F O R M S . O R G

Paradigm shift: From software as a service to service as a software. Photo Courtesy of 123rf.com | iqoncept

the first instance and the inability to scale in the second. To address the limitations of traditional software development (see

Figure 1) a new paradigm is required: service as a software, where man and machine come together and work in harmony. Businesses will need this

Figure 1: Software or services?

A NA L Y T I C S

M A Y / J U N E 2 017

S ERVIC E AS A SO F T WA R E

“Iron Man” model – where decision scientists don exoskeletons of software and tools to attack big data (and small!) – to find new ways of solving business problems. THE MORE THINGS CHANGE . . . When we look at business problems and classify them by their underlying nature, logic and math, it turns out that many are similar. For instance, when providing a recommendation, which occurs across many business functions, the process still involves making some

choices from a universe of possibilities to improve the probability of a relevant message and drive consequent action. Examples of this include: • What offers do I suggest to a customer in a direct marketing campaign? What channel should I use to communicate with them? • What products should I display on the home page of my website for a returning customer? Or on the shelf of my store? • What additional services should I offer to someone who has bought

Figure 2: How a business problem is similar across different industries or applications.

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

or subscribed to one or more of my existing services? TOOLS TO ACCELERATE BETTER DECISION-MAKING When designing software for solving problems, although it is easy to find techniques and algorithms, there is a need to apply an intelligence layer and framework on top of these algorithms that can adapt to the context and content of different use cases. Today, the differentiation is no longer just in the techniques and algorithms, but in how they are configured and stitched together to solve business problems. Therefore, an adaptive solution framework we call Meta-software that addresses different business problem classes where the relevant techniques are wrapped in intelligent workflows and selection classes – including recommendation, classification, forecasting, segmentation, attribution, optimization, etc. – is essential to enabling better and faster solutions. When designing software for solving complex business problems, it is easy

enough to find algorithms for problems of classification, pattern recognition, optimization, recommendation and so on. However, one cannot ignore human input in problem definition, solution design and decision support. An example would be product recommendations by a wealth manager for his or her well-heeled clients. Would you leave it completely to software to make the recommendations on a matter of money? Doubtful. When it comes to muddy and fuzzy business problems, augmented intelligence still performs better than artificial intelligence. In today’s business world, problems are becoming more granular and more interconnected. By bringing man and machine together to blend heuristic and algorithmic solutions and preserve the flexibility of customization, business leaders can improve the efficiency of software development and accelerate the deployment of business solutions. ❙ Deepinder Dhingra is an apprentice leader at Mu Sigma.

Request a no-obligation INFORMS Member Benefits Packet For more information, visit: http://www.informs.org/Membership

A NA L Y T I C S

M A Y / J U N E 2 017

MO B IL E E COSYST E M T RA N S FORMAT ION

From CSP to DSP Text mining will play a pivotal role in the transition.

BY (l-r) SOMNATH DE, SAIBAL SAMADDAR AND UPASANA MUKHERJEE

he ubiquitous availability of high-speed Internet and increasing smartphone ownership are at the forefront of the new era of digital transformation. The entire mobile ecosystem is experiencing significant growth, including the following projections: • There will be 5.8 billion smartphones users by 2020 as compared to the 2.6 billion at the end of 2015 [1]. • Data traffic will grow by a compound annual growth rate (CAGR) of 49 percent by 2020 [2].

A N A LY T I C S - M A G A Z I N E . O R G

• Mobile Internet penetration will reach 60 percent by 2020 from 44 percent in 2015 [2]. However, despite the growth in the overall mobile ecosystem, communications service providers (CSPs) across the world are facing saturated markets, experiencing slow subscriber growth and stagnated revenue as their core services are being increasingly commoditized by new Internet giants and over-the-top (OTT) content players. Operators are constantly striving to keep up with the rapidly changing preferences

W W W. I N F O R M S . O R G

Evolutionary journey of the telecommunication industry: from a “smartphone era” to a “hyper-connected world.” Photo Courtesy of 123rf.com | everythingpossible

of customers who rely more on recommendations from friends, colleagues, acquaintances and online customer opinions across various social media sites than normal corporate marketing messages. This requires a paradigm shift in how CSPs engage with their customers. The evolutionary journey of the telecommunication industry from a “Smartphone era” to a “hyper-connected world” is illustrated in Figure 1. The new business models of individual telcos will be the key enablers in this pathbreaking journey. As CSPs transition from a conventional service provider to digital service providers (DSPs) who provide the functional platform that enables the customers to

A NA L Y T I C S

go to the market (e.g., using self-service mobile payment apps to pay utility bills generated by smart meters), they need to move toward a sustainable business model. It will also be imperative for them to prioritize business requirements in order to identify the key use cases that will reap significant benefits while at the same time being relatively easier to implement. Following are some of the use cases that can leverage enormous amounts of unstructured text data from social media, call center data, browsing histories, log files, network analyzer, etc. to derive actionable insights: • real-time congestion and customer offload management;

M A Y / J U N E 2 017

MO B IL E E COSYST E M T RA N S FORMAT ION

Figure 1: Road ahead for the communications industry.

• real-time personalized offers based on browsing history, device, location, live interactions; • product/service innovation, e.g., payment banks, mobile money, etc., and product/service pricing; • preventive action on network failure; • Capital expenditure and operating expenses (CAPEX/OPEX) optimization using network function virtualization (NFV) and softwaredefined networking (SDN); and • call center, workforce and inventory optimization. Thus, moving forward, operators need to move toward a disruptive business

A N A LY T I C S - M A G A Z I N E . O R G

model and chalk out a business strategy for an enterprise-wide digital transformation with focus on the following three key areas: 1. digital customer journey map, which succinctly portrays the rapidly changing expectations and experiences of customers across multiple channels/touchpoints; 2. enterprise-wide digital adoption, e.g., virtual office, virtual stores and virtual retailers; and 3. CAPEX/OPEX optimization: Operator CAPEX is on the rise, so this is the key area that could differentiate a successful DSP from an unsuccessful one.

W W W. I N F O R M S . O R G

http://meetings2.informs.org/wordpress/analytics2017/analytics-career-fair/

MO B IL E E COSYST E M T RA N S FORMAT ION

Operators can efficiently plan their network design by segmenting customers according to their daily travel plans. This will help them evaluate and update their geographical networks to optimize the network spend and provide better customer service. This would significantly increase their customer engagement, reduce the cost of network deployment and reduce CAPEX. RE-DEFINING THE DIGITAL CUSTOMER JOURNEY The key to increasing customer engagement in the age of digital transformation is obtaining a comprehensive view of a customer by creating a digital customer journey map that helps operators pitch personalized product/services to their customers. The first step in developing a customer journey map is developing a “persona” for the digital customer to map his or her expectations and experiences at each stage of his or her life cycle. Insights generated from this digital persona about specific customer traits, positive/negative experiences, browsing history, online behavioral patterns/ trends can be incorporated into cross sell-up sell/customer retention models to identify the various digital touchpoints (website, email, social media, mobile) through which contextual offers 40

A N A LY T I C S - M A G A Z I N E . O R G

can be pitched to more profitable customers. This will also enable the operators to optimize the marketing spend on campaigns. Let’s illustrate how the digital customer journey can be re-defined by operators with the help of a scenario: Mr. Sharma has two postpaid connections, and he is an active user of Internet and mobile banking. His wife also has a postpaid connection in her name. She has been browsing websites on international holiday destinations and actively liking Facebook travel/tourism pages. While a CSP pitches international roaming packages to his wife based on her browsing history, a DSP creates an exhaustive, comprehensive view of customer, i.e., the persona of Mr. Sharma, by incorporating household, account, transactions, social media feeds, positive/negative sentiments, etc. to determine from Mrs. Sharma’s web browsing history that the entire family is going on an international vacation and identifies personalized cross-sell up-sell opportunities, e.g., selling travel insurance through mobile banking apps. A successful DSP will also incorporate insights based on customer-customer relationships and household information in their product recommendations. W W W. I N F O R M S . O R G

Mr. Chopra works in the same department as Mr. Sharma and is also likely to take an international vacation. Here the DSP can cross-sell international roaming offers, international data pack offers and travel insurance to Mr. Chopra as well. TECHNOLOGY APPROACH

Figure 2: Analytics maturity level of CSPs.

Following are some of the key tools and technologies that will be used in the transition from a CSP to a DSP in the next few years: Natural language toolkit. NLTK provides easy-to-use interfaces for building Python programs to work with human language data. It provides more than 50 corpora and lexical resourcesÂ such as WordNet, along with a suite of text processing libraries for classification, tokenization and stemming, as well as tagging, parsing and semantic reasoning. Machine learning. Machine learning algorithms operate by building aÂ modelÂ based on inputs and using that to make predictions or decisions, rather than solely obeying explicitly programmed instructions. Machine learning algorithms can be divided into two main groups: supervised learners that are used to construct A NA L Y T I C S

predictive models and unsupervised learners that are used to build descriptive models. While a supervised learning algorithm aims to unearth the relationship of other variables with the target variable thereby predicting a target of interest, in an unsupervised learning algorithm there is no target to learn and no single variable is more important than others. As CSPs are increasingly adopting an enterprise-wide digital transformation strategy, they are moving toward maturity level 3.0, i.e., content analytics that includes sentiment analytics using NLP, text analytics and artificial intelligence (AI). While organizations in maturity level 2.0 use data-driven insights from their conventional analytical models to take business decisions, a transition to level 3.0 requires incorporation of insights from the vast amounts of unstructured M A Y / J U N E 2 017

MO B IL E E COSYST E M T RA N S FORMAT ION

internal/external data in the existing analytical models to understand the interrelationships between key drivers better, identify key pain points and get an integrated view of their business. This will enable them to arrive at more efficient and sustainable business decisions compared to level 2.0. Moving beyond level 3.0, organizations can use the text data for cognitive analytics and robotic process automation (RPA), e.g., extracting competitors’ pricing data from their websites. CONCLUSION The path to becoming a successful digital service provider from a communication service provider is fraught with the increasing challenge from OTT players and new startups such as Uber, Netflix, Spotify, Airbnb, Skype, which with their disruptive platform-based business models are putting CSPs under tremendous pressure by targeting their core services and consuming the bandwidth. It will be imperative for the operators to respond to the challenge by launching innovative services, particularly in the areas of mobile money and machine-to-machine (M2M) services, e.g., launching new voice calling, messaging apps and own content/video platforms. NLP, and text analytics, machine learning and RPA will prove to be the key pillars to 42

A N A LY T I C S - M A G A Z I N E . O R G

this disruptive transformation in the next few years. This, along with building a sustainable partner ecosystem by providing data as a service and opening up alternate monetization avenues, will significantly improve both the top line along with the bottom line and enable the operator to maintain a competitive advantage over other players in the long run. ❙ Somnath De is a director in the Data & Analytics practice of KPMG, India, where he leads the D&A initiatives. He has more than 18 years of consulting experience and has undertaken multiple data science engagements for leading organizations across multiple industry sectors. Saibal Samaddar (saibalsamaddar@kpmg.com) is an associate director in the Data & Analytics practice of KPMG, India. He has more than 10 years of consulting experience in data and analytics while working for various telecom clients. He holds a master’s degree in electrical engineering. Upasana Mukherjee is a consultant in the Data & Analytics practice of KPMG, India. She holds a postgraduate diploma in management from Indian Institute of Management, Indore, and has worked in multiple advanced analytics and data science engagements spanning multiple industries. EDITOR’S NOTE: An earlier version of this article was first published on AIM. REFERENCES 1. GSMA Intelligence, 2016, “The Mobile Economy.” 2. China Academy of Information and Communications Technology (CAICT), GSMA Intelligence, 2016, “Mobile Operators: The Digital Transformation Opportunity,” June 2016.

W W W. I N F O R M S . O R G

Franz Edelman Award Winner

Daniel H. Wagner Prize Winner

Holiday Retirement

Université Laval, SDC, & The Foresty Research Institute of Sweden

“Senior Living Rent Optimizer (SLRO ), an Innovative Revenue Management System, Achieves Significant Revenue Gain” TM

“Calibrated Route Finder - Social, Safe, Environmental and Cost-Effective Truck Routing”

INFORMS Prize Winner

The Walt Disney Company

U.S. Air Force

UPS George D. Smith Prize Winner

INFORMS O.R. & Analytics Student Team Competition Prize Winner

Operations Reseach Program, United States Air Force Academy

University of Warwick, Nottingham University, & Cardiff University

Awarded at the 2017 Edelman Gala in Las Vegas, Nevada

For more information on the conference, visit http://meetings.informs.org/analytics2017

B U S IN E S S TO B U S I N E S S

Lessons learned from a decade of B2B eCommerce Four common points of failure in B2B eCommerce initiatives, including ways to avoid re-platforming, massive customization and other costly misfires.

BY STEVE SHAFFER or mid-size manufacturers and distributors, and many other B2B (business-tobusiness) organizations for that matter, eCommerce has become a major challenge for the operation. Despite the early promise of added sales and increased efficiencies, these companies are often experiencing what can only be described as a failure to meet the initial goals and objectives of the eCommerce system itself.

A N A LY T I C S - M A G A Z I N E . O R G

At some level, these project misfires are often attributed to factors such as poor rates of adoption after an initial launch, a platform that canâ&#x20AC;&#x2122;t map to complex B2B processes, and/or a need for major and expensive customizations. Although these factors certainly impact the success of the system, the actual issue stems from an underestimation of the differences between B2C (business-tocounsumer) and B2B commerce. Adding fuel to the fire is a common message of W W W. I N F O R M S . O R G

B2B organizations operate in a vastly different manner from B2C when it comes to the buyer’s journey. Photo Courtesy of 123rf.com | alexmillos

“convergence” from B2C eCommerce vendors eyeing a huge opportunity projected by many analysts to be more than $1 trillion by the year 2020. The fact is that B2B organizations operate in a vastly different manner from B2C when it comes to the buyer’s journey. Behind common functions such as user navigation, shopping carts and menu structures lies a complex set of processes, business rules and integration requirements that not only vary tremendously from a B2C buying scenario, but can even be unique to organizations within the same industry. Trying to customize a B2C retail eCommerce engine to a B2B commerce environment is the epitome of the “square peg in a round A NA L Y T I C S

hole” metaphor. Most of the time, it simply doesn’t work. The good news is that after several years of experience building eCommerce systems specifically for B2B, we’ve learned the functionality and capabilities that can be standardized and the processes that typically require customization. We can use these lessons to build stronger, more efficient manufacturing and distribution eCommerce systems. However painful it may seem, we need to examine the major points of failure many B2B initiatives have already faced. In my experience, there are usually four common areas where mid-size manufacturers and distributors struggle when it comes to designing, implementing and M A Y / J U N E 2 017

B 2 B eCOM ME RC E

maintaining their eCommerce systems. These four areas typically map to several different scenarios: 1.The need for massive, expensive platform customization. Too many software decisions are made on the easy promise of an “out of the box” solution from a vendor with an eCommerce solution based on a B2C retail engine. Unfortunately, those operating within the retail industry are not fully aware of the many complexities surrounding B2B operations, particularly those within the ever-changing manufacturing and distribution industries. Although there are similarities between B2B and B2C eCommerce requirements at a design level, what most retail eCommerce vendors fail to understand is that the underlying purpose of B2B eCommerce is different from that of B2C. B2C eCommerce is meant to increase sales. The goal of B2B eCommerce, however, is to maximize efficiency in a complex world of “many to many” relationships between buyers and sellers. The system must acknowledge and handle the fact that B2B is driven by the need to connect complex business processes and data from the back office in order to create a more robust, user friendly experience across the enterprise. Those B2B business processes and data must drive that experience in 46

A N A LY T I C S - M A G A Z I N E . O R G

order for the technology and the project itself to succeed. B2B eCommerce is about making it easier for people and customers to engage and do business. eCommerce engines built for retail simply cannot handle this complexity without expensive customization efforts, as they are built to encourage more sales. Having said that, it is important to point out that B2C shoppers and B2B buyers do share some common needs when using an eCommerce system. From a purchasing standpoint, both require a user experience that is familiar from a consumer perspective, whether it’s incorporating best practices around menu structures and labels, or using a typical shopping cart icon. The experience has to be relatable to the individual and that requires using common consumer navigation standards. Yet the ability to seamlessly navigate a shopping cart experience is where the convergence of B2B and B2C ends. 2. A rush to launch. Mid-size manufacturers and distributors often adopt a sense of “transform or die,” which creates added pressure to not only the development cycle, but also the communication methods surrounding the new eCommerce capabilities. Adoption within B2B eCommerce can be a tricky, and at times frustrating process. Sales may feel that they’re now “out of the loop” W W W. I N F O R M S . O R G

in terms of the buying cycle. Customer service representatives may view the eCommerce site as just another channel they have to manage. Buyers may not feel comfortable or well-educated in terms of gaining maximum value from the system. Both internal and external stakeholders, from existing and new customers, to employees, channel partners and others, need the value proposition of the new eCommerce not only communicated, but marketed to them. This includes robust training components and highly developed help sections, not to mention building a bridge between field representatives and the online system. The lack of strong communications before, during and after the launch of the new eCommerce system creates a major point of failure all on its own. Key stakeholders must be identified at the beginning of the project, and communications requirements must be outlined and begun well before the launch of the system. Anything that is related to educating internal staff, partners and customers regarding the capabilities of the new system such as required training components, common visual and messaging assets, and other critical communications tools, needs to be identified during the analysis phase, not during or after implementation of the new B2B eCommerce system. A NA L Y T I C S

3. Poorly designed integration with enterprise applications. B2B commerce requires strong integration between the organizationâ&#x20AC;&#x2122;s critical backend processes and software, and the eCommerce system. In most cases, the complexities of the B2B operation require communication between sophisticated ERP (enterprise resource planning), fulfillment, and other enterprise applications. Those applications are often ill-equipped to integrate with a B2B digital experience, which involves much more than just adding a shopping cart. A strong B2B eCommerce solution must unify the entire commerce environment and support a wide range of buying scenarios, from transactions that are 100 percent self-service to those that are almost completely sales-supported. Acknowledgement of this requirement often comes too late in terms of the eCommerce launch, and is one of the reasons we see so many initial B2B eCommerce systems re-platformed within a year of launch. Unlike B2C eCommerce, there is a myriad of business and operation scenarios that are unique to the buying cycle of B2B commerce. In fact, the differences between B2B and B2C commerce often reflect the nature of the stakeholders involved. The relationship is many to many: many solutions, many buyers and influencers, and M A Y / J U N E 2 017

B 2 B eCOM ME RC E

many different contractual agreements. Accounting or payables departments may want access to download historical invoices or pay outstanding balances. A salesperson may need specific information about a buyer prior to a meeting. Fulfillment departments may require information on multiple shipping locations and ship-to contacts. Each of these complex scenarios, and others, needs to extract and update data from enterprise systems, often in a real-time manner. Enterprise application integration requirements should be identified and taken into account during the software selection process, not afterwards. And certainly, they cannot be based on the promises of an eCommerce software salesperson that it only requires a simple “plug in.” 4. Burdensome, multi-application environments. Mid-size manufacturers and distributors often try to solve specific needs by purchasing individual applications for each unique requirement of the B2B commerce environment. In addition, solutions primarily built for B2C usually include features that result in more overhead, but simply don’t translate to B2B needs. This scenario creates a complex web of tools and solutions that cannot easily communicate with each other or the organization’s

A N A LY T I C S - M A G A Z I N E . O R G

enterprise systems. Many of these applications are cloud-based, making integration even more complex. In some cases, the purchases and implementation have been made entirely outside of the IT department, turning support into an ongoing nightmare. A mashup of applications cobbled together as needs are identified results in the opposite of an efficient, unified eCommerce environment, one that is also destined for failure. The rush to provide a usable eCommerce system, poor communication and solutions chosen without enough planning and analysis can all create massive dysfunction for the organization as well as the development team attempting to launch a successful B2B eCommerce system. The answer to creating a unified B2B commerce environment lies in selecting a platform that can be easily customized to meet the unique needs of your organization without requiring massive customization or the purchase of extraneous, additional applications. In other words, choosing and customizing software that, where eCommerce is concerned, is specifically built for the complexities of the B2B world. ❙ Steve Shaffer, CEO of Insite Software (“Built for B2B,”), has more than 20 years of experience as a technology entrepreneur and executive.

W W W. I N F O R M S . O R G

GOLD SPONSORS

SILVER SPONSORS

https://tinyurl.com/analytics2017-sponsors

DATA AN A LYS I S & MODE LI N G TOOLS

Leveraging big data for better business decisions

BY DAVE OSWILL orking with big data is fast becoming a key step in the process of scientific discovery and engineering. This is happening as technologies such as smart sensors and the Internet of Things (IoT) are enabling vast amounts of detailed data to be collected from scientific instruments, manufacturing systems, connected cars, aircraft and other systems. Significant value lies in this data as it may show important physical phenomena

A N A LY T I C S - M A G A Z I N E . O R G

or provide information on the operating environment, efficiency and health of a system. With the proper tools and techniques, this data can be used for rapid scientific discoveries and to incorporate more intelligence into products, services and manufacturing processes. For many engineers and scientists, who must consider implementing these data-driven solutions into their enterprise, this can be a daunting process due to the systems commonly used to W W W. I N F O R M S . O R G

The first challenge in working with big data is gaining access to large data sets that may be stored in various types of systems. Photo Courtesy of 123rf.com | illustrator

store, manage and process this valuable data. Software analysis and data modeling tools have been enhanced with new capabilities that allow engineers and scientists to use familiar syntax and functions to unlock the complexity of the data they are collecting to make more effective design and business decisions.

databases (SQL/NoSQL), IoT data aggregators and data historians, to

ACCESSING LARGE SETS OF DATA The first challenge in working with big data is gaining access to these large data sets that may be stored in various types of systems ranging from Figure 1: Access a wide range of big data. shared file systems, A NA L Y T I C S

Source: The MathWorks, Inc.

M A Y / J U N E 2 017

B IG DATA

distributed platforms such as Hadoop. The data may consist of delimited text, spreadsheets, images, videos and other proprietary formats. To effectively work with this data, engineers and scientists need scalable tools, such as MATLAB, that can provide access to a variety of systems and data formats. This is especially crucial in cases where more than one type of big data platform or data format may be in use.

result. There are certain capabilities that simplify this exploration process, making it easier for engineers and scientists to observe, clean and effectively work with big data, including: Summary visualizations, such as binScatterPlot (Figure 2) provide a way to easily view patterns and quickly gain insights. Data cleansing removes outliers and replaces bad or missing data to ensure a better model or analysis. A programmatic way to cleanse data enables new data to EXPLORING AND PROCESSING LARGE SETS OF DATA be automatically cleaned as itâ&#x20AC;&#x2122;s collectAfter accessing the data and before ed. (Figure 3). creating a model or theory, itâ&#x20AC;&#x2122;s important Data reduction techniques such as to understand what is in the data, as it principal component analysis (PCA) may have a major impact on the final help to find the most influential of your data inputs. By reducing the number of inputs, a more compact model can be created, which requires less processing when the model is embedded into the products or services. Data processing at scale enables engineers and scientists to not only work with large sets of data on a desktop workstation, but use their analysis pipeline or algorithms on an enterprise class system Figure 2: binScatterPlot in MATLAB. such as Hadoop. The abilSource: The MathWorks, Inc. ity to move between systems 52

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

without changing code greatly increases efficiency. CREATING A MODEL Imagine collecting years’ worth of data. What is valuable in this data? Often, in order to analyze the data and create an intelligent and pre- Figure 3: The two types of machine learning methods provide different algorithms tailored for different problems. dictive model, machine learnSource: The MathWorks, Inc. ing is required. INCORPORATING BIG DATA FOR Machine learning uses computational REAL-WORLD SOLUTIONS methods to “learn” information directly There are a number of platforms availfrom data without relying on a predeable to IT organizations for storing and termined equation as a model. It turns processing of big data that fall into two out this ability to train models using the categories: 1) batch processing of large, data itself opens up many use cases for historical sets of data, and 2) real-time or predictive modeling such as predictive near real-time processing of data that is health for complex machinery and syscontinuously collected from devices tems, physical and natural behaviors, enBatch applications, such as Spark or ergy load forecasting and financial credit MapReduce, are commonly used to anascoring. lyze and process historical data that has Machine learning is broadly divided been collected over long periods of time or into two types of methods, supervised across many different devices or systems. and unsupervised learning, each of which These applications are typically used to contains several algorithms tailored for look for trends in data and develop predicdifferent problems. tive models. • Supervised learning uses a training Streaming applications that process in data set which maps input data to real- or near-real time, such as Kafka, may previously known response values. be coupled with a predictive model to add • Unsupervised learning draws more intelligence and adaptive capabilities inferences from data sets with input to a product or service such as predictive data that does not map to a known maintenance, optimizing equipment fleets, output response. and monitoring manufacturing lines. A NA L Y T I C S

M A Y / J U N E 2 017

B IG DATA

Incorporating models into products or services is typically done in conjunction with enterprise application developers and system architects, but this can create a challenge. Developing models in traditional programming languages is difficult for engineers and scientists, while recoding models can be time-consuming and error prone, especially if the models require periodic updates. To alleviate this issue, enterprise application developers should look for data analysis and modeling tools that are familiar to their engineers and scientists, while also providing production ready tooling such as application servers and code generation for deploying models

into their applications, products and services. To truly take advantage of the value of big data, the full process â&#x20AC;&#x201C; from sourcing data to developing analytical models to deploying these models into production â&#x20AC;&#x201C; must be supported. IT managers and solution architects can use modeling tools to enable the scientists and engineers in their organizations to develop algorithms and models for smarter and differentiated products and services. Simultaneously, the organization is being enabled to rapidly incorporate these models into its products and services by leveraging production-ready application servers and code generation capabilities that are found in these tools. The combination of a knowledgeable domain expert who has been enabled to be an effective data scientist, along with an IT team capable of rapidly incorporating their work into the services, products and operations of their organization, makes for a significant competitive advantage when offering the products and services that customers are demanding. â?&#x2122;

Figure 4: Integrating models with MATLAB. Source: The MathWorks, Inc.

A N A LY T I C S - M A G A Z I N E . O R G

Dave Oswill is product marketing manager at MathWorks, developers of MATLAB.

W W W. I N F O R M S . O R G

PLATINUM LEVEL

EXECUTIVE LEVEL

SCIENTIFIC SOLUTIONS

YEARS

CHAMPION LEVEL

https://tinyurl.com/analytics2017-sponsors

CO RPO RATE P RO F I LE

Praxair: Making our planet more productive

BY LARRY MEGAN (left) AND KRISTIN BRUTON

or more than 100 years, Praxair has taken something as fundamental as air and turned it into ways to make food taste better, processes operate cleaner and breathing easier â&#x20AC;&#x201C; in short, to make all our lives better by fulfilling our mission of making our planet more productive. As a Fortune 300 company, we develop products and technology that impact more than 20 different industries. You see our work every day, often without

A N A LY T I C S - M A G A Z I N E . O R G

realizing it. We provide oxygen for respiratory therapy, hydrogen to help purify crude oil into gasoline, carbon dioxide to add the fizz in our beverages and argon to enhance the robotic welding systems that build new automobiles. We provide products to our customers through three primary supply modes: large process plants, cryogenic liquid and packaged gases. Large process plants: For our largest customers, such as refineries and

W W W. I N F O R M S . O R G

Praxair plant in Utah with argon, nitrogen and oxygen storage tanks. Photo credit: Ted Kawalerski

steel mills, we design, build and operate large process plants adjacent to their facilities. These plants act as a vital utility to those customers, similar to electricity and water, and provide an uninterrupted supply of industrial gases to support their operations. As a result, a high degree of automation and data analytics is needed to ensure that we continuously maintain safe, reliable and efficient operation. Cryogenic liquid: Medium-volume customers, such as hospitals and universities, typically have liquid storage

A NA L Y T I C S

located at their facility, which is then used to provide product throughout their operations. We monitor their inventory in real time and then deliver product using our tanker truck fleet, all without the customer needing to place an order. This proactive, vendor-managed service model provides high reliability for our customers while enabling us to effectively manage our costs. As with large process plants, this service model requires a variety of analytics, from optimizing assets to scheduling daily deliveries.

M A Y / J U N E 2 017

CO RPO RATE P RO F I LE

Packaged gases: We sell a wide variety of smaller packaged gases to laboratories, hospitals and other customers. These products can range from a cylinder of nitrogen to specialty gas mixtures needed for emissions testing, advanced manufacturing and semiconductor fabrication. We also offer a variety of services such as embedded regulators and telemetry. We also own the cylinder assets, which helps ensure continuous supply to our customers. Packaged gases are a very transactionintensive business with many distinct products, which leads to many opportunities to use analytics to manage the supply chain, understand margins and better target the sales force. CULTURE DRIVES STRATEGY The core of Praxair’s business model is a culture of productivity – continuous improvement across all aspects of our business, including how we operate our plants, how we deliver our products, how we manage our business processes and how we collaborate with our customers. This culture pervades all levels in the organization. We have a longstanding history of effectively deploying tools such as Lean and Six Sigma, but as our productivity model has matured, it has become much more challenging to find the low-hanging fruit 58

A N A LY T I C S - M A G A Z I N E . O R G

in productivity. As a result, our current efforts to sustain our productivity momentum are focused on three key areas: innovation, digitalization and advanced analytics. Praxair’s Global Analytics team plays a key role in this evolving productivity strategy. We support Praxair’s mission by making our people more productive – developing the decision support systems needed to enable our business clients to make better decisions faster. We provide a wide variety of systems that improve our different business functions regardless of where they are on the analytics maturity curve, from descriptive and diagnostic systems that help Praxair’s businesses more effectively manage and share their data, to predictive and prescriptive systems that optimize operations, logistics and sales. Our wide variety of skills enables us to meet our customers wherever they are on the continuum of analytics, and positions them to meet ever more challenging business needs. GLOBAL REACH, LOCAL IMPACT In industrial gases manufacturing, the majority of the products sold are manufactured roughly 200 miles of where they are produced. This high degree of localization leads to a business model where P&L (profit and loss) W W W. I N F O R M S . O R G

responsibility lies with geographical business units to ensure that the company best meets the needs of its local customers. The Global Analytics team reflects this strategy â&#x20AC;&#x201C; we have a hybrid global/local organization designed to balance centralized development while embedding key technical capabilities within the Praxair truck on its way to meet the needs of customers. regional business units. Photo credit: Ted Kawalerski The centralized corporate team is organized into Advanced flexibility while being close to business Analytics and Visualization, Business needs. and Supply Chain Optimization and Complementing the corporate teams Smart Operations functions. Its portfolio are local teams in China, India and Mexico. includes simple visualization tools that These local business teams integrate help business leaders effectively mancorporate-level applications into local age data across heterogeneous data needs. Many of our applications require sources so they can get quick insights customization based on differences in into their business performance; comthe local business climate and regulatory plex optimization models to manage statutes. These local teams enable fast strategic, tactical and operational deciand impactful replication of solutions to sions in our supply chains; and sophistithose geographies. These teams, which cated algorithms to manage all aspects have similar skill sets as the corporate of plant operation, from automating team, solve local issues that would startup to maximizing and sustaining previously have been invisible at the efficiency during normal operation. To corporate level. They identify problems ensure alignment with business priand solutions for their local customers orities, we are a shared organization and quickly meet their needs. Often these between R&D and Global Operations, local solutions have been subsequently which allows us to have development leveraged for replication globally. A NA L Y T I C S

M A Y / J U N E 2 017

CO RPO RATE P RO F I LE

Besides this horizontal integration across the business, the entire team works to apply analytics across all business functions including our on-site, merchant liquid and packaged gases businesses, and a broad range of business processes including sales, product management and customer service. Some recent examples of our programs follow. BROAD REACH In our on-site plants, predictable production reliability is critical to our business. We have deployed several programs using advanced analytics to better increase our predictive reliability. Cryogenic air separation plants, which produce oxygen, nitrogen and argon, require compressors driven by large electric motors. The unexpected failure of one these machines is very disruptive to our operations and supply chain. Our programs utilize advanced statistics and visualization to monitor several hundred machines worldwide and alert the operations staff when a machine is at risk for a potential failure. This provides the local team with the insight they need to effectively manage their maintenance programs. Furthermore, the data from all the machines is sent back to a centralized control center so that subject matter experts can assist local operations to manage issues. Future machine learning 60

A N A LY T I C S - M A G A Z I N E . O R G

applications will help us better identify patterns across multiple data sources to better identify at-risk machines. Managing the merchant liquid supply chain is a second example. These liquid plants, while all making the same basic set of products, often vary in capacity and efficiency. As merchant liquid customers may receive shipped product from multiple locations, continuously optimizing this supply chain can be challenging. External factors, such as varying customer demand and time-of-day electricity prices, make the system quite dynamic. Through the development of sophisticated forecasting tools and large mixed integer linear programming models, we can determine optimum operating scenarios on a continuous basis and share these with our logistics and production teams. The sophisticated planning tools, which plan over a weekly time horizon, then guide the operational tools designed for minute-to-minute optimization of the plants and logistics. Finally, on the business side, the team has significantly impacted Praxairâ&#x20AC;&#x2122;s revenue management strategies over the past several years. For example, in our packaged gas business, weâ&#x20AC;&#x2122;ve developed cooperative game theory methods to allocate shared distribution and production costs to determine product cost at the SKU level and service cost at the order level. W W W. I N F O R M S . O R G

Praxairâ&#x20AC;&#x2122;s goal is to give its people the opportunity to work through the entire lifecycle of a problem. Photo credit: Ted Kawalerski

This solution adopts linear programming models to determine the cost to serve for existing customers as well as machine learning algorithms to estimate costs for new customers. The deliverables help us better manage business and productivity decisions. For example, understanding cost differences among different locations provides additional insight into our supply chain optimization. OUR PEOPLE As with any organization, our people are our most valuable asset. To develop A NA L Y T I C S

our staff, our goal from day one is to give them the opportunity to work through the entire life cycle of a problem. This includes working with our business partners on the value proposition, leading a crossfunctional team from initial development through beta testing, and working with the business to deploy tools in the field to bring value. Such opportunities help our people develop strong leadership, communication, business and project management skills, in addition to technical skills. At our core, weâ&#x20AC;&#x2122;re very strong in technical areas such as data visualization, advanced M A Y / J U N E 2 017

CO RPO RATE P RO F I LE

statistics, linear and nonlinear optimization and advanced process control. Our strength is in our breadth as we must be ready to solve a wide variety of business problems from the simple to the complex. We also recognize the need for strong external partners. We have a long history of working with leading researchers in operations research and process systems to develop the platforms for our future innovations. Strong relationships with SUNY-Buffalo, Carnegie Mellon, McMaster University and UT Austin, among others, keep us as at the leading edge of academic research while providing an opportunity for our staff to engage and mentor graduate students. As an example, we have recently been working with several institutions under Department of Energy funding to demonstrate a state-of-theart industrial scale steam methane reforming plant. As a team, we installed a new visual monitoring system, developed a new control strategy to operate the plant, and demonstrated a cloud-based solution for decision deployment. Going forward, we will have

substantial engagement with the DOE’s new Clean Energy Smart Manufacturing Innovation Institute to further Praxair’s mission and drive innovation generally within the manufacturing industry. WHAT DOES THE FUTURE HOLD? This is an exciting time to be engaged in advanced analytics. The digitalization of manufacturing is a key trend for future growth, and we are excited to be at the leading edge of that in the process industries. In the next few years, we will be developing cutting-edge applications that leverage “big data” infrastructure to solve key challenges in machine reliability and fleet safety, using new technologies to deliver real-time information and collaboration capabilities to people in the field, and delivering new tools to help optimize our supply chains on a global scale. While the industrial gases industry is more than 100 years old, there is plenty of room left for innovation. ❙ Larry Megan (Larry_Megan@praxair.com) is director of Praxair Global Analytic, as well as Praxair’s representative to the INFORMS Roundtable. Kristin Bruton is communications manager at Praxair.

Join the Analytics Society of INFORMS

For more information, visit: http://www.informs.org/Community/Analytics/Membership

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

Is the largest association for analytics in the center of your professional network?

IT SHOULD BE. • INFORMS Allows You to Network With Your Professional Peers and Others Who Share Your Interests • INFORMS Connect, the New Member-only, Online Community Lets You Network With Your Colleagues • Unsurpassed Networking Opportunities are Available in INFORMS Communities and at Meetings • INFORMS Offers Certification for Analytics Professionals • Take Leadership Roles to Help Build Your Professional Profile • INFORMS Career Center Provides You With the Industry's Leading Job Board

Join Online Today! http://join.informs.org

CAREER B U I LDE R

Storytelling: The Write Stuff New book explains how to present analytics to a nontechnical audience. Hint: Shorter and clearer is better.

BY DOUGLAS A. SAMUELSON

“T

he battle was already lost in the opening paragraph,” one member of an analytical team recounts. “A federal agency was soliciting congressional support for a multi-million-dollar IT cloud system that would serve several agencies. The technicians invested substantial effort composing papers that explained the concept, operational benefits, customers, timetable, financial savings and cost in hopes of securing needed funds. But the authors had failed to make their case understandable to non-IT people.” Fortunately for this effort, the team member telling the story is Carla D. Bass,

A N A LY T I C S - M A G A Z I N E . O R G

colonel (retired), U.S. Air Force, who was asked to “enhance the congressional readability” of the papers. She completely rewrote the papers and accompanying presentations, working closely with the subject matter experts to ensure her revisions didn’t inadvertently skew the message. Here is an example of how the writing techniques found in her recent book [1] transformed the products, ultimately persuading Congress to fund the project: Before: Currently, there is a multiplicity of non-interoperable collaboration IT tools in use across the various agencies. This not only limits information sharing potential, but also generates increased

W W W. I N F O R M S . O R G

costs. (Italics added to emphasize unhelpful wordiness.) After: Government agencies use many incompatible IT tools for collaboration, a costly practice that impedes sharing information. Analysis: Identify the subject and verb, the first step to transforming verbal mush into a succinct statement. The subject is “government agencies” and the verb is “use.” These modifications eliminate, “There is … in use across the various.” Eliminate useless words. Don’t use words that hog space (e.g., replace multiple with many and non-interoperable with incompatible). Hogging space also includes the practice of using two words when one will suffice. Thus, replace “limits … potential” with “impedes.” And “generates increased costs” with “costly.” Finally, clarify “This …,” a confusing word in the complexity of the preceding sentence. Bass was subsequently asked to compose a series of 90-word elevator speeches on various aspects of this same cloud system: What is an IT enterprise? How does the IT enterprise function? Is the cloud secure? What is the IT enterprise implementation timeline? How do you manage information in the cloud? How do you access information in the cloud? What about information sharing, security and the IT enterprise?

A NA L Y T I C S

“Given the intended audience, the project at stake and constraint on length of the text,” she notes, “every single word mattered!” These elevator speeches were subsequently used in testimony and included in other updates to Congress. Many similar experiences and thankful teammates convinced Bass to write her book. Do analytics professionals really need such guidance? Here’s the opinion of one of her acquaintances, a senior military operations research analyst who teaches in a military graduate school: “My course in combat modeling and simulation was intended for graduate students who were going to become operations research analysts. They took the course as they were preparing their thesis en route to graduation. The point of the course was the use of modeling and simulation to accomplish various Department of Defense studies, and many bright students used models in their research. Since I was known as a stickler about communicating study results, many came to me as thesis advisor or reader. They often got bogged down in getting their work described and were mostly unsuccessful in reaching their intended audience with the results and merits of their research. “In review sessions, we often banged away at their writing which was usually too technically deep and unclear in describing results. The process of revision for clarity

M A Y / J U N E 2 017

STORYTEL L IN G and creating only strong purposeful written descriptions usually produced a shorter and more understandable product. It always worked, and in two years some 25 students turned into O.R. analysts and accomplished study report writers.” The professor continued, “My first attendance at an operations research symposium was marked by one of the senior mentors telling everyone from the podium, ‘No decision was made on analysis that was not understood.’ The room went dead silent as we all tried to get meaning from these strange words. While it took a while to sift through the audience, it was apparent something cerebral had taken place. Several months later I was preparing for my first decision briefing. The presentation was to be given to a non-analyst highranking decision-maker. It dawned on me that the usual analyst thoughts were just not going to work, and these would only be ignored rather than be useful in helping to reach a decision. To this day, I believe mentors’ words, rewriting the briefing, cutting out the colorful mathematical jargon, and clarifying the facts while explaining the merits and disadvantages, got me the ‘thank you very much,’ and ‘I know exactly what I want to do’ at the end of briefing.” GENERAL PRINCIPLES In the book, Bass details 10 basic principles for how to improve writing:

A N A LY T I C S - M A G A Z I N E . O R G

1. Eliminate useless words. 2. Shorter is better; don’t hog space. Shorter words, shorter sentences. 3. Eliminate redundancy. 4. Lead with the basics. 5. Rely on verbs. Instead of “served in a coordinating capacity,” say “coordinated.” 6. Avoid professional jargon, unnecessarily detailed information and phrasing that requires the reader to dig for the message. When you must use a professional term of art, explain it – preferably in a footnote or technical appendix – or cite an explanatory reference. 7. Use tethers; all terms in a phrase should refer back to a clear base. For example, “She composed and translated technical information into layman’s language” leaves the sharp-eyed reader wondering how the person composed information into layman’s language. The reference is unclear. Better: “She composed technical information and translated it into layman’s language.” 8. Be clear: Who does what to whom? Instead of “Significant efforts by the staff went into the creation of a new dining facility,” write “The staff worked hard to create a new dining facility.” 9. Keep the focus. Lead the reader to your conclusion with as few distracting, irrelevant topics as possible. 10. Proofread carefully; your credibility is at stake. Misspellings and

W W W. I N F O R M S . O R G

grammatical errors quickly undercut the effectiveness of your writing. And don’t rely on just the computer’s spelling and grammar checker; many of them are too forgiving of misuses that lead to making you look silly. Begin with a “hook” that quickly engages the reader: What is this about and why should you care? Conclude by repeating and summarizing the most important point and indicating what the reader should do about it. NOT JUST FOR TECHNICAL REPORTS Bass’ approach works for fundraising letters, performance appraisals, awards nominations and resumes, too. She recounted (not in the book) a friend who purchased an historic building to renovate for his home design shop, banquet facility and guest rooms. He had written a letter to solicit investors. She spent two hours rewriting a one-page letter – reorganized it, sharpened the language, added focus and better explained the unique opportunities this building in a country town offered. “As he’d written it,” she said, “the key headline (need money) was buried in the second paragraph, following a rambling opening paragraph. Financial specifics were located at the end of the letter. I grouped these and opened the letter with them.

A NA L Y T I C S

I relocated how funds would be applied from the opening paragraph to after the proffered opportunity to invest. I recast the entire tone, infusing it with focused excitement. Instead of, ‘We need your help in order to open these businesses,’ I phrased, ‘We offer you the opportunity to participate in this singular transformation.’ I added a colorful, enticing, brief description of the surrounding area and activities that will draw clients. I concluded on a strong note of community support: “We hope you will join us and participate in this exciting and prosperous business venture. “He got the $75,000 he needed.” In another example from the book, consider the following rewriting of a colleague’s resume: Before: Assisted the Office of Computer Defense with technical and programmatic support of the portfolio with over $400 million in annual budget. Developed numerous highly technical proposals and presentations through substantial research and analysis of data from multiple sources to inform senior corporate leader and gain further support in delivering improvements to existing technology and the development of new technology aimed at keeping corporate computer networks ahead of the adversary.

M A Y / J U N E 2 017

STORYTEL L IN G After: Developed and presented to senior corporate leaders technical proposals to improve current technology and develop future IT capabilities to better defend computer networks from attacks by both state and non-state adversaries. Supported a portfolio valued at $400 million, annually. Analysis: “Assisted with ... support of” is a boring, unfocused opening that also begs the question, “Assisted how?” Begin each bullet in a resume with a solid, action verb – in this case, “Developed and presented.” Again, write concisely “With over $400 million in annual budget” is more concisely stated as, “valued at $400 million, annually.” Eliminate useless words (italicized). Delete “Substantial research and analysis of data from multiple sources” because this is expected as foundational for proposals and presentations; therefore, these steps should not be mentioned. I can state from my own experience that even very good writers can benefit from a critique by other good writers. For all the times I’ve advised others to emphasize accomplishments rather than capabilities in a resume, I was surprised several years ago when a professional told me I had failed to do that in my own resume. My summary statement began, “Experienced operations research analyst with extensive

A N A LY T I C S - M A G A Z I N E . O R G

background in...” listing methods and techniques. After that rude awakening, it begins, “Proven trustworthy problemsolver with a solid record of successes in...” and a list of application areas. This works much better. I can also state from experience that just reading Carla D. Bass’ book and working through the many instructive examples and exercises in it will make you a better writer, but it is not sufficient. It often takes in-person critiques to help you identify your bad habits, and – again from my experience – the people who most need improvement are often the last to realize it. (This observation applies to more than just writing skills.) Having a few non-technical people review your writing and point out what they don’t follow readily is a fine quality control method. The ancient Jewish sages had it right: The truly wise person is the one who learns from everyone. The book is a good start, though. ❙ Douglas A. Samuelson (samuelsondoug@yahoo. com), D.Sc. in operations research, is president and chief scientist of InfoLogix, Inc., a small R&D and consulting company in Annandale, Va. He is a contributing editor of OR/MS Today and Analytics magazine.

REFERENCES 1. Carla D. Bass, “Write to Influence!” Orlean Press, Marshall, Va., 2016. 2. Personal interview and email correspondence, March-April 2017.

W W W. I N F O R M S . O R G

SPECIAL ADVERTISING SECTION

CLASSIFIEDS

MARVEL At Disney, we're storytellers. We make the impossible, possible. We do this through constantly collecting, organizing and analyzing all forms of data allowing our analysts to bring stories to life and narrate the future of Disney experiences. We are a global, diverse team that drives business results and strategies using advanced analytics. In a rapidly changing consumer and market environment, with ever increasing touch points and data, our emphasis is on leveraging innovative and optimal solutions to provide actionable insights.

The UVA(,r�fsNEf Studios HAS RELEASED

30 RESORT HOTELS IIMl!lO'letltmt

•WALT DISNEYWORLO•RESORT

4 SHIPS av�UISE

• DISNEYLAND,.RESORT

• HONG KONG DISNEYLAND RESORT • SHANGHAI DISNEY RESORT

1100 '

FEATURES

,4,ui/d!:!:_ !Wei

33 000 Roolis DAIL y • DISNEYlANO�RESORT PARIS • DISNEY CRUISE LINE•

�fst..i�p Consumer Products and Interactive Media wci��t.� LARGEST LICENSOt

MFcK�<f Mo43f:1s r! 1

if:$i=i;' S PORTFOLIO OF MOBILE APPS GENERATED 4 6 .

BILLION

MINUTES OF USAGE

THAT'S APPROXIMATELY

34% OF ALL

CHARACTER FRANCHISE IN THE WORLD .

• •

FULL-LENGTH

SPOR TS APP USAGE

JOB OPPORTUNITIES WITHIN ANALYTICS INCLUDE:

BUSINESS ANALYTICS

DATA SCIENCE

DATA ENGINEERING & TECHNOL OGY

CONSUMER INSIGHTS

•

DECISION SCIENCE

CO N FERE N C E P R E V I E W

Ready for Rotterdam: 2017 INFORMS Healthcare Conference Forum for O.R. professionals to come together to optimize health service operations and outcomes.

All over the world, healthcare organizations are being challenged by obstacles related to the aging of the population and other global trends. It is vital that the complex and growing demands in healthcare delivery are tackled swiftly and optimally. With the abundance of data available, operations researchers across the globe can pool resources and share solutions to come up with innovative ideas to remedy shortcomings within current healthcare systems. The 2017 INFORMS Healthcare Conference, to be held July 26-28 in Rotterdam, Netherlands, is the ideal forum for O.R. professionals to come together to optimize health service operations and outcomes. The conference with be led by conference chair Joris Van de Klundert, Institute of Health Policy and Management, Erasmus University, along with program chairs Edwin Romeijn, Georgia Institute of Technology, and Sandra Sulz, Erasmus University Rotterdam. The program is organized into nine tracks of top issues that are impacting the healthcare industry today. These tracks are Disease and Treatment Modeling, Healthcare Data Analytics and Machine Learning, Health and Humanitarian Logistics, Health

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

Picturesque Rotterdam will host the 2017 INFORMS Healthcare Conference on July 26-28. Photo Courtesy of 123rf.com | rudi1976

Information Technology and Management, Health Operations Management, Health Systems in Low and Middle Income Countries, Medical Decision-Making, Personalized Medicine, and Public Health and Policy-Making. When attendees are not engaged in high-level talks they can take part in some of the other events that round out the program for the conference. The INFORMS Health Applications Society sponsors a student paper competition. Students are asked to submit either oral or poster presentations that will be evaluated by leading healthcare scholars on quality, novelty and importance of methodology, contribution to healthcare research and potential for impact on practice. The finalists must present their work during a special session at the conference. There are also poster sessions during the conference that are not related to the student paper competition. These poster session

A NA L Y T I C S

presentations allow authors to present projects that are in the early stages of development, and thus benefit from the interactive critique, suggestions and encouragement from colleagues working in similar areas. The conference will take place at the De Doelen International Congress Centre, in the heart of Rotterdam, which is an emerging world leader in the healthcare and medical industry. A group rate is available at the Rotterdam Marriott, which is linked to the Congress Centre and just a two-minute walk away from Rotterdam Central Station. There are only a limited number of rooms booked at the INFORMS group rate, so make your reservations as early as possible. â?&#x2122; For more information on this conference, including registration, the venue or the program visit: http:// meetings2.informs.org/wordpress/healthcare2017/.

M A Y / J U N E 2 017

CO N FERE N C E P R E V I E W

ISMS Marketing Science Conference Event set for June 7-10 at the University of Southern California.

The 39th Annual ISMS Marketing Science Conference will be held June 7-10 at the University of Southern California in Los Angeles. The ISMS Marketing Science Conference is an annual event that brings together leading marketing scholars, practitioners and policymakers with a shared interest in rigorous scientific research on marketing problems. Topics include but are not limited to branding, segmentation, consumer choice, competition, strategy, advertising, pricing, product, innovation, distribution, retailing, social media, Internet marketing, global marketing, big data, machine learning, choice models, game theory, structural models and randomized control trials. â?&#x2122; For more information and to register, click here.

The University of Southern California in Los Angeles, site of the ISMS Marketing Science Conference. Source: USC

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

2017 INFORMS ANNUAL MEETING OCTOBER 22â&#x20AC;&#x201C;25 | HOUSTON, TEXAS

You are invited to present to 5,000 plus attendees, join intriguing plenary presentations, panel discussions, and tutorials, or submit for one of our numerous oral and poster tracks focusing on operations research and analytics.

IMPORTANT DATES May 15 - Abstract Submission Deadline August 1 - Poster Competition Submission Deadline September 1 - Poster Submission Deadline September 1 - All Presenters Must Register September 29 - Early Registration Deadline

SUBMIT AN ABSTRACT OR REGISTER TODAY http://meetings.informs.org/houston2017

http://meetings.informs.org/nashville2016

FIVE- M IN U T E A N A LYST

Hamilton and duels Dueling has had a long and storied history among mathematicians.

BY HARRISON SCHRAMM

I have, like many others, been caught up in the Hamilton craze. Duels are a central feature of the Broadway show, with three prominently featured, two of which were fatal. Since neither I nor Analytics editor Peter Horner have been able to procure tickets via the normal means, we will have to settle for the next best thing – math about Hamilton. Below is a letter that I imagine written by a statistician to a number theorist who had offended him. Should Peter receive tickets before I do, I may challenge him to a duel. Dear Sir, I am writing to you this evening to express my displeasure over your recent Twitter comments, particularly your insinuations regarding the insignificance of my p-value. Dueling has had a long and storied history among mathematicians. As you are no doubt pondering your impending death, I would like to remind you of the story of French luminary Evariste Galois. When challenged to deadly conflict, he spent his last night on this earth as I imagine you will, preparing his treatise on group theory. It is this treatise for which he is now remembered. The name of the woman involved – if it was truly a woman at all – is forgotten to history. Upon taking the field, he was shot dead by his rival. It would seem that you and your fellow number theorists should keep to your pencils, but the handling of weapons should be relegated to other men.

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

Year

Wounded

Killed

1777

1801

1804

1806

1817

1820

1823

1826

1827

1831

1832

1838

1839

1842

1847

1853

1856

1859

1863

1865

1867

1877

1882

1887

Table 1: Recorded U.S. duels and outcomes.

In any event, I look forward to admiring the brilliance of your contribution post-mortem. I will certainly take pride in having forced the brilliance from your mind before I force the blood from your Figure 1: Fatality by duel year. veins. Now to matters at hand. I have produced a representative list of duels and their outcomes for your consideration (see Table 1).” As you wisely eschew operations with actual data (a policy you should have adopted for actual weapons), I will summarize for you: There is a 70 percent chance that our duel, should it be consummated, will be fatal for at least one party. Should you somehow escape death, you will see that there will be a 22 percent chance of horrific maiming, which I would find satisfying in its own right. You may take some comfort knowing that depending on the year, the fatality of our encounter is variable, with fatality minimized before the U.S. Civil War (Figure 1). I have no explanation for this observation, save that perhaps mangy curs such as yourself showed more forbearance. To summarize, I demand satisfaction. I have the honor to be your obedient servant, H. Schramm ❙

Harrison Schramm (Harrison.schramm@gmail.com), CAP, PStat, is a principal operations research analyst at CANA Advisors, LLC, and a member of INFORMS.

A NA L Y T I C S

M A Y / J U N E 2 017

THIN K IN G A N A LY T I CA LLY

Payload balance

Where to distribute the delivery payload?

The International Space Station requires occasional resupply of food, water and equipment from payloads that are delivered by rockets. Because rockets require their payloads to be balanced, deciding where to place cargo is critical to mission success. The accompanying image shows a delivery payload. There are four cargo sections within the payload labeled A, B, C, D. In order for the payload to be balanced, the total weight of cargo in section A must be equal to that of section D. And the total weight of cargo in section B must be equal to that of section C. Note that section A does not need to equal section B (nor does section C need to equal to section D). There are 15 packages each of varying weights that must be included in the delivery. The weights of the packages in kilograms are as follows: 70, 90, 100, 110, 120, 130, 150, 180, 210, 220, 250, 280, 340, 350, 400. Each section must contain at least three packages and no section can hold more than 1,000 kg. QUESTION: What are the weights of the sections when the payload is balanced? Send your answer to puzzlor@gmail.com by July 15. The winner, chosen randomly from correct answers, will receive a $25 Amazon Gift Card. Past questions and answers can be found at puzzlor.com. â?&#x2122;

BY JOHN TOCZEK

John Toczek is the senior manager of analytics at NRG in Philadelphia. He earned his BSc. in chemical engineering at Drexel University (1996) and his MSc. in operations research from Virginia Commonwealth University (2005). He is a longtime member of INFORMS.

A N A LY T I C S - M A G A Z I N E . O R G

W W W. I N F O R M S . O R G

I S T H E P E R F E C T TO O L F O R MODELERS ACROSS INDUSTRIES If you are not already using GAMS for your optimization, here are some reasons to start right now:

Cutting Edge Modeling Build your models in a language similar to their mathematical description The GAMS language is powerful, yet easy to learn We offer responsive technical support by highly competent experts

Design Principles that make a Difference Switch between platforms, solvers, data sources Integrate GAMS seamlessly into your application with our APIs Protect your investment with our unconditional commitment to backward compatibility

Innovation through Research We invest in the future by developing innovative modeling concepts GAMS has a large user base in more than 120 countries

Co n

t fr fo act ee r a u s tr ia l!

Maturity through more than 30 years of experience

High-level algebraic modeling system which incorporates all major commercial and academic state-of-the-art solvers

sales@gams.com www.gams.com Phone: (202) 342 0180