BIG DATA: TURNING INSIGHT INTO ACTION
BIG DATA IS A VALUABLE BUSINESS ASSET AND A USEFUL TOOL FOR COMPANIES THAT KNOW HOW TO GATHER, PARSE AND ANALYZE IT. AND AS ALL BUSINESS BECOMES GLOBAL, IT’S INCREASINGLY ESSENTIAL TO REMAIN COMPETITIVE.
BY ROBERT S. BENCHLEY, WITH ILLUSTRATIONS BY HARRY CAMPBELL

big_data

Every week, when 285 million shoppers walk through the doors of Wal-Mart’s 4,740 U.S. stores, Henry Holaday (BBA ’08) is waiting for them — virtually, at least. Holaday is the shopper insights and category solutions manager for Campbell Soup Company’s shelf-stable juices business (V8 and its V-Fusion and Splash brand extensions). He needs to know what shoppers are going to buy. He figures it out by using data — and lots of it.

Because Campbell and Wal-Mart have a shared goal - driving growth in the business - they also share data. Holaday pulls together data from a variety of sources: Syndicated Data provides him with total U.S. market tracking by retailer, manufacturer, brand and item; survey data in which 100,000+ households provide information about their purchases; and market basket analysis of which items share the same shopping cart.

Retail Link, Wal-Mart’s own database, offers proprietary information the company uses to collaborate with its suppliers. “We look at distribution, pricing, shelving and merchandising,” Holaday says. “They tell us if we have the right product assortment, price, placement and promotions — store by store, market by market, in real time.

The Retail Link data also gives us insights into merchandising opportunities with other Campbell brands, as well as brands from other companies. It’s a partnership in growing the business.”
space
BIG BUSINESS IN REAL TIME

Welcome to big data, currently the hottest two words in business. Truth be told, it’s tough to get consensus on exactly what big data is, except … big, and useful. Robert Plant, associate professor and vice chair of the School’s computer information systems department, writes about the topic regularly for the Harvard Business Review, among other publications. “There is so much data out there that I see it as a new currency,” Plant says. “In fact, some companies are buying up certain types of data to prevent other companies from having access to it.

Today, data is a company’s most-valuable asset.” Data has so much value in business, in fact, that the School plans to launch a one-year Master of Science in Business Analytics degree program next fall.

It’s been possible to collect reams of data for years, but it’s only recently that technology has advanced enough that we now have the tools to capture, store and analyze massive data sets. Add in smart people who understand how a business, industry or consumer group works, and businesses can, with an impressive degree of accuracy, predict actions and activities in complex systems ranging from the weather to the stock market to consumer purchasing behavior.

The data is available in incomprehensible volumes, just waiting to be mined, and more is produced every day — layers upon layers of medical records, social media posts, legal filings, census data, corporate transactions, cell phone calls, retail sales, scientific research and more. Almost all of the daily activity on Earth, great and small, ends up stored somewhere. And increasingly, what is getting crunched is real-time data, not historical records.

How much data are we talking about? Think about it this way: Wal-Mart, a company founded in the days when offices were filled with typewriters, now generates enough data to fill 60 million five-drawer file cabinets every hour.

Using all this data in business involves capturing, combining and analyzing seemingly unrelated bits and bytes of information to make business decisions. It’s about finding patterns, connections and relationships among those pieces of information and using sophisticated algorithms and other analytical tools to make timely decisions and predict future events. “Big data doesn’t just mean getting clever answers out the other end,” Plant says. “You have to be pretty smart about what you ask it.”

Many companies have learned the hard way that simply collecting reams of data won’t help their business. Hong Kong-based Mandarin Oriental Hotel Group, for instance, is trying to gain a better understanding of its guests by combining two different sets of data: social media and traditional loyalty measurements.

“Historically, our industry has relied on data warehouses to collect information and build a 360-degree view of our guests,” says Monika Nerger (BS ’95, MBA ’99), the company’s global CIO. Data the company uses includes the dates a guest stayed at one of its 27 properties around the world, how often a guest stays at Mandarin Oriental properties, how many different properties a guest has stayed at and the value of each stay. “Our goal is to tailor the experience in a way that is dynamic to each guest,” Nerger says.

But conventional customer relationship management systems simply can’t keep up with guests in real time, so the company has added in social media data. Now, all that data translates into ideas the hotel can immediately put to use. “With social data interjected in real time, we are more quickly able to understand sentiments — what guests think about our brand — as well as individual behaviors, such as whether a guest spends time at our spa but hasn’t stayed at the hotel,” she explains. “Actionable insight is what is needed to transform our business.”

candado

LOCATION, LOCATION, LOCATION … AND DATA

Actionable insight has long been a mantra in the real estate industry, where proprietary databases of information have a history of influencing a company’s success. What’s changing is just how much of that information is now available publicly. In residential real estate, private companies such as Zillow are just beginning to bring data into public view, but in commercial real estate, much of the key data is public record and readily available, says Matt Rotolante (BBA ’97, MBA ’01), managing director at the Miami office of the national commercial real estate advisory platform Sperry Van Ness.

Factors such as demographic trends, traffic patterns, crime statistics and consumer purchasing all have an impact on a real estate parcel’s desirability and, ultimately, its value. Rotolante has always used data to help clients make investment decisions.

Now, however, he has more data, can access it anywhere and is able to analyze it quicker. It is “rapidly shifting the old paradigms” in real estate, he says. “My grandfather used to trudge around with Miami-Dade plat books. I now access that same property data and more using Google Earth functionality for anywhere in Florida from my cell phone. I use it to help my clients forecast trends and understand why building in a certain location will work.” The ability to mine all that public data and make smart recommendations from it, he adds, means more accurate and better decision-making for customers.
space
BIG DATA’S GLOBAL IMPACT

Data is especially vital in today’s global economy, in which managers often have to oversee operations in far-flung locales. For instance, you can buy a Coke just about anywhere in the world, and Javier Polit (BBA ’87) helps make sure you get one when and where you want it. Polit is CIO of The Coca-Cola Company’s Bottling Investments Group, a $16 billion business with 90,000 employees in 20 countries on five continents that processes, packages and distributes the many beverages owned by its parent, The Coca-Cola Company. It delivers to 6 million outlets each year to provide the 1.8 billion servings of Coca-Cola products consumed every day. Polit is responsible for creating and executing the information technology strategy that holds it all together. “When I look at big data, I see it as a big advantage,” says Polit during a call from Guatemala. His company analyzes consumer behavior right down to the outlet level. It looks for actionable insights on immediate consumption (single servings) vs. future consumption (multipack purchases). It also monitors the sales impacts of culture and weather. Polit offers one example: “In India, it can get too hot to go shopping,” he says. When that happens, the company needs to change its shipment and distribution plans.

Right now, The Coca-Cola Company’s Bottling Investments Group is in the middle of a significant global initiative to improve its collection and use of data. “Dealing with as many countries as we do, we still need one common data repository, one version of the truth, where everybody in every country can go and get the same information,” Polit says. This, he notes, is where big data becomes hard work. With the necessary databases and infrastructure already in place, he is working on the master data management piece — the common data structure across all countries. “If you don’t have that, you don’t have all of the information you need to make the right decisions,” he says. “At the same time, as a global company, you need to leave some freedom in the framework. You may have a different strategic initiative in India than in Germany or Uruguay, and you need to give the locals the freedom to deal with their own marketing execution. As long as the freedom isn’t 60 to 70% of the whole, you’re OK.”
space
HEALTH CARE’S DATA PRESCRIPTION

As the use of big data expands, one of the industries poised to see the most transformation from it is health care. Health care reform has been a major catalyst in getting the whole industry focused on big data, especially for use in managing the chronic diseases that consume a huge portion of overall health care costs, says Jim McDowell (MBA ’85), senior director of health care strategy at Oracle. Data analytics provide insights into better treatments, while helping drive process improvements. “We need to learn how to take better care of people and drive costs down,” he says. “Studies show that one leads to the other. A combination of less-expensive hardware, new software tools and new data sources like social media are creating real opportunities.”

As Medicare and insurance companies increasingly tie reimbursement to patient outcomes rather than billings, technological change for health care providers is being driven, in large measure, by changes in compensation that penalize poor results and inefficiencies, adds Steven G. Ullmann, professor of management at the School and director of its programs in health sector management and policy and Center for Health Sector Management and Policy. “Health care has not had any negative financial impact for inefficiency,” he says. “Now, however, the federal government and the insurance companies have said enough is enough. Incentive structures are changing. Medicare is leading the way, and private insurance companies are following right behind.” 

That change involves a huge datamanagement challenge for Jim Wilder (MBA ’12), senior vice president and CIO of Sheridan Healthcare, a provider of anesthesiology, emergency medicine, neonatology and radiology staffing to hospitals. His company deals with everything from electronic medical records to government regulations, quality analytics and, of course, billing.

Although Wilder is uncomfortable with the words “big data,” he recognizes that data does play a huge and growing role throughout the health care industry. Just how much of a role varies between hospitals, and it varies within hospitals by medical specialty, too. The operating room, for instance, Wilder says, is “the last frontier” for full automation of electronic medical records, and anesthesiologists at some hospitals still deal with paper records (although Sheridan is trying to equip them with digital pens to capture their writing). At the other end of the scale is Sheridan’s proprietary platform for managing neonatal ICUs and the electronic health records of the littlest patients. Everything is digitized, analyzed and used to make decisions about both individual patients and best practices for the future.

box

 MANAGING AND PROTECTING DATA ASSETS

For Wilder, whose systems also interact with his hospital clients’ systems, data security can be as important as innovation in its collection and use. Even when the data isn’t sensitive, like health care records, it is still valuable. Business competitors, foreign governments and organized criminal enterprises have all been implicated in the theft of corporate and consumer data. The more data that is out there, the greater the opportunity for bad acts, says Sandy Goldstein (BBA ’81, MBA ’85), CEO of the data security and forensics firm Capsicum Group and chairman of the School’s Entrepreneurship Programs Advisory Board.

“Large corporations are well aware of the threats,” Goldstein says, “but small- and medium-size companies are all over the map in terms of addressing data governance issues. Data moves in a stream like water, and none of us is fully protected. Our security is really our ability to quickly react to potential breaches, because you’re not going to stop the all threats.”

Security Consultant Cem Gürkök (BBA ’00, MBA ’02) agrees. “It’s not a question of if, but when, an incident will occur for the average company,” he says. “A company needs to be prepared to secure its data by educating employees about threats, having a proper incident-response policy, and constantly monitoring its systems. It also needs to have big data instrumentation in place to process security-related information.” And don’t forget, he adds, that other types of disasters — like fire, flood or power outages — also can take down your company’s data. You need to have multiple backups at different locations and plan for those events, too.
space
TURNING DATA INTO RESULTS

Used well, big data can significantly impact a company’s bottom line. Henry Holaday describes one way Campbell Soup Company and Wal- Mart turn insights into results. Certain items sell better in certain stores.

“The high-velocity items need greater emphasis during the weekly store visits by our representatives to make sure they are fully maintained and stocked on the shelf,” he says. When a best-seller slows down, it’s a concern. It could happen because the store has run out of the product and needs a delivery, or simply because an empty shelf has not yet been replenished from in-stock inventory.

Campbell created an algorithm that analyzes the data, determines the reason for a sales drop, and alerts either the sales rep or the store to remedy the situation. Holaday says improvements in Campbell’s data systems dropped false alerts to reps by 17% and increased the on-shelf availability of its products.

That translates into significant revenue. “Millions of dollars in sales can be lost if a consumer sees an empty shelf and buys a competitive product or walks out of the store,” he says. But having the data — and knowing what to do with it — prevents that from happening.
space
A NEW MASTER’S PROGRAM

BIG DATA IN THE SCHOOL OF BUSINESS CLASSROOM

“Business professionals need to be comfortable with data-driven decisionmaking,” says Anuj Mehrotra, vice dean of graduate business programs and executive education. That’s why learning to analyze data for actionable insights using business scenarios has been part of the School’s curriculum for some time.

Sara Rushinek, professor of computer information systems, pits teams of students against each other in the health informatics and stock market classes she teaches. It can get competitive, just like in the business world. “We’re exploring algorithmic forecasting and programming,” she says. “The students look for patterns of data that assist in decision-making. Typically, each team is made up of students in different majors, so they bring different strengths. The teams compete to see which can use available data to generate the greatest profits.”

In fall 2014, the School plans to launch a new Master of Science in Business Analytics degree program, which is pending approval from UM’s Faculty Senate. The degree is for professionals who need a structured approach to working with data in a business context. “Graduates of this program will have the ability to work with big data, develop and use data-mining methods and software tools, understand and use decision models and develop and use predictive models,” Mehrotra says. “They will be technically trained to understand and tackle large data issues in any type of organization.” — R.S.B.
space
Who Owns Your Data?
In an online exclusive story, the School’s faculty examines other aspects of Big Data: privacy, data breaches and what companies can do when data gets away.
space

Fall 2013
big data
ideas
make
state
links past

Do you expect the Affordable Care Act to have a direct impact on you during the next six months?

Banner Fall 2013