It seems that almost every organization is on the “big data” bandwagon. But, as this author writes, rallying around the newest, new thing is one thing. Understanding what big data is and how it can move your organization forward to lead the pack is quite another. Readers will learn how to use big data to create sustainable competitive advantage.

“Big Data:  A New Competitive Advantage” blazed a headline in a recent Ivey Business Journal[1]. But the fact is that while such claims are common, they may also be misleading.  “Big data” is not new and creating and sustaining a competitive advantage is rarely easy.  The key question is “How do we extract the kinds of useful, meaningful managerial insights from big data that will result in a sustained competitive advantage?”  This article will suggest several ways that managers can mine big data to enable their firm to establish and sustain competitive advantage.

 

Background

Many firms now manage vast daily inflows of raw data. However, data by itself is just a cost item.  What’s creating the excitement – and headlines – is the possibility that used properly and strategically big data has the potential to create competitive advantage. Barry Beracha, the former Chief Executive of the Sara Lee Bakery Group, has been quoted as providing the maxim of the data-driven management school:  “In God we trust.  All others bring data.”[2]  However, few executives can process modest amounts of raw data, let alone gigabytes of the same that flow into many firms daily.  Further, this data is often of very high quality, precluding the use of the traditional wailing of “the data is no good,” often just a convenient excuse for ignoring it. 

Many are now talking about the value and use of big data. But the fact is that there are very few business analytics professionals who understand the business and the business issues and the analytics required to add-value and extract relevant information from the data.   Analytics is to management as a light bulb is to darkness:  it is illuminating and helpful in revealing both future opportunities and pitfalls.  Descriptive analytics seeks to understand past data and is widely used. Predictive analytics seeks to understand the future. This is a challenge for many firms, since it brings in risk (the future is uncertain) and the need to manage risk.  Prescriptive analytics tries to identify what the firm should do. This too presents a major management challenge since humans have intuitive ideas as to what the firm should do that may conflict with the results of prescriptive analytics.  

In our new book[3], Analytics for Managers, co-author Greg Zaric and I argue that analytics has been and can be a source of competitive advantage. Tom Cook led an analytics group at American Airlines (AA) that created a competitive advantage for AA for more than 15 years (from about 1985-2000). Joe Hinson, General Manager of Operations Research at Federal Express, chaired the corporate strategic planning committee and reported directly and regularly to CEO Fred Smith.  “FedEx is an archetype of a company that has succeeded by applying the scientific methods to its operations.  Models and analysis have informed many of [FedEx’s] crucial, business-shaping decisions.  In cases in which[advanced analytics] wasn’t used … the company performed poorly.”[4]

A 2013 survey found “67 percent of respondents reporting that their companies are gaining a competitive advantage from their use of analytics,” up from 58 percent in 2012 and 37 percent in 2010[5].

Analytics, however, is often misunderstood and seen as a threat by many senior managers. But analytics at its best informs decision makers, leading to improved decisions. Analytics does not make decisions but input from analytics can help intelligent, knowledgeable, experienced people make better decisions.  Creating a competitive advantage from big data requires both the big data itself and the skills of the analytics professional.

 

Starting into business analytics

Executives frequently ask how their firm can start a business analytics function. In fact, there is no standard approach.  Hiring an analytics champion to develop a team that can produce useful and valuable input to firm decision-making, while being sensitive to the management and personal issues involved, generally requires poaching someone from another firm.  If this firm is a direct competitor, conflicts over intellectual property (IP) rights might result.  Increasingly, a common alternative is to buy analytics from a domestic or offshore supplier.

IBM, Accenture, SAS Institute and others have invested heavily to become major domestic and global suppliers of analytics to business and governments:

“Over the past five years, IBM has invested more than $14 billion in 24 analytics acquisitions. Today, more than 8,000 IBM business consultants are dedicated to analytics and over 200 mathematicians are developing breakthrough algorithms inside IBM Research.”[6]

India has emerged as a major offshore supplier of analytics due to its history and focus on mathematics and statistical training. Dhiraj Rajaram, an MBA from Chicago and former strategy consultant for Booz Allen Hamilton and PricewaterhouseCoopers, founded Mu Sigma in 2004. Mu Sigma has emerged as an analytics powerhouse:

“With over 2,500 [analytics] professionals, and over 75 Fortune 500 clients across 10 industry verticals, [Mu Sigma is] one of the largest decision sciences and analytics companies. Our interdisciplinary approach to problem solving using cross-industry expertise corroborates our sustainable engagement model with our clients, making us one of the most preferred analytics and decision sciences partners among our contemporaries.”[7]

Other offshore analytics suppliers started as captives of multinationals and were eventually spun off. American Express was a leader in setting up a captive analytics supplier in India, while General Electric established GENPACT in countries that focused on education in mathematical sciences; centres were set up in India, China and eventually Hungary.

“Driven by a passion for process innovation and operational excellence built on the legacy of serving GE for more than 15 years, [GENPACT’s] 60,500+ professionals around the globe deliver services to its more than 600 clients from a network of 74 delivery centres across 20 countries supporting more than 30 languages.”[8]

Firms that outsource some or all of their analytic processes offshore include Wal-Mart, Allstate, Goldman Sachs, HSBC, Citibank and Royal Bank of Scotland.  Outsourcing and/or offshoring analytics needs to be managed with great care, especially if the analytics or data involved is critical to the competitive package of the owner (more details see “Competing with Analytics by Taking Analytics Offshore”[9].)

 

Outsourcing and competitive advantage

The ease of access to outsourced analytics complicates the task of creating competitive advantage through the use of big data and analytics. Ease of access also makes sustaining any competitive advantage more difficult.  However, many firms manage to do this by recognizing and exploiting the factors that make it difficult for competitors to replicate the work.  Examples of big data prescriptive analytics at Dell and the Industrial and Commercial Bank of China (see[10] for more examples) showcase two common approaches.

Dell was founded on a configure-to-order (CTO) business model, where customers chose the various components of their PC and the custom PC was then assembled and shipped.  Using CTO, Dell grew to become one of the top PC brands. But this business model began to tire about 2007 and Dell decided to expand to new channels including retail, and value-added networks in the emerging nations.  This channel expansion required Dell to specify a product line of “off-the-shelf,” pre-configured PCs, which they could have designed experientially. But instead, Dell turned to prescriptive big data analytics.  By accessing their billions of records of customer-purchase history and using advanced analytics that evaluated millions of combinations of possible product configurations, Dell analysts were able to define a product line of pre-configured PCs that (allowing for some trade-ups) captured almost three quarters of Dell’s notebook and desktop sales.   Today, these configurations make up about half of Dell’s sales.[11]

This is an example (there are many others) of big data and analytics informing a major business decision.  It was a strategically important, one-time decision that was made by using proprietary data, where the combination of data and advanced analytics contributed to a great decision.  The competitive value of the data and analytics appears in the resulting product line, which is likely sustainable. 

Industrial And Commercial Bank of China (ICBC) has some 16,000 branches throughout China and is by many measures (market capitalization, deposits) the world’s largest bank.  The branch network is a key strategic asset of any retail bank and managing the branch network is a core banking competency. The branch network was seen as ICBC’s most important service and marketing channel, since the branches were the most effective means of customer acquisition.

As China’s economy expanded and modernized, the centres of business and customer activity moved around, with new urban districts and satellite cities emerging, and personal wealth increasing. These fast-changing conditions and the competitiveness of the Chinese banking market make it essential for ICBC to quickly identify new high-potential locations for branches, as well as moving or reconfiguring existing branches. ICBC needed to determine how many branches should be in each city and their locations, identify new high-potential regions for expansion, and improve branch location decision-making in order to minimize costs and avoid poor location choices.

In 2006, ICBC partnered with IBM Research to begin development of ICBC’s branch network optimization system, which was driven by analysis. Each city was divided into tens of thousands of 100-metre square cells, and the business activity and demographic data for each cell were identified from Geographic Information System databases.  This data was used to calculate a base, market-potential value for each cell, which was then refined using human opinion and optimization models to produce a final market potential estimate.  Finally, large-scale optimization coupled with expert judgment were used to find the cells which offered the best locations, taking into account market potential, competitors’ locations, and other ICBC branches in the neighbourhood. 

This process has been implemented with great effect in over 40 major cities in China.  ICBC attributed US $1.04 billion in deposit increases in Suzhou to the use of this system for improving branch locations.  ICBC has trained 500+ employees to use the system and now sees branch location as a strategic problem that is ongoing and needs constant attention.

The massive data required to address this problem and the ability to acquire, maintain, and process this rapidly changing data using complex large-scale optimization makes it very difficult for a competing bank to replicate the system quickly.”

This example uses public-domain data, so the data is not the source of advantage. Rather, it is the complexity of the analytics, which few competitors can copy (providing that ICBC has some control over the IP developed by IBM).  Further, there is a first-mover advantage from capturing the best locations.  Consequently, this work should provide a competitive advantage to ICBC for some time.

 

Many executives have not embraced the Big Data movement 

These examples illustrate the considerable benefits of big data and analytics-based decision-making. So then, why doesn’t everyone do this?  A large part of the answer is the strong resistance from managers who are not comfortable with data or analytics.  There is a scene in the movie Moneyball[12] where the Oakland A’s scouts are discussing their recruiting strategy for high school prospects. Billy Beane (the decision maker) arrives and informs them that the data shows that this recruiting strategy does not work and that the A’s are not going to use it anymore.  The scouts, who have built their careers on their special skills in evaluating and recruiting high school players, react to this news quite sharply.  Senior executives are also likely to react strongly and negatively if their deeply held intuition is challenged by results derived from data that used analytics that they do not understand.  

Executives who are more comfortable with data and analytics recognize the fact that the future is so uncertain that it is difficult, even after the fact, to tell a good decision from a bad one based on its outcome.  A horrendous decision can run into a stream of good fortune and turn out well, while a great decision can hit bad luck and be a disaster.  It is therefore important, when assessing a decision or a decision maker, to look at the process by which decisions are made.  Increasingly, when we look for great decision makers we have actually begun to look at the way they handle the data and analytical input to their decision-making processes. 

Many C-level executives have strong people skills but weak quantitative skills. However, in a data analytics world, she who understands the spreadsheet often exerts disproportionate control over the discussion.  Those C-level executives who make an effort to achieve an understanding of data and analytics will be able to control the discussion and capture the great benefits from data-driven decision making.    

McGuire et al.¹³ point out that Big Data has the potential to replace some managers. But, on the positive side, we are seeing more and more situations where analytics-based systems are enriching the manager’s role.  For example, personal loan approval systems, airline and hotel pricing systems, and supply chain optimizers routinely make the millions of decisions that are required daily to manage these complex businesses.  These are cases where feedback loops are short, and where it is better to have fast, consistent and robust decisions made by software algorithms than to have humans try to keep up with a non-stop demand for decisions. Consequently, the manager’s job is enriched when it is changed from being an operations-level, decision-making function to a managerial role, where the individual is the supervisor of a complex system.  The manager now needs to understand both the operations system and the analytics that drive the decision making for that system. Otherwise, the manger will override the decisions from the algorithm too frequently and so reduce its effectiveness.

While the wise use of big data and analytics can improve management decision making generally, the greatest promise of big data and advanced analytics is also the most elusive, namely that it will lead to new business models[13].   Google is reinventing people management using advanced analytics[14].  Others see that firms that use analytics and big data can segment markets according to the needs of the individual customer, a capability that creates opportunities for new kinds of businesses.  This is a decided advantage, as many businesses exist largely because they do not know the needs of individuals. For example, the retail store exists as a display of variety.  Retailers or supermarkets that know the precise needs of their customers do not need to display variety and may not need the costs of retail space. 

 

The future for the senior executive

Successful senior executives will play a leading role in directing the firm’s efforts to glean useful, relevant information and knowledge from big data. They will also be involved in developing and directing innovative methods and business processes.  Those executives who rely on instinct for decision-making will be increasingly challenged to understand and accept new ideas and innovations emerging from big data and analytics.

The impact of big data is destined to appear first in highly competitive industries.  Competitive advantage works both ways:  If your firm can create an advantage using big data and analytics, then your competitors can also do so, putting your firm at a disadvantage.  This can occur swiftly, as in the case of Peoples Express Airlines, where a highly successful and growing air carrier was put out of business in less than three months because it was unable to respond to what founder and CEO Donald Burr called “sophisticated computer programs” that were able to immediately match or undercut his prices.[15]  The successful senior executive will need to constantly sense the environment and be aware of leading-edge analytics being used by competitors.

The world of big data will require new organizational structures:  Chief Data Officers and Chief Analytics Officers are appearing (although the titles are not yet standardized).  Finally, the executive skill set has expanded and the job of the C-level executive has becoming even more complex.  Leadership, visioning and people skills remain important. However, the executive must now also have the skills to critically assess the innovations emerging from data and advanced analytics, incorporate their own knowledge and experience, and arrive at decisions and actions that move the firm forward.  Those executives who doubt the benefits of investing in big data and analytics may find, like Donald Burr, that their firms have no response to “sophisticated computer programs.”^15.