Global Resource Access (R = G): Dynamic Real-Time Reconfiguration of Resources

Visibility in the global supply chain is almost a prerequisite for managing the complex web of product and information flows. The capacity to reconfigure resources globally can start with a simple trend analysis of the key metrics across different markets and product categories. But this beginning should be expanded to a capacity for rapid response to changes in either external market demand or internal process capabilities available at a given point in time.

For example, U.K.-based Aviva plc, the largest insurance company in Europe, is architecting its global customer service processes to constantly search for innovation and efficiency gains to deliver value to its customers. It is common for customer support call centers to use technologies to route calls to appropriate agents (agents with specific skills and temperament based on customer needs) within an office. Aviva’s insurance underwriting and claims business processes are designed to dynamically leverage the appropriate competencies from its global service centers, ranging from Australia to the Philippines, India, Europe, and Canada. Aviva’s focus is on enhancing the consumer’s experience (N = 1) by dynamically routing customer service requests to different parts of the world to provide the best service for that customer without compromising the cost of that service. This requires a capacity for realtime matching of customer profiles with agent skill profiles on a global basis.

In its global customer support processes, Aviva worked with its partners, including a business process outsourcing (BPO) firm called 24/7 Customer in India, to capture metrics in every subtask of the entire customer engagement process to better understand its customers. The process adopted by 24/7 Customer is visible with performance metrics such as customer satisfaction and time to resolve the problem. Outcome measures such as cross-sale and transaction completion are tracked in real time for each call. In order to accomplish this dynamic routing, Aviva must have visibility to the type of customer, loads, and the quality of agents and their skills in various locations. In an article published in the Economic Times in India in 2006, Richard Harvey, Aviva Group CEO, says, “Because we take a lot of care to measure customer satisfaction on a completely arm’s-length basis, we can demonstrate that our customer satisfaction from India is as strong as or even stronger than the United Kingdom.”

An additional benefit of this transparency in its global processes is that it enables Aviva to constantly monitor the best-in-class process execution across its global centers and disseminate that knowledge to other centers. John Ainley, HR director at Aviva, admits that the company is building a culture within the organization to promote competition in process performance across its global centers, to prepare its employees to emerge out of the “not-invented-here” syndrome and to accept process innovations from other centers. This leads to continuous improvement across all centers. Aviva has certainly taken a lead in reconfiguring global resources to create customer value in the insurance industry. But it is not alone.

A visit to the Chennai (India) office of the Dallas-based Perot Systems reveals a new level of visibility in its processes and a capability to predict and reconfigure resources for its global clients. The business process service unit of Perot Systems provides backoffice support to a number of hospitals and health insurance clients. Its Chennai center has developed a customized technology platform that integrates operations, HR, and finance business processes in a single portal. The Chennai team has disaggregated every process assigned to them and carefully identified both the skill requirements and performance metrics around each task. For example, each claim can be broken into subtasks. Each subtask requires a specific skill. One can identify the performance metrics appropriate for each subtask. Such a detailed understanding of the business process (granularity) is a key ingredient in their success.

Granularity is as important as visibility. Granularity allows managers to examine in depth the process steps, as well as the appropriate skills needed to perform them. The training modules required for each task at Perot Systems processes are digitized so that individual agents can take a set of e-learning courses at a time convenient to them. As the back-office business processes for large health insurance clients are executed in its Chennai office, the integrated platform automatically tracks the performance of every process step by every agent in every work shift. The best and worst performance levels across the organization are derived in real time through live data. Performance goals for each agent are redefined periodically with an analytical engine to enable continuous improvements in their processes and hence value for their global client. The same analytical engine also computes profitability for every client at the end of each shift.

Anurag Jain, vice president for business process services at Perot Systems in Dallas, states that this integrated platform in the company’s India office allows it to assess the performance of its employees in a direct and transparent way by which individual employees are presented with their performance in a task as compared to the mean, best, and worst 10 percent of performers in that task within the organization. It is not surprising that this BPO unit of Perot Systems has bagged several awards. And Mr. Jain has now been promoted to the position of India head at Perot Systems, which means he is leading the overall consulting, applications, insurance, and BPO units in India. Perot Systems has also extended into a new business service that helps engineering services firms apply lean manufacturing concepts to their operations.

While this may be viewed as an invasion of privacy, the reality is that firms are beginning to operate at a new level of visibility to individual performance, a performance that is measured and compared with others in the organization. In a high-performance organization, there may be no place to hide for the employees, agents, or their managers. Vardhman Jain, heading the offshore BPO Chennai center of Perot Systems, claims that its primary motivation was to create a transparent culture in which there is constant peer pressure to perform as well as incentive to improve processes.

A majority of its process improvement effort emanates from its own agents, akin to a Toyota production system. He adds that the company immediately spots development areas of employees who are unable to perform at expected levels and assigns training modules that specifically improve performance in targeted areas.

This same transparency in the company’s processes and analytics also enables it to accurately measure the cost incurred for each global client, and hence, the related client profitability. Performances of individuals or the profitability associated with a customer are not exercises performed periodically; rather they are performed continually. The company’s platform provides instant profitability of each client as it executes its processes. The analytical model can also predict future run rates of revenue based on demand patterns. This creates a capacity for Perot Systems to know profitability levels of potential engagements. This level of granularity and the capacity to execute the engagements allows Perot to submit proposals of great accuracy.

Large IT systems vendors in India, such as Infosys and TCS, have developed capabilities to constantly monitor the demand and resources needed for new IT services in their global markets. These firms recruit about 25,000 people annually, and their business models demand that they train these new recruits rapidly. These firms manage around 3,000 projects on site and offshore globally. They need to build capabilities to track latent demand for expertise in specific IT tools such as J2EE (Java to Enterprise Edition) or technology such as RFIDs and use these insights to manage their talent supply chain. Their annual training budget exceeds half a billion dollars. They need to understand resource needs and performance at the project level and profitability and experience at the customer level. Their challenge is to anticipate global demand for services, recruit and train for the right skills rapidly, and deploy resources to the right projects for the right clients globally to maximize long-term profitability. This is an analytical problem akin to a quantitative assignment problem familiar to operations researchers.

Nirvana, an emerging BPO company in Bangalore that serves global financial services clients in customer support and other back-office processes, is yet another example of a company’s unique applications of analytics and process discipline to constantly improve its understanding of customers and deliver value through global resource leverage. In addition to business process visibility and metrics- and measurements-based decisions in daily management, Nirvana has further integrated analytics-driven insights into its decision processes to build a capacity for dynamic resource reconfiguration. For example, while typical BPO organizations record at most 10 to 15 percent of the customer calls for customer support from India, Nirvana records 100 percent of the customer calls. This enables Nirvana to build a real-time customer profile based on both transaction data and keywords searched from customer conversations recorded digitally and mined for insights. In addition, Nirvana’s IT infrastructure also tracks the voice amplitude of each customer during the service call to sense the customer’s frame of mind or temper. For example, the voice of a male customer calling from Dallas is tracked and compared to the typical voice profile from similar callers. The variation in a customer’s voice amplitude is tracked in real time to be used as one of the inputs to build real-time customer insights and alter the company’s services appropriately, if needed. For example, Nirvana’s analytics engine based on data from multiple sources (transaction data, voice recordings, and keywords used by customers) has helped a large U.S. financial institution predict propensity to switch to a competitor at an individual customer level. This information has enabled the company to proactively alter its services to some of the high-risk customers and reduce its customer churn rate by
15 percent.

Similarly, consider the collaboration between the multi-billion dollar online retailer Overstock.com in the United States and 24/7 Customer in India. Virtual stores and sales chat agents are common in online retail sites because they try to enhance customer experience through either automated or human support “online chats” with customers. Unlike physical stores, online retailers, such as Amazon.com, eBay, or Overstock.com, have millions of visitors every day, and the majority of these visitors have no intention to buy and can easily switch to other shopping sites at the click of a mouse. Hence these retailers look for analytics to identify the right customers to engage in chat.

In this partnership with online retailers, 24/7 Customer experimented with analytics to crack the science of determining the right filters to apply in inviting customers to chat and at the same time matching the appropriate resources (that is, agents) for a given customer to enhance overall customer experience. First, the process of selecting customers and assigning agents is made visible to the U.S.-based retailer, and performance outcomes are transparent. Second, for individual customers who are invited to chat, the past data about those customers and their current requests or queries are combined to identify the appropriate agent to be assigned to that chat, illustrating real-time reconfiguration of resources.

The performance of agents, in terms of closing sales and overall customer experience and loyalty, is constantly assessed as feedback inputs to this analytics engine. The goal here is not to optimize product-agent selling output but to develop a real-time analytics engine that uses data from multiple sources to assess agents based on a set of customer, product, and experience attributes to determine the best available agent to talk to a given hot-lead customer. This process has also improved the performance of some agents by over 60 percent because it matches the right agents (based on their strengths and knowledge in specific product and customer categories) with the right customers. Now, if the company extends this by allowing customers to define profiles of the agents it would like to chat with, we will be moving closer to anticipation of demand and resource needs and cocreation of value.

It must be obvious that in order to perform analytics for insights, we need to focus on the visibility, granularity, accuracy, and timeliness of data. Visibility to the processes is a necessary first step. The premium paid by large businesses for logistics services offered by UPS or FedEx is not for mere visibility. These businesses are also paying for accuracy, timeliness, and the ability to reroute the businesses’ packages based on their current needs—that is, the capacity to reconfigure resources. Dave Barnes, senior vice president and CIO at UPS, states that his company has undertaken several time and motion studies to continuously optimize every step in the package delivery processes. These studies have revealed methods for loading the trucks in better ways through new heuristics and analytical methods such as training their drivers to fasten their seatbelt with their left hand while turning the ignition key with their other hand. Package routing information is constantly tracked and planned for each delivery truck, allowing for any changes in the routes if required either by the customer or by other interferences such as traffic or weather.

The examples of Li & Fung, UPS, the Department of Defense, 24/7 Customer, Perot Systems, and Nirvana illustrate increasing sophistication in the means available to create visibility and transparency to business processes. These examples also highlight the learning capability to reconfigure resources in real time, continuously improving the skill base of employees such that consumer needs and employee skills can be matched, and finally, building a personalization component in activities that appear simple and commonplace, such as delivery of parcels. These advances call for the integration of analytics with explicit business processes defined with fine granularity. Such integration demands extreme levels of training and intense measurement of both people and business processes. These systems are measurement intensive, and they prosper with the capacity for real-time feedback and corrective actions.

R = G needs to be appropriately configured to serve N = 1. The building blocks of analytic capabilities for R = G are depicted in Figure 3.2. It should now be obvious that visibility to processes and data within global supply chains (R = G) is crucial for building the multiple layers of capabilities that are critical for dynamic reconfiguration of resources. This visibility also helps managers anticipate consumer behaviors such that they can add or subtract appropriate resources to the whole supply network. In this process, we will also be able to get new insights—be it for operational improvement as in the case of UPS or for strategic redirection and course correction as in the case of the DOD supply chains that require integration of three distinct supply chains into one.

by C.K. Prahalad and M.S. Krishnan, Via The New Age of Innovation: Driving Cocreated Value Through Global Networks (2008)

Global Resource Access (R = G): Visible Global Supply Chains

The capability to leverage global resources will demand new levels of visibility and agility in managing logistics of physical goods and resources (globally) to meet unique demands of customers

Let us start with a well-known example. Access to global resources requires the capability to tap into a complex web of resources, expeditiously, and at the best global price. Li& Fung, a premium global trading group covering high-volume, time-sensitive goods including fashion accessories, furnishings, handicrafts, and home products, is a good illustration of process innovation through analytics. Li & Fung started as a pure trading company, sourcing its products from China for exports. However, within a decade Li & Fung had put in place a global network for managing supply chains for a large number of retailers in Europe and the United States. Unlike the traditional trading business model, Li & Fung does not own any production facility or large warehouses. As stated by Victor Fung, the CEO:

Everybody thinks that a trading company is just taking an order from the right hand and giving it to the left hand. The idea is that, maybe foreigners don’t know which factory to go to, so you perform an introductory role, maybe a quality control role, and there it stops. . . . Whenever we go in, we don’t just give them [the suppliers] an order and hope that they know what to do. We hand-hold them through the whole process. That’s why we say we almost are a virtual factory. . . . It is the way we orchestrate the production, come up with samples, and feed them information. All that is going way, way beyond that original matching function.

Li & Fung manages a large number of quality-conscious, cost-effective producers who can effectively deliver orders on time for customers such as JCPenney. More recently, the company identified the need to expand in locations near Europe and the United States to cut lead time for delivering physical goods. The overall business model of Li&Fung is based on the “end-to-end business process knowledge”—that is, from the point-of-sales data emanating from a specific branch of JCPenney in the United States (for example, how many white shirts, cotton, size 16 inches/32 to 33 inches, pattern XYZ) to its ability to replenish the inventory in that location through articulating its supplier network in maybe three countries.

The complexity of the company’s supplier network demands a capacity to manage information regarding regulatory restrictions across countries, managing the skill base of its suppliers and hiring in specific locations, and finally, integrating all this information to provide a seamless one-stop shop for its customers. The insights derived through the company’s accumulated data on various markets and individual supplier capabilities enable the company to deliver unique value. This system cannot function without a detailed, constantly updated understanding of all the suppliers—capacities, capabilities, costs, skills, and distances. This also demands a detailed understanding of the customers’ needs—urgency, quality, locations for delivery, and profitability. R = G must start with this level of visibility to all variables that can impact the appropriate resource configuration—“plant A in Thailand to serve JCPenney in Dallas for this order”—decisions.

In an R = G world, establishing the visibility to the entire chain is a good first step. Schneider Electric is the world’s largest manufacturer of electrical distribution systems and components. The company has a healthy growth rate; sales are U.S. $8.8 billion, and it has 70,000 employees in 130 countries. Schneider’s purchasing organization procures for four leading markets (each worth U.S. $1 billion): raw materials and means of production, fabricated metallic and plastic components, electronic and electrical devices, and nonproduction services. The global purchasing operation works with these four markets and a total of 33 commodity groups and multiple country organizations. The complexity of a supply chain such as this makes business analytics a necessity to effectively compete. Many large companies operate such supply chains without full visibility, and the consequences are obvious as they expose their supply chain to unknown global sources. For example, Serge Vanborre, a senior manager at Schneider Electric’s purchasing headquarters, says, “We want to know who is buying what from whom. We want to know the global purchases, be able to do an analysis in order to repartition our purchases and verify if the supplier policies are followed.”

A well-known example of visibility in a global supply chain is exhibited by leading logistics firms such as UPS and FedEx. For example, Atlanta-based UPS moves over 15 million packages around the world in a day, and it provides complete visibility to the end consumer on each and every packet. FedEx recently integrated the software systems of its ground, air, and freight businesses to provide full visibility to all of its customers and employees for the 6 million plus packages it handles every day.

Similarly, in one of the largest radio frequency identification device (RFID) projects implemented so far, Unisys has created full visibility for the global supply chain of the U.S. Department of Defense (DOD). Prior to this new system, the department operated three different supply chains for the army, navy, and air force with minimal integration. Furthermore, there was almost no visibility. In contrast, the new system connects global suppliers with 30 centers of DOD to any location in the world from Taiwan to Tacoma, providing complete visibility through RFID tags. As a result, when the military runs out of spare parts for a tank in Iraq, it has the capacity to locate the floating warehouse in the nearest ship instead of having to source the parts from the nearest depot, which in the past has often been far away.

Similarly, Homeland Security demands that it know where the cargo shipments that reach U.S. ports have been. So far, this has been an elusive goal. For example, a Sara Lee innerwear shipment manufactured in Pakistan and loaded in a container at Karachi can travel through a feeder ship to Mumbai, India, and then to Sri Lanka, through the Suez Canal, to Nova Scotia, and finally to New York. The items in the ship are invisible during their circuitous course of travel in the sea for almost a month!

by C.K. Prahalad and M.S. Krishnan, Via The New Age of Innovation: Driving Cocreated Value Through Global Networks (2008)

Analytical Tools Provide Business Insights

Traditionally, managers depended on experience and intuition to develop insights—“gut feel,” if you will. Most often a gut feel is based on past experiences.  Gut feel and intuition are important but in a fast-changing competitive environment, experience of the past is less and less valuable. Foresight, not hindsight, is of value.

Foresight is a result of understanding, through structured and unstructured data, the unfolding of competitive dynamics. There is value in identifying new patterns of relationships, predicting the
behavior and evolution of systems, and mitigating risk. In an N = 1 world, the behavior of individual consumers as well as broad patterns of change must be understood. In R = G, the capabilities of each vendor in the ecosystem in terms of costs, time, and quality levels must be understood and matched with the specific demands of a single consumer at a point in time. Furthermore, given the complexity of the entire ecosystem, the impact of change in any single variable, such as order entry, will have a ripple effect on other related subsystems such as inventory, spare parts, and manufacturing lead times.

A “small change” in order entry could trigger multiple changes in the totality of business processes. Managing the systemwide impacts of changes cannot be left to the gut feeling of managers. However, individual managers can, based on their experiences, interpret the signals differently (especially in a rapidly evolving system). Hence, foresight based on the real-time analyses of both structured and unstructured data is indispensable. Intuition and gut feeling are still useful, but not as a substitute for analytics.

Keeping business processes current and compliant with all changes and at the same time gleaning insights about the evolving behavior of consumers and the supply network require a commitment to analytics. Consider, for example, the Indian IRS. It is known that not everyone in India pays his or her taxes adequately. The IRS can safely start with the assumption that there is significant tax evasion in the country. In order to deal with widespread tax evasion, India’s tax agency is building a database of declared income and consumption patterns, such as travel, purchases of bigticket items such as cars, plasma TVs, deposits and withdrawals from banks, stock market activity, and the like, to spot patterns of tax evasion. The focus is on identifying individual taxpayers (N = 1) for further investigation. This project calls for deriving insights based on data from multiple sources.

A similar initiative is in place at the IRS in the United States as well. In the United States, the cost of tax avoidance is estimated at $350 billion. Tax evasion, around the world, is a moving target. In order to predict these behavior patterns, complex analytic models have to be developed. Data from a wide variety of sources must be pulled together to see the emerging patterns. Microsoft recently announced its purchase of a small start-up health search engine called Medstory, Inc., that applies advanced analytics to structured and unstructured medical and health information in journals, government documents, and the Internet to present an enhanced customer experience in access to health information. The desired result is a personalized information search based on one customer’s family history, prior medication, age, and gender.

Analytics must be driven by strategy. For example, in order to price health insurance for each diabetic consumer (patient), we need analytics, which in real time monitors behaviors (compliance on predetermined routines) but can also forecast likely behaviors. Analytics can also show where to allocate resources and how to optimize the “resource network.” Should a call from an irate and important customer in New Zealand be routed to India or Australia? This is a real-time decision, one of thousands, to which the firm must respond creatively. Insights also result from consumer concerns and comments. Understanding and researching blogs and chat rooms is another important source of insights. The capability to use analytical modeling tools is critical in every aspect of value creation, from understanding customer preferences and behaviors to supply chain management, global resource reconfiguration, skill management, and risk mitigation. We (i.e. CK Prahalad and MS Krishnan) will illustrate the power of analytical tools with applications focused on leveraging global resources to serve individual customers in global markets first (R = G), followed by illustration of such tools in moving toward N = 1.

by C.K. Prahalad and M.S. Krishnan, Via The New Age of Innovation: Driving Cocreated Value Through Global Networks (2008)

Real-Time Analytics as a New Source of Competitive Advantage

In previous entries excerpted from Chapter 2: Business Process as an Enabler of Innovation, CK Prahalad and MS Krishnan have identified business processes as the enabler of an innovative culture through their impact on both social and technical architecture. As a critical intermediate step between strategy and operations, the quality of business processes (granularity, flexibility, and clarity) determines the capability of firms to compete effectively. By definition, in a rapidly changing competitive environment, business processes cannot be static. The dynamics of an industry dictate the rate of change in business models and strategy. Business processes must keep pace with this rate of change in the strategy of the firm. More important, business process capability may suggest new ways of competing.

Competitiveness favors those who spot new trends and act on them expeditiously. Therefore, managers must develop insights about new opportunities by amplifying weak signals. These weak signals emerge from insights derived through a deep understanding and interpretation of a wide variety of information. For example, recognizing that SMS (text) messaging using a cell phone will be an important method for settling small payments is critical for the longterm success of Visa and MasterCard.

Spotting new trends requires comprehension of consumer expectations and behaviors and technological changes, as well as the nature of the supply chain and opportunities for its improvement. How does one spot trends early? Can a firm develop tools that aid in building insights? The new competitive landscape requires continuous analysis of data for insight. Analysis that is only episodic and ad hoc (as when a senior manager commissions a specific study, say, to assess the impact of oil prices on shopping patterns) or periodic (such as actual sales compared to forecasts) will not suffice. Traditional analytical approaches are often asynchronous with business changes. Hence, delays in recognizing, interpreting, and acting on the trends are emerging as critical impediments to competitiveness.

Every firm accumulates a voluminous amount of transaction data (for example, sales transactions) and equally large volumes of unstructured data (for example, video clips and advertisements). Managers need a mechanism to understand the accumulated information and extract valuable insights. Real-time analytics seize the opportunities and mitigate the risks in seeking to have global resources serving single customers.

The terms analytics and analytical models are used to describe a class of mathematical applications that permits businesses to crunch everything from picking stocks in trading rooms rapidly (in less than a millionth of a second) to identifying specific advertising messages based on your search at any time in Google. Some recent trends are helping firms build this capacity. Algorithms and quantitative methods used in analytics are evolving to help managers derive insights, often combining structured transaction data (numbers) and unstructured data as in documents, images, and video. Digitization of business processes, the Internet, and evolving ICT architecture enable real-time predictive modeling. These capabilities are at the heart of effective management in an N = 1 and R = G world In the subsequent entries, this will be demonstrated).

The link between data, analytics, and insights is shown in the figure above. As you can see, the quality of insight depends on both the quality of data and the quality of analytics. Models that are not built specifically to inform on strategic priorities are of little value of line managers. More important, insights that are not available when decisions have to be made are of little value. In this chapter, we will assume the availability of high-quality data that capture the millions of transactions in a company—be they sales, warranty claims, orders placed, or payments to suppliers. (The quality of data is a major concern in many firms. Data collection often is not standardized across the firm. Increasingly, data are also collected in a highly decentralized fashion, for example, by delivery agents with handheld devices. Rather than engage in a detailed technical discussion on how to “clean up databases,” therefore, it is assumed that the data quality is acceptable to perform analytics.)

by C.K. Prahalad and M.S. Krishnan, Via The New Age of Innovation: Driving Cocreated Value Through Global Networks (2008)