PDF File: 

ASC5_wp_haydock.pdf

SAS

Supply chain intelligence reveals opportunities to reduce costs and stimulate revenue growth and it enables companies to understand the entire supply chain from the customer’s perspective.

Before examining the supply chain of a particular business, it may be advantageous to understand the motivations behind supply chain improvements. The confluence of several recent business and technical developments has made the study of supply chain improvements extremely relevant:

1. An understanding and an appreciation that serving a customer is done by many firms in collaboration.
2. The ability to capture and analyze data from enterprise resource planning systems across the firm, combined with business intelligence methods, can reveal opportunities to take out additional costs previously not recognized.
3. A desire by executives to understand the enterprise as a whole.
4. Dramatic improvements in integrating optimization and forecasting software, which have improved the output of computational solutions.

This paper will provide the reader with a perspective of the elements of supply chain intelligence through examples.

Figure 1: Supply Chain Intelligence, Breadth, and Objectives

Driving Forces Create the Need For Supply Chain Intelligence

A prominent management style throughout the 1990s was to drive down into an organization the responsibility and authority to drive a division's profits. The smaller the division, the further down into the organization one could drive the responsibility for profits. In organizations effective at using this management style, each employee felt responsible for the success of his respective division.

So far, nothing is wrong with this model. In fact, using this model effectively seems to have been one of the key success norms for chief executive officers during the decade. CEO portfolios contained assets made up of the divisions within their companies. The typical measure of success was whether a division was No. 1 or No. 2 in its respective marketplace. Funding and careers were made or broken with the ability to generate results. Corporate headquarters were typically run with a minimum of staff; in fact, in the early 1990s companies often moved headquarters personnel out of headquarters and into the operating divisions so they could be closer to the firm's customers and where their contributions could be solidly measured by performance in the marketplace. This remains the preferred corporate management model, and it may be for quite some time — it clearly drives business results.

One of the model's disadvantages is that it tends to decentralize some core supply chain functions into the divisional operating units. Managers responsible for the results want control over most, if not all, supply chain functions producing those results. Procurement is a good example, where the trend for each division to do its own strategic sourcing began to emerge. Logistics is another example, where each division moved its own goods from manufacturing through to the consumer. To help these autonomous divisions do their jobs, companies purchased enterprise resource planning (ERP) systems and supply chain management systems highly targeted to solve a few specific challenges, such as inventory management, materials requirements planning, invoicing, transportation routing systems, and shop floor control. Divisions wanted to get a handle on their performance as well as achieve operational excellence, so they installed their own "instance" of these ERP and supply chain management systems.

This management style and strategy worked well until the U.S. economy began to slow down in 2000. It was difficult for corporate chiefs to go back to operating divisions and try to make them more efficient. The operating units had been driven to efficiencies for nearly a decade, and there was little to be gained at reasonable costs. So they began focusing on "cross-organizational" efficiencies — especially in logistics, manufacturing, and strategic sourcing. Managers began to ask: "Can I take a horizontal look across the operating divisions and gain efficiencies that I can't get by tasking vertical operating companies to be more efficient?"

Firms are highly complex organizations with multiple divisions, making an "information slice" across divisions complex and multidimensional. The information systems that collected the data and solved some of the reporting challenges inside individual operating units were not designed to address this multidimensional information slice. Addressing this type of information challenge requires business intelligence methods and techniques that look across multiple business dimensions and "mine" the data for its informational content — turning information into insight and, in the case of a supply chain, finding opportunities where costs can be taken out and top-line revenue can be added.

Corporations have a lot of experience with a similar multidimensional problem — customer relationship management (CRM). In the early 1990s, firms learned that they had multiple relationships (or multiple dimensions) with the same customer, yet they would treat that customer, from a sales and marketing point of view, as if each operating unit with part of the relationship "owned" the customer. This not only confused customers flooded with promotions to cross-sell and up-sell into other related divisions, but treating that customer in that way typically costs more than that customer's profit contribution to the firm. To develop a more meaningful relationship with a customer, companies began to collect data across the enterprise about the stimulus and response behavior of customers, segmented those customers into like clusters of behavior and then treated those customers most efficiently relative to their ability to return profit.

There are still a lot of efficiencies to be gained in the efficient treatment of customers, and CRM professionals admit they are somewhere in the middle of their journey in solving this problem. But in the few years that this methodology has become a corporate staple, companies have made impressive gains in both adding top-line revenue and cutting the cost to treat. The breakthrough was using business intelligence methods to better understand the customer's needs — taking information slices through the enterprise to develop customer intelligence.

Supply chain practitioners consider that the intelligence required to efficiently source, make, store, ship, and ultimately sell and deliver the product or service to a customer is an area of opportunity not well-leveraged in most companies. We are seeing evidence in our customer base of centralized corporate function beginning to take "horizontal" views across operating divisions of the enterprise to look for additional cost efficiencies.

Strategic sourcing is a pay and how much inefficiency or "slack" could be eliminated from the system by taking a multidimensional approach across the enterprise.

Another trend motivating the need for multidimensional views of the supply chain is the drive to look not just at the internal workings of a single firm, but to look at the performance of the supply chain across multiple firms that comprise the entire end-to-end supply chain. The inspection of both process and cost structure from the supplier's supplier to the firm's supplier to the firm to the firm's customers and to their customer's customer through to the end user is key in taking out redundancy not valued by the consumer. Examining supply chains and how they compete against other supply chains redefines corporate competition.

Collaboration between trading partners, using business intelligence methods, and seeking efficiencies become core competitive skills throughout the supply chain. So the drive for efficiencies is a battle on two fronts: to get the firm's internal house in order and reduce costs in the supply chain by looking across multiple internal divisions; and to reduce supply chain costs by looking across all the elements that make up the supply chain — from raw materials to ultimate consumers.

The need for response-based supply chains that can adapt continuously to changes in demand, logistics situations, inventory positions, manufacturing backlogs, and materials acquisition has fueled the trend toward intra-company collaboration. This sensitivity to changing conditions is best illustrated in event-driven management. The motivation is the opportunity to: never miss a sale; take costs out of the system on very short notice; and look for ways to delight existing customers.

In exploring more efficient techniques to treat customers, companies have learned that the data has to pass through multiple processes in its journey from raw observation to business intelligence. Developing one version of the supply chain depends on having the right set of information technology competencies to handle large data repositories that allow for complex views of the data, the right set of computing technologies that allow for massive amounts of information to be processed quickly and, finally, a business intelligence software platform to perform the analysis to gain visibility into optimal supply chain design and execution configurations. We call this multidimensional capability supply chain intelligence.

Supply Chain Intelligence

A supply chain could have many definitions depending on the industry, but most industries include a supply chain made up of a set of geographically dispersed facilities where raw materials, intermediate products, or finished goods are sourced, manufactured, stored, transported, and sold. A supply chain network represents the flows of goods and services through these five general activities, but equally important, the supply chain network represents the information flow through these same activities.

Executives today are faced with a tough economy, increased competition, fickle consumers and a shrinking product lifecycle, requiring smarter, more strategic decisions along the supply chain. Supply chain intelligence reveals opportunities to reduce costs and to stimulate revenue growth. This ability to add top-line growth, combined with the capability to take a multidimensional view across the enterprise to remove inefficient costs, distinguishes supply chain intelligence systems from either ERP or SCM systems. Supply chain intelligence takes a broad view of the entire supply chain to reveal the processes that truly add value from the customer's perspective throughout the product lifecycle.

The need for both descriptive and prescriptive supply chain models that extract information from ERP systems and other supply chain transactional systems has never been greater. Traditional supply chain management solutions reflect the way most organizations currently think of supply chain functions: make, store, and ship. Supply chain executives understand all too well that their solutions must include an accurate interpretation of demand across all product lines, and that an important source of cost control comes from the strategic sourcing process. Once demand is added to traditional supply chain functionality, demand management becomes paramount.

This demand intelligence capability lies at the intersection of CRM and the supply chain that serves that customer and is a core component of supply chain intelligence. Similar logic applies to strategic sourcing and procurement: By adding supplier intelligence to supply chain intelligence, we immediately begin to think about the management of the sourcing process and the relationship with key suppliers.

Elements of Supply Chain Intelligence

Webster's Dictionary defines "intelligence" as "the capacity to apprehend facts and propositions and their relations and to reason about them." Combining this definition with that of a supply chain presented previously describes a capability requiring insight and understanding of some complex processes linking demand, logistics, storage, manufacturing, and sourcing. The words "insight" and "understanding" are necessary because it is not enough for executives to get answers from a computer system on what their positions are along a supply chain; the information needs to contain an explanation of how to harmonize the main interacting elements of supply chain design and supply chain execution. Following are brief descriptions of core supply chain intelligence components:

Demand Intelligence

The optimal supply chain is built around the needs of the end customer. The perfect supply chain would include only those activities that the end customer values and is willing to pay for. Therefore, a keen understanding of the customer's behavior is the first step in developing an accurate supply chain intelligence capability. Demand intelligence can be broken out into three areas that provide insight into the customer:

• Demand management
• Pricing optimization
• Inventory replenishment

Demand management for a consumer products company would include those functions that help the firm develop effective trade promotions with its retail trading partners and develop a forecast of demand and product or service usage for replenishment. This can be done by accumulating point-of-sale and other sales factor data and using various computational techniques such as time-series analysis or causal methods such as regression analysis.

An additional degree of complexity sets in when multiple products need to be forecast within multiple product families in multiple geographic locations for multiple time periods within multiple market segments. In companies with large numbers of products in the marketplace, this multiplicity can make accurate forecasting formidable. Having an automatic high-performance forecasting capability that feeds an accurate interpretation of future sales into replenishment systems that help set inventory policies is a vital component in maintaining customer satisfaction and stocking properly.

Pricing optimization relies on accurate demand forecasts and observations of customer behavior. Pricing optimization seeks to set a price for a product that is acceptable to the customer that also maximizes the firm's profits. Pricing optimization variables include demand, the cost of the product or service, available inventory, the velocity of recent sales of the product or service, and the profits or revenues the firm requires. As these elements adjust over time, the pricing optimization functionality should allow prices to adjust accordingly. The dimensions described in demand management are the same dimensions that must be computed for price optimization applications.

Inventory replenishment planning takes the information contained in demand management and pricing optimization to determine the optimal inventory position. This inventory position is a trade-off between capacity utilization objectives and customer service. An accurate demand forecast can dramatically lower the cost to serve. The optimal inventory decision will take into account this stochastic information and produce policies that will drive inventory stocking rates, lead times from manufacturing and what gets manufactured when, product quality, and order completeness.

Figure 2 illustrates the cost advantages of having a better forecast versus achieving a certain level of customer service. The goal is to achieve a desired customer service rate at a lower cost.

Figure 2: The Value of Achieving a Better Forecast

Logistics Intelligence

Logistics intelligence takes the changing "landscape" in demand intelligence and determines the optimal response, maximizing customer value and how it affects manufacturing production capacity and schedules, the logistics network, and inventory policy. Logistics intelligence in this description consists of two functional areas:

• Supply chain network optimization
• Maintenance and spare parts logistics

Supply chain network optimization looks at most, if not all, of the components of the supply chain of a firm at one time and determines the optimal resource allocation across the enterprise for a given demand pattern. This process takes into account existing and potential suppliers, facilities, their capacities, the location and buying patterns of markets, the products that flow though the network, the raw materials used to make the product, the steps in the product assembly process, the resources used to produce, store, and transport the product from a cost basis across each activity, and the transportation network and the alternatives.

The model's outputs are in the form of an optimal scenario for a time frame and would include the optimal quantity of raw materials required as input into the sourcing strategy; the optimal levels of manufacturing activity and resource consumption for each facility as input into the production strategy; the optimal movement of intermediate and finished product from manufacturing facility to distribution center to final customer markets; and the optimal configuration of facility location. The supply chain network optimization models form the basis for strategic and tactical supply chain design across most industries. Figure 3 is a typical graphical representation of an efficiently designed supply chain resulting from a single scenario of a supply chain network optimization model.

Figure 3: Output from a Supply Chain Network Optimization Model

Maintenance and spare-parts logistics is a wide topical area of which this paper will only touch on a specific aspect of the overall function. A brief description of what the U.S. government is doing relative to an initiative entitled "condition-based maintenance" illustrates some of the modern thinking in this area.

Conditioned-based maintenance is a form of proactive equipment maintenance that forecasts incipient failures. Maintenance actions are based on real-time or near-real-time assessments of equipment condition obtained from embedded sensors. It contrasts with reactive (run to failure) and preventative (scheduled) maintenance concepts. The goal is to improve the military's operational readiness and mission reliability while reducing costs.

Condition-based maintenance capabilities allow the detection of changes in the operating parameters of a piece of equipment that will allow the prediction of an impending failure to a part or predict the part's operational life. With this information, corrective maintenance can be scheduled rather than waiting for a system failure to occur. The intelligence procedures for condition-based maintenance use trend analysis and expert recommendation software to drive predictions of failure and to recommend alternative courses of action and maintenance activities.

Supply-and-Demand Alignment

In our supply chain intelligence model, supply-and-demand alignment will refer to the ability to understand two sets of metrics:

• The cost to serve across the supply chain
• The performance management metrics that measure best practices across the supply chain

In the previous description of supply chain network optimization, the cost structure along all aspects of the supply chain was a significant consideration. Laying end to end all the processes contained in a supply chain and assigning a cost to each of the tasks or activities as a product moves through each process would provide the total cost to serve a single customer using that supply chain. Activity-based costing assigns cost to an activity as it utilizes resources within the supply chain. A capability that combines the cost to serve on an activity basis and the ability to develop complex supply chains in a graphical and logical manner would lie at the intersection of supply chain design and supply chain collaboration.

The measurement of the effectiveness of supply-and-demand balance within a firm is probably best represented by the supply chain operations reference model (SCOR). The scope of the SCOR model includes all processes from the initial demand signal (an order) to the final signal that demand has been satisfied (the payment). The SCOR model is specific to a product or a family of products and is a measure of performance and process effectiveness. One of the tenets of the SCOR model is that a supply chain must be measured and described in multiple dimensions. These dimensions include reliability, responsiveness, flexibility, cost, and efficiency of asset utilization.

The SCOR model is a cross-industry model that decomposes the processes within a supply chain and provides a best-practice view of supply chain processes. The model is typically implemented in most organizations by developing a geographic representation of their supply chain (similar to Figure 2) and then converting the geographic view into a process-flow view with specific measures for each process component. The model becomes potent when used as a collaboration tool to improve supply chain processes between trading partners that comprise the end-to-end supply chain.

Production and Process Intelligence

This addresses the key manufacturing concern — quality. This quality issue is manifest in the increasing complexity of industrial processes. Early risk assessment of process abnormal deviations by having the data to determine the root cause of failure analysis is a vital component of quality and has enormous potential for savings and higher overall plant efficiencies.

Higher quality products usually lead to significant increases in market share. The supply chain intelligence core components to achieve quality are:

• Production or process intelligence
• Product warranty

To illustrate the benefits of production or process intelligence, the lifecycle of a drug offers a pertinent example. The lifecycle of a drug is somewhat different from that of other process industry products. It typically takes 10 years to turn out a potential new drug after testing thousands of new compounds. The potential new drug must be tested for both safety and efficacy. This involves a variety of trials for toxicity and to alleviate disease. Finally, the process-development activity produces a chemical or biochemical to manufacture.

The manufacturing process produces the active ingredient, which normally involves chemical synthesis and separation to yield a complex molecule or, in the case of a biochemical, fermentation and purification. Material from the process must pass quality control checks prior to being used downstream in the final set of processes. The need to avoid cross-contamination of products and the requirement for validated cleaning and changeovers result in long downtimes between product manufacturing. All of these tasks must be documented for compliance with Food and Drug Administration guidelines. All of the data, which comes from a multitude of sources across the enterprise, must be collected and tracked for each step of the processes described above. The compliance requirements track the manufacturing processes at each step from molecule to retail shelf for quality.

If production and process intelligence provides the view of quality from manufacturers' perspective, the warranty guarantee provides the view of quality from consumers' perspective. Looking at the consumer products industry, we find that a product warranty has substantial value to the consumer. In effect, it provides an "insurance policy" that dissipates the risks in purchasing the product.

From the consumer goods manufacturing perspective, the intelligence gathered from warranty data can help deal with issues such as which warranty claims are fraudulent, what are the new issues of quality facing product returns, how much money should be set aside for claims and whether failures can be predicted before they occur.

Supplier Intelligence

Supplier intelligence is the ability to understand the relationships between the firm and its major suppliers. This article began by articulating how getting a true understanding of supplier relationships across the enterprise is one of the best ways to take costs out of the supply chain. But this understanding requires using business intelligence methods to take a multidimensional view of the data across the enterprise. Supplier intelligence is comprised of two major functions:

• Supplier relationship management
• Materials and cost risk management

Getting an accurate picture of total purchasing spend is difficult at best, and in a large enterprise it is formidable. Industry analysts have identified supplier relationship management as a prime source for reducing costs and take-out within the supply chain. Spend analysis includes an intelligence capability to understand buying patterns and their trends, usage patterns for particular products, and how supplier dependencies have changed over time.

The intelligence surrounding risk management has great opportunity for cost efficiencies. Risk management in this context is an ability to lock in the price of a highly valuable raw material or component. A candidate component is somewhat scarce, essential to the product or service, and has experienced considerable price volatility recently. The ability to lock in the price of this precious material could be the difference in most cases between profitability or loss in that product line.

"One Version" of the Supply Chain

The elements of supply chain intelligence have been articulated in this paper and have a dual use. Some firms will want to execute around individual applications that can handle pinpointed and targeted tasks without wanting to implement an entire system. Other firms will need and want to do more. Underlying all of the applications described in the prior section is the data architecture that will be referred to as the Intelligence Architecture.

Figure 4 : Business Dimensions of the SCOR Model

Figure 4 represents an example of an Intelligence Architecture. This architecture draws from such data sources as: