Supermodelling allows a company to take an end-to-end view of its supply chain and make adjustments in production, distribution, and inventory practices to meet changing market demands
The outlook for global supply and demand is constantly changing, particularly under the current economic circumstances. The situation is fluid: Fundamental market dynamics shift, the balances and trade-offs in cost equations change, and solutions such as outsourcing and localised production may lose their value. In response, global companies reassess and reshape their supply chain networks and operations, taking future developments into account. No one can accurately predict the future, of course, but companies can plan ahead by defining potential scenarios, risks, and options, and then assessing the likely outcomes of each.
Manufacturers and logistics service providers often struggle with this complex task. They typically apply ad-hoc spreadsheets and inconsistent data-collection methods rather than tailored analytical tools and standardised procedures for gathering relevant data. The tools most companies currently use to help them analyse changing situations in operations and supply chains are not perfectly suited for the job. Most are either very detailed, making this exercise cumbersome and time-consuming, or they are limited and unable to analyze mid- and long-term scenarios. Moreover, many existing statistical models are based on a regression of historical data. But historical models are inadequate when fundamental supply chain parameters such as demand, markets, products, and cost drivers are facing significant and unprecedented changes.
Supermodelling offers a solution. This modelling method takes an integrated view of the end-to-end supply chain, from market-demand scenarios through order management and planning processes, and on to manufacturing and physical distribution. It studies historic and “as is” market and order data as well as “to be” market scenarios and demand forecasts. Such scenario-based simulation leads to better strategic supply and demand balancing because new products, expected price changes, and options for physical network changes are dynamically incorporated into the model. Finally, Supermodelling—conceptualised and tailored to companies’ specific business conditions—provides a fact-based approach for making difficult but necessary decisions that may encounter political and emotional resistance within the company.
How is Supermodelling different? In general terms, Supermodelling provides a computer replica of a real or planned supply chain system— what one might call a “model world.” The scope and content of the model—entire value chain, highly detailed breakdown of data, full transparency of feedback loops, and high reliability of options—is more comprehensive yet no more complex than the typical supply chain optimisation tools. Supermodelling’s broader focus can address a wider range of questions and issues, such as volatile commodity pricing and availability, shifting perceptions of market players, and conflicting trading or purchasing activities, that are not covered by traditional supply chain models.
Other types of supply chain optimisation tools improve physical networks by looking at transportation, distribution, and labour costs in isolation—an approach that may produce unexpectedly costly results. Supermodelling, on the other hand, not only examines physical production and distribution costs but also takes into account operations planning aspects such as supply management, manufacturing planning, and delivery management. In other words, it assesses the impact of various cost and value drivers, such as labour, transportation, technology, and productivity, on the entire network. Since Supermodelling can alter those parameters to develop different scenarios, it can demonstrate how those changes might affect customer behaviour or supply chain performance.
Supermodelling’s output is more visual than that of traditional supply chain optimisation tools. The level of detail it produces is based on a careful assessment of what is required to respond to both strategic and tactical questions. Outputs typically represent key measures in finance, performance of physical process flows and virtual information flows, capacity utilisation, stock levels, and customer service, all in the context of various rules and constraints imposed over time.
Accordingly, the model is able to demonstrate the expected benefits of reducing lead times by streamlining business processes, managing or reducing variance, and improving responsiveness and flexibility. This allows users to compare end-to-end supply chain scenarios—from quote to delivery to cash— with each other and with the current, as-is situation. A “supermodelled” replica of a supply chain thus provides the scope needed to determine the appropriate course of action based on future demand scenarios and trends.
The objectivity of this approach makes it a helpful tool for achieving consensus among stakeholders. Ultimately, Supermodelling enables “boardroom experimentation.” It allows companies’ management to test hypotheses and see a visualisation of the answer right away. They can use the simulations to identify how best to balance demand and supply, examining such options as whether—and when—to open or close factories, move production to a different location, or shift inventories between distribution centres.
For example, a global manufacturer of health care equipment used Supermodelling to optimise its order-to- cash process, including the physical network, from sourcing through manufacturing to the customer, as well as the information flows. For this company, Supermodelling was particularly powerful at speeding up and simplifying the decision-making process because multiple scenarios could be run with the participation of key stakeholders. They could instantaneously see the impact of proposed changes in the supply chain network and consequently make better, consensus-driven decisions, even when opinions regarding a particular situation had been divided prior to using the supermodel.
To obtain optimal results, a Supermodelling exercise should follow a four-step process
1. Establish a baseline, simulating the as-is scenario to validate and calibrate the model. Run a base-case simulation, applying projected demand over time by product and by region. Identify areas that will require new supply decisions.
2. Define and evaluate the major external trends that are likely to have a long-term impact on supply and demand.
3. Identify potential changes to investigate, with the aim of minimizing costs while maintaining the best balance between supply and demand. Set up scenarios that reflect those options in the model, and then simulate market and supply performance.
4. Evaluate results by comparing output in terms of key performance indicators. Continue iterations, with multiple runs and “what if” sensitivity testing, until the most effective solution becomes evident.
For a brief look at how one company applied this process in a distribution-network analysis, see the sidebar “Four steps to Supermodelling success.”
Future scenarios, year by year Companies that have employed the Supermodelling approach have been able to reduce costs and free up working capital. They also generally do a better job of matching supply to demand as it evolves from month to month and from year to year. This modelling approach allowed the health care equipment manufacturer mentioned above to avoid a costly physical supply chain setup in Asia and substitute a more cost efficient solution based on insight provided by the supermodel. A change in sourcing and better timing of order fulfillment deadlines allowed it to adopt a leaner physical distribution network, cheaper transportation, and fewer stockholding locations. The model not only delivered tens of millions of euros in potential savings but also developed a solution that the whole business could support.
A more detailed example is the case of Velux, a manufacturer based in Denmark. A major player in the global building materials sector, its products include roof windows and skylights, many types of decoration and sun screening, roller shutters, installation products, and remote controls and thermal solar panels for roof installation. Velux has manufacturing suppliers in 11 countries and sales companies in nearly 40 countries.
In 2007 and 2008, a project team consisting of Velux’s supply chain and manufacturing strategy specialists and supply chain consultants (including the authors) developed a dynamic supply chain model to study alternative ways of managing information and physical material flow between production, stockholding points, and markets. The primary objective was to build a supply chain model for strategic planning and evaluation of options that would be specific to Velux. Through evaluation of different scenarios, the model would support strategic manufacturing initiatives and ongoing sales and operations planning for the six-month to five-year time frame.
Today the Velux Supply Chain Model (VSCM) is used in the company’s windows and flashings group for long-term planning. The model is predicated on a baseline setup; its “basis year” employs actual production and sales data and incorporates future expectations for sales, productivity, and projected costs for raw materials, labour, and transportation. The model provides Velux with the ability to evaluate different scenarios by showing year-by-year development in capacity utilisation, product and/or component flows, and even investment costs.
VSCM has enabled the company to increase both the number of alternative scenarios to study and compare as well as the scope, relevance, and quality of the output. Velux has used the model to examine several European manufacturing and supply chain scenarios. On the basis of that analysis, the company made important decisions that significantly impact manufacturing, logistics, and financial value drivers, such as optimising production’s environmental footprint and designing a longer-term sales and operations planning process, to name just two.
Capitalise on change
Many companies are content to establish supply chain processes and structures, and then allow them to continue as is until they become obsolete and problems arise. Or they may find that business is developing in such a way that their existing supply chain operations and processes are unfit to exploit market opportunities or meet market challenges. In today’s competitive environment, that is no longer a viable way of conducting business.
If a supply chain modelling effort is to provide truly valuable decision support, then it must be based on a deep understanding of a company’s specific situation, requirements, and key issues. For this reason, it’s necessary to have a collaborative effort among model developers, analysts, the company’s own business experts, and other stakeholders throughout the supply chain.
What makes Supermodelling an appropriate tool for achieving that objective is that it gives companies that are looking to capitalize on change the ability to re-examine their production and delivery networks by taking into account cost and value drivers in the end-to- end supply chain, all within the context of future growth. This approach can make visible to company management the potential payback for selecting a particular course of action.
Four steps to Supermodelling success
A successful Supermodelling implementation requires working through four basic steps. The following is a real life example of how one company applied this modelling technique.
The company used to have a decentralized structure with stockholding delivery and service centres (DCs) in 16 European countries. Initial analysis indicated that a more centralized production and distribution process might significantly reduce costs without compromising service.
The primary challenge was to identify how many and which warehouses to close, and what level of service to offer from the remaining locations. That decision would have to be made in tandem with changes in order management processes and stockholding policies. To address those considerations, the project team developed a dynamic supply chain “supermodel” scoped to simulate and study alternative ways of managing information and physical material flow between production, stockholding points, and end customers.
The project proceeded as follows:
1. Development of a baseline case. The model was calibrated to simulate one year as-is operation, and a baseline scenario was created. Confidence in the model was established because it could simulate and reproduce history with numbers for total annual logistics costs, inventory levels, and delivery performance that approximated what had actually been the case that year.
2. Development of future trend scenarios. By manipulating inputs to the model, several simulated scenarios were produced, along with the associated effects on costs and service. This allowed the team to understand the issues and draw up a short list of realistic potential solutions.
3. Evaluation of options. With 16 DCs in the “as is” situation, a simple optimization exercise was infeasible. Complicating matters was the fact that all of the country directors were against closing “their” warehouses. It was important, therefore, to help them objectively evaluate and compare alternatives.
The model run proposed closing down various groups of DCs and simulated the outcomes. The model was tested for all kinds of “what ifs”; it proved and visualised how stock and transportation costs could be balanced more effectively, without compromising service to local customers. The solution that proved best—cutting back from 16 DCs to three— would allow the company to cover all of its markets in Europe while enjoying a 20-percent saving in logistics costs.
4. Evaluation of results. The model scenarios and options were developed by the company’s wider European logistics organization, and members of the group jointly selected the best solution. The modelling approach helped to establish consensus among the stakeholders, avoiding the dangerous route of making decisions based on political and emotional resistance to change.
Tim Lawrence is the global leader of PA Consulting’s supply chain and procurement services, Knud Erik Wichmann is Denmark’s practice leader in Decision Sciences.