Across industries, companies that rely on transported materials for their operations have gained hard-earned knowledge from major disruptions over the past couple of decades, most notably the financial crisis of 2008, the COVID-19 pandemic, and geopolitical developments. During times of sudden upheaval, companies must quickly ramp down operations and then ramp up again once demand bounces back, adapt their execution, and revisit long-term plans. The supply chain has a special role to play, as companies rapidly shift their focus from cost-effectiveness to maximized throughput—all with the same assets and infrastructure.
A poor response can have cascading effects, such as facility shutdowns or missed delivery obligations. In recent years, approaches that use data and analytics to identify answers and make recommendations to specific business problems have proved to be particularly relevant in bringing clarity to operations planning and thus improving supply chain resilience. However, the vast majority of companies have yet to implement such approaches in their supply chain, leaving a serious gap in their planning capabilities.
A recent McKinsey article1 examined three value chain approaches that can address this supply chain gap—simulations of reality, optimization of plans, and real-time control-tower monitoring (see sidebar, “How analytics supports different planning approaches”). Here, we examine the second approach in more detail, as it ranks among the most powerful tools businesses can use to navigate complex and changing environments, especially disruptions.
How analytics supports different planning approaches
Supply chain analytics can support planning efforts that broadly fall into three types. While optimization has been at the center of this article, two other tools are simulation and monitoring. All three require expert knowledge of the system, but simulation also relies on large historic data sets (exhibit).
Optimization models are prescriptive analytics tools. The main output is an optimal plan for the current environment. Optimization is most useful when an organization must create an ideal plan from scratch that factors in complex rules and constraints. A typical use case for optimization is building a monthly plan for operational production as part of the sales and operations planning process.
Simulations are descriptive analytics tools. In a simulation, a digital twin is constructed with uncertainty measures and operational inputs. Simulation is more appropriate for comparing or modifying existing plans in the face of uncertainty. Whereas optimization is more of a “black box” process, simulation produces explainable KPI-driven reports. A typical simulation use case involves examining an existing material network with uncertain production quantities, demand, and transit times to identify first- and second-order bottlenecks.
The third analytics approach is real-time monitoring. Most companies have a control tower serving as an operating nerve center. However, fully digital end-to-end control towers can increase resilience by expanding the scope of real-time monitoring to anticipate and respond to upcoming or potential supply chain disruption. Real-time monitoring is the most useful when seamless communication between various functions and integrated decision making are critical.
A combined approach is often best. For example, consider detailed railway scheduling. The complex rules and interactions require optimization methods to create valid plans, while simulation can then validate the performance and robustness of plans and determine the most effective delay-mitigation policies. Real-time monitoring ensures seamless communication between individual hubs in case of unexpected disruptions.
Organizations that want to get the most out of this powerful approach design their associated optimization tools and processes along five best practices: improve information flows between teams, elevate customer centricity, bridge the gap between long-term planning and day-to-day operations, understand true operational constraints, and use scenario analysis to ask critical “what if” questions.
As companies in manufacturing industries have discovered, following these design practices can increase supply chain throughput by 10 to 15 percent in the short term, with no change in assets or overall configuration. Organizations can also reduce costs by 5 to 10 percent and CO2 emissions by 10 to 15 percent over the long term while increasing operational flexibility and resilience toward disruptions. This article details the design best practices that support this effort and how companies can get started integrating the necessary capabilities into their business.
Optimization in sales and operations planning
Optimization in operations planning involves determining the optimal choices for a set of decisions in a given business environment and business target. This type of optimization generally works best with prescriptive models that provide the ideal set of decisions as an output. The elegance of optimization is its ability to adapt to not only changing business environments but also shifts in the business target—for example, from minimum cost to maximum throughput, highest yield, zero environmental impact, or a combination of multiple factors.
For these reasons, optimization is the ideal approach to readjust a company’s operations as outside factors or strategic priorities change. For example, an agricultural company recently experienced dramatic rises in production costs combined with limited transport capacity, creating significant gaps in its ability to manage existing resource constraints. The company was able to respond by shifting its operations planning to an optimization approach that almost completely closed this gap.
Would you like to learn more about our Manufacturing & Supply Chain practice?
In essence, organizations can embrace optimization to support better, faster planning and thus increase value capture and resilience. Five best practices can sustainably improve supply chain decision making across the full coordination process (exhibit). To do so, the design of the corresponding optimization tools and processes must address all decision layers—from strategic orientation to operational execution. The best practices for optimization design explicitly address distinct layers and can thus enhance transparency and planning effectiveness.
1. Elevate customer centricity by enabling a pull principle for demand-driven operations
Customer-centric thinking is a foundational element for any type of operations planning. Tools and processes must be able to quickly translate demand into sequence and volume of production units and share this information with individual production sites. Optimization tools and processes can make these decisions rapidly and objectively, enabling both automated plan adjustments based on changing customer demand and faster responses by contact center agents to customer requests. Satisfying those requests can have a significant impact on operations. A commodity metals company was able to accelerate its decision-making process and thus more quickly react to urgent customer requests resulting from demand fluctuations caused by rapidly changing spot prices.
2. Bridge the gap between long-term planning and day-to-day operations
Short-term plans are typically shaped by current constraints, while long-term plans depend on outside factors that are likely to change. Effective operations planning addresses the in-between period, in which value can be lost in the gap between tactical vision and concrete implementation. Optimization tools and processes can translate mid- and long-term plans into detailed operational schedules, automatically accounting for dynamic conditions and complex operational constraints while allowing users to refine and explore possibilities. The gap is bridged in the other direction by feeding information on actual operations and deviations from control-tower monitoring back to optimization or simulation tools to continually minimize any loss in value. The agricultural company mentioned above was able to ensure its day-to-day operations remained in line with the long-term strategic goals and tactical monthly plans despite significant changes in the overall business environment.
3. Improve information flows between operations and marketing and sales
Organizations need to enable a consistent flow of information among stakeholders. Integrated supply chain planning tools already bring together information from multiple systems and business functions, creating transparency and empowering decision makers while enabling analytics tools. Optimization tools and processes can be added to improve decision making. A tangible example of this interplay can be a sudden change in production capacity due to an unforeseen dumper breakdown. Integrated planning tools ensure all relevant stakeholders are aware of this change, but only optimization tools will actively steer and synchronize the decisions of the entire organization toward the ideal target in this new environment. Actions could include modified marketing and pricing of the respective products or the realignment of the supply chain. Comprehensive control-tower dashboards are then used by teams from various functions to support real-time decision making.
For instance, a mining company was able to quickly and regularly make updates to its product portfolio based on recommendations from the marketing and sales team. This coordinated rapid response enabled the company to keep its portfolio closely aligned with market demand at all times.
4. Understand true operational constraints by dissecting the infeasible plans
The exercise of identifying and avoiding infeasible plans often leads planners to review and adjust their operations. This infeasibility can come in the form of forced stoppages when operators are tasked with following an impossible plan or situations in which plans are unable to meet all known operational constraints. Control-tower dashboards can aid in understanding these constraints and providing feedback. Optimization tools can then progressively integrate all these constraints into actual operations planning, thus providing consistently feasible solutions. In this way, the impact of optimization tools results in higher adoption, as they reduce the overall frustration level across the organization associated with these infeasible solutions. For one automotive company facing a shortage of semiconductor chips, a control-tower tool in combination with optimization-based processes generated more than $100 million in margin improvement.
5. Use scenario analysis to ask critical what-if questions
Planners must have an operational process that can run, understand, and evaluate scenarios for planning and scheduling. This process produces what-if questions that can inform discussions with sales, customers, and third parties and support better decision making. Optimization tools allow decision makers to focus on the “what if?” and receive immediate and risk-free feedback on the consequences, thus streamlining and lowering the barriers to asking insightful questions. Recently, a pharmaceutical company was able to improve overall throughput and on-time delivery by using optimization tools that enabled asking what-if questions regarding rush orders, staffing shortfalls, and capital expenditure investments.
Get started with an appropriate business opportunity
Many companies have yet to make significant investments in optimization for their operations planning. Often, organizations have not even undertaken the analysis to select the business opportunity. In addition, discussions around optimal business targets that cover the most relevant trade-offs are often ignored, as they reside between functions and are thus considered “off limits.” Picking the wrong operations element or business target for optimization diverts finite resources from what truly matters and represents the biggest risk of failure.
The key is to identify a process that satisfies three criteria: first, the overall decision space must be so large that an individual planner can’t explore and understand all the possibilities at once; second, the quality of outcomes must be objectively measured and assessed according to a well-defined and accepted target; and third, the potential for improvement must be measurable and ideally quantifiable—for example, costs, throughput, CO2 emissions, profit, or a combination of multiple factors. The mining company focused on profit by adjusting the supply chain in line with the optimal product portfolio, and the agriculture company emphasized the marginal costs of incremental production. The metals company selected a combination of throughput and yield while maintaining high customer satisfaction. The automotive company minimized lost revenue coming from supply chain disruptions. And the pharmaceutical maximized on-time, in-full delivery.
Companies must ensure their optimization program identifies business opportunities and associated targets that are not properly covered by the existing manual processes. However, any optimization-based approach must retain an adequate level of human judgment and expertise to account for unexpected situations and outliers. Optimization should aim to augment and empower human decision makers but not replace them.
Common pitfalls in data gathering and infrastructure
Companies should avoid directing time and resources to elements that aren’t necessary for high-functioning operations planning. For example, contrary to conventional wisdom, large longitudinal data sets are not required. Optimization for supply chain planning can typically be built by drawing on knowledge gained through expert interviews as well as snapshots using a small amount of the latest business data. Both elements are present today, as organizations need access to this information to monitor supply chain performance in the first place.
Future-proofing the supply chain
Similarly, companies should avoid focusing on building new data-centric infrastructure as an initial enabler for operations planning—either to capture and ingest data more frequently or to host new, complex analytics solutions. Instead, operations planning solutions can run infrequently, use small-scale data, and typically be implemented alongside and connected to existing tools and systems. In this way, these solutions are complementary to common end-to-end planning cloud solutions and maximize the value organizations can extract from existing resources.
In a recent case study, the large mining player developed an optimization-based process to improve the supply chain of its trains, vessels, and mine operations over short- and medium-term horizons. The process was aligned with the company’s strategic target regarding customer specifications and overall production goals. Embedding the optimization process into overall decision making allowed the company to improve operations significantly. This move alone captured many millions of dollars in cost savings.
In a world characterized by increasing volatility and major disruptions, the maturity of operations planning has the potential to increase the performance spread between first movers and laggards. So far, bigger corporations have taken the lead in implementing approaches to optimize their operations planning, given the resources and capabilities required. Whether optimization is within the reach of all companies is still an open question.
