Proactive Planning for Demand Uncertainty

Risk management is an emerging area of supply chain management. To address the myriad
risks facing companies today, i2—together with its visionary customers—has been developing innovative solutions such as i2 Sales and Operations Management, i2 Scenario Manager and
i2 Inventory Optimization.

In this article, Blake Johnson of Stanford University offers his perspective on this important issue.

The current economic environment has made all of us aware of the risks and market uncertainty that surround our organizations. The question many businesses are asking is, "How can we make managing these risks part of our daily operations?"

At the root of the problem: uncertainty in supply and demand is undeniable; however most core planning processes today are deterministic—a legacy from a more predictable era.

This disconnect is visible in operational performance on a day-to-day basis—in excess inventory, late deliveries, missed market opportunities and utilized assets—and in the financial performance this operational performance drives. The toll on supply chain teams is enormous. Changes in the plan typically mean long hours of fire-fighting, debates over accountability and strained relationships.

While the consequences of failing to proactively manage supply chain uncertainty are not new, the magnitude of their impact over the last 12 months has led many companies to conclude that now is the time for change in core planning processes.

If you can't measure it, you can't manage it

Most companies know what their key risk factors are, and also know the levers that drive their exposure to them. For many companies, sourcing and outsourcing relationships are among these risk factors. For asset-intensive companies, capacity and tooling decisions are often critical. And for nearly all companies, demand uncertainty is a key risk factor—whether it arises from short product lifecycles, market and economic uncertainty, changing technologies or increased competition.

Even if these risks are acknowledged, businesses face two key challenges when managing these risks:

Inability to quantify risks. Few companies have processes to quantify their key risks. While short-term demand variability is analyzed to set inventory buffers, the key planning decisions with the big dollar impact—the sourcing and capacity decisions identified above—are based on best-guess forecasts, extending months and quarters into an uncertain future.
Inability to quantify the financial impact of supply chain decisions. Most companies lack the ability to calculate the financial risk of their supply chain planning decisions. And if you can't measure it, you can't manage it. This gap is the number-one factor companies cite as limiting their ability to identify and implement the supply chain decisions that would best position them for an uncertain future.

Range planning makes it possible to measure—and manage—risk

The good news is that range planning processes that overcome these obstacles are now available. They rely on three key capabilities:

Range forecasts that quantify supply and demand uncertainty over the planning horizon
Range plans that prepare for the variety of possible outcomes captured with range forecasts
Range performance that quantifies the operational and financial performance and risk of range plans

Range planning is focused on a simple idea: identify the set of things that could happen, and plan proactively so that you can perform as well as possible across any of these potential outcomes.

We follow these same common sense steps in our everyday lives. For example, if there is a chance of rain, we perform a quick range calculation to decide whether to carry an umbrella. If the chance of rain is 90 percent, we'll definitely bring one. On the other hand, if the chance is only 10 percent, we may balance the burden of carrying the umbrella with the low risk of getting wet. The role of range forecasts in the range planning process is quantifying the "odds" of key supply chain risks in a similar way.

But the stakes matter as well as the odds. For instance, if we are going to an important job interview or are getting married, we might opt to bring an umbrella even if the chance of rain is only 10 percent, because getting wet would have a significant negative impact. Range performance enables the quantification of the stakes. In business, for example, a 5 percent stock-out risk might be acceptable, or even too low, for a very low-margin product—but extremely costly for a high-margin product.

In businesses that use range planning, range performance is the key output that the supply chain organization delivers to the rest of the company. Range performance specifies, for each possible demand outcome in the range forecast, "If this specific demand level occurs, this is the performance we will deliver." This makes clear accountability possible, even in the face of uncertainty. As a result, cross-functional and cross-organizational alignment on risk become possible in key processes such as sales and operations planning, as well as the management of customer, supplier and partner relationships.

Range forecasting, range planning and range performance require these key process steps:

Range forecasting: What's the longer-term weather outlook?

We all recognize that insight into short-term uncertainty—e.g., daily or weekly demand variability—is essential to making near-term decisions, such as setting the right inventory targets. In order to proactively manage risk over the planning time frame, uncertainty over future months and quarters must be quantified. It's not enough to predict the weather today; the role of range forecasting is to predict what weather is coming.

The key range forecast input is data about the accuracy of a company's current point forecasts—including error, bias and volatility by forecast horizon. This data quantifies the uncertainty, or predictability, of future demand over the planning horizon, and enables product-level range forecasts to be generated on an automated, scalable basis.

Range forecast accuracy can be further refined by slicing and dicing forecast error data by product line, stage of product lifecycle, past market or economic conditions, and other key drivers of forecast performance. For new, highly differentiated products or markets, expert judgment from marketing and sales may also be important.

Range planning: Should we bring an umbrella? Or sunscreen?

The role of range plans is to proactively plan for the future uncertainty captured by range forecasts. Range plans prepare the business by matching planning decisions to the level of confidence an organization has about different possible levels of future demand.

For example, the worst-case scenario defined by the low end of a range forecast represents a level of demand that the company has a high degree of confidence in. This confidence creates an opportunity to save costs. For example, plans for this level of demand can leverage lowest cost and long lead-time sourcing, as well as production and transportation strategies such as low-cost country versus domestic, ocean versus air, volume purchases, large batch sizes and full truckloads. These cost savings go straight to the bottom line.

On the other hand, the potential demand between the low and the high end of the range forecast requires flexibility. Because range forecasts quantify the amount of flexibility required by lead time, companies can create range plans that deliver this flexibility at the lowest cost, including postponing commitments to leverage better information as it arrives over time.

Range performance: Stepping out with confidence

The final step is selecting the range plan that best meets business goals. Range performance makes this possible by quantifying the operational and financial results of implementing a range plan under each possible supply and demand outcome captured by range forecasts.

Since traditional planning processes are deterministic in nature, they only provide visibility into the future performance that will occur if demand actually matches the forecast—which it almost never does. (It would be like meteorologists always giving their weather forecasts with 100 percent probability—and being correct.) Because companies have no insight into future results under other potential demand outcomes, they must "fly blind" as they attempt to manage off-plan performance risks. Because these risks can't be seen or anticipated, they often provide unpleasant surprises.

In contrast, range performance provides a complete, up-front view of future performance and risk across potential outcomes when that risk can be evaluated and proactively managed. Supply chain organizations that use range planning typically provide a "menu" of range performance options from which management can choose. For example, a range performance option may provide lower costs but less flexibility—and more risk—than another option.

Supply chain teams that present these kinds of range performance options to key stakeholders in general management, finance, and sales and marketing almost always get a very positive response. A typical comment is, "This is what we've always wanted from the supply chain—future performance options we can count on, with risk exposures and performance trade-offs we can see." All parties leave with clear alignment on the cost and risk trade-offs that have been identified as best for the business. Whether or not they choose to carry an "umbrella," they can step out confidently, knowing they have made the decision that best balances all key costs and risks.

Making it happen: Supply chain leadership is critical

The key to successfully institutionalizing supply chain risk management is organizational change. Making the management of uncertainty part of core operational planning processes requires changes in demand and supply modeling, supply chain planning, and performance measurement and reward systems. Because the supply chain organization owns the planning decisions that drive operational performance, only the supply chain organization can facilitate this transition.

The ability to proactively manage uncertainty in supply and demand—a critical success factor in today's economy—positions the supply chain organization in a leadership role and makes it essential to its counterparts in general management, finance, and sales and marketing, as well as to the future success of the company.

Blake Johnson is a consulting professor of Management Science and Engineering at Stanford University. He regularly works with leading companies on risk management and other pressing business challenges.