Taming the Bullwhip: Driving Supply Chain Efficiency with Forecasting

Taming the Bullwhip: Driving Supply Chain Efficiency with Forecasting

Best-in-class demand forecasting :

  1. Is consumption-driven,
  2. Incorporates comprehensive data inputs, and
  3. Leverages technology and artificial intelligence.

What’s the result of the above elements? The ability to generate timely, dynamic statistical demand forecasts at a level of precision that builds confidence across the organization. In fact, consistently generating a high-quality, high-confidence, high-accuracy demand forecast is a critical first step in improving total company performance. Why? Because the demand forecast is the starting point for all upstream supply chain planning and execution.

By beginning with a more accurate picture of the end in mind, food companies—especially those operating in fresh and perishable categories—can accelerate business processes and operating improvements that lead to competitive advantage. Here’s how:

 

Sales & Operations Planning arrives at a single version of the (forecast) truth that guides the organization

Sales & Operations Planning (S&OP) is a structured, regular process that aligns all functional areas under a unified set of assumptions in order to enable and coordinate decision-making. S&OP integrates demand, supply, operations, sales, marketing, and financial planning into a single game plan for the business—along with linking strategic and operational plans—and provides a decision-making framework to maximize sales and profit. 

Best practice S&OP starts with an unconstrained demand-driven forecast. Through cross-functional collaboration, it builds to a consensus demand plan that sets the overall level of upstream supply chain operations output and other activities to best satisfy the sales forecast. What’s the result? An enhanced ability to meet the general business objectives of profitability, productivity, and customer satisfaction.

 

The higher quality the S&OP starting point (the demand forecast), the higher quality and more reliable the output.

Leveraging big data analytical platforms, machine learning, artificial intelligence, and agile forecasting models to quickly and efficiently incorporate a wide array of disparate demand signals generates significantly higher-quality, more reliable forecasts—for both baseline demand and demand-shaping activities. An upside to the use of analytical platforms: this “starting point” forecast is unbiased by manual overrides, functional goals, perceptions, and misconceptions. Furthermore, the time spent gathering, updating, integrating, and reconciling competing data from multiple disconnected spreadsheet forecasts generated by various departments is eliminated (not to mention the potential for human error!).

Beginning an S&OP session with a statistically sound forecast (one that is supported by facts and proven estimating techniques) takes emotion out of the cross-functional dialogue that is so essential to arriving at the consensus plan. The result, of course, is that the whole process is both accelerated and elevated. Further, with world-class forecasting models, robust what-if analyses can be conducted quickly and transparently, allowing the S&OP team to proactively consider multiple scenarios and upside/downside risks. Importantly, the transparency and visibility of all inputs and assumptions driving the demand forecast generates confidence across the organization and eliminates compensating manual “adjustments” by functional departments in the upstream supply chain. The result: alignment around a single version of the truth that guides the entire organization.

 

The Bullwhip is tamed

Supply chain analysis has shown that consumption patterns (true demand) are significantly less erratic than the ordering patterns of those in the upstream supply chain (retailers, wholesalers, producers, and suppliers). Without alignment and visibility around a single, truly consumption-driven forecast, each upstream process generates its own forecast. Biases, errors, and “safety stocks” are compounded along the way, as each forecast reflects the order history and patterns, price fluctuations, and real or perceived availability of inputs of each immediate downstream member across the supply chain. This creates what is known as the “bullwhip effect,” in which a +/-5% change in actual consumer demand can impact upstream suppliers by as much as +/- 40%.

A highly visible, high-quality, and high-confidence demand forecast helps tame this bullwhip—reducing both cost and waste—while still achieving consumer and customer satisfaction goals. Here’s how:

 

1. Inventory and out-of-stock risk are simultaneously reduced.

Having the right product in the right amount—and in the right place at the right time—with the maximum possible remaining shelf life is critical, especially for highly perishable foods. To move closer to the “gold standard” of continuous replenishment at point of sale, upstream inventory must be aligned with downstream demand: finished goods, raw materials, ingredients, packaging, etc.

Better understanding of expected sell-through, of course, facilitates production scheduling, capacity planning, and material resource planning. For fresh produce, this means planting, harvesting, sourcing, and packaging plans can be more effectively aligned with true consumer demand. For fresh prepared foods—for which variety and selection are important for category success—each retailer has their own unique “recipes”, and the item assortment may vary by store due to shopper demographics, in-store “finishing” constraints, and more. Understanding consumption demand by store, item, day of week, and time of day reduces out-of-stocks and allows producers to reduce order-to-delivery lead time while optimizing line and labor utilization and product freshness.

More accurate demand forecasts support a holistic approach to inventory management, allowing optimal safety stock targets to be established by item across all tiers of the supply chain based on lead time, in-store processing requirements, expiration dates, and demand volatility. This further reduces working capital investment while meeting customer service requirements. Ultimately, improved forecast accuracy can provide the confidence needed to source ingredients and schedule production in anticipation of customer orders, leading to improved order fulfillment, operating margin and GMROI.

 

2. Production plans more efficiently align with inventory and customer service targets.

Balancing production efficiency with demand and inventory requirements is challenging as consumer tastes become more volatile, product assortments grow increasingly complex, and shelf life becomes shorter. Thus, the more dynamic and precise demand forecasts become at a granular level, the more production teams will be able to anticipate, plan, and build finished and work-in-process inventories across all plant locations and lines. What’s more, these inventories will adequately consider:

  • Use-by dates
  • Complementary items (i.e. sauces, dressings, utensils, packaging, etc.)
  • “Finishing” requirements and constraints at each ship-to location
  • Internal equipment change-over requirements
  • Production sequencing
  • Raw material/ingredient lead times

 

3. Distribution and logistics plans are optimized.

It’s a simple fact: full trucks are less expensive to ship than partially filled trucks. But many products—especially new items, specialty items, seasonally available products, and individual SKUs of fresh prepared foods—do not have sufficient demand on their own to “fill the truck.” Optimizing transportation costs by “over-shipping” to fill a truck, on the other hand, results in shrink and waste. With high-quality demand forecasts across all SKUs, producers have the ability to pre-plan and build mixed loads that optimize distribution and logistics efficiency, while increasing delivery frequency, remaining shelf life, and freshness. In addition, demand forecasts by location help ensure that inventory is optimally positioned across the entire distribution footprint—both internal and external—in order to optimize service levels and delivery lead times.

 

Growth opportunities are illuminated and quantified.

True consumption-driven demand forecasting is critical, even in supply-constrained categories. We often hear from growers and other food producers that they don’t need to forecast demand because they do not grow/raise/catch enough product to supply all the marketplace demand that they perceive exists. Therefore, they allocate supply to their customers, typically on either a basis of a percent of total requested or a percent of total volume represented the prior year. But the reality is that forecasting true consumer demand in this situation is just as critical to current and future success, if not more so!

If customers know—or perceive—that they will be supplied only a portion of their total requirements, they may game the system by ordering more than they really need so that their “ration” is closer to their expected true consumer pull-through. Basing demand forecasts solely on customer orders gives suppliers no visibility into “real” demand for their product. Understanding true consumer sell-through, however, allows the supplier to strategically allocate available short-term supply while judiciously planning additional capacity for longer-term supply growth (including by land acquisition, geographic expansion, incremental contract growers/sources, plant improvements, labor growth, or other means). Communication between producers, suppliers, and customers is essential. Sharing information about demand forecasts, capacity and inventory helps alleviate customer anxiety and lessen the need to engage in “gaming,” while strengthening the partnership.

 

A demand forecast is not just a projection of future business: it is a request for product and resource deployment across the organization in order to maximize marketplace opportunities. Given how central the demand forecast is to upstream supply chain operations, it’s clear that consumer-driven demand forecasting must become a regular and highly visible process that is performed efficiently, rapidly, and accurately by leveraging the latest data, analytics, and technology available.

Fortunately, with Crisp technology, it’s easy—and fast—to make the leap from spreadsheets to high-speed, automated forecasting that incorporates all available demand signals and uses the latest forecasting models and machine-learning algorithms. Why? The Crisp demand forecasting platform is 1.) cloud-based; 2.) able to gather, organize, and analyze trillions of signals from disparate sources; and 3.) driven by computing power that provides actionable data. Even better: it can be set up in 10 minutes—and customized forecasts can be generated in even less time!

 

Ready to get started? Beginning is easy, output is fast, and impact is high, which means the decision is simple. Book a demo or contact us to learn more about how Crisp consolidates every byte of your data for optimal forecasting.