A Shift in Supply Chain Planning Priorities?
For several years there has been a commonly held belief that focusing on improving demand forecast accuracy should be a top priority for Chief Supply Chain Officers (CSCOs). But according to recent reports by Aberdeen Group, we are seeing evidence of an important shift in this thinking. According to multiple analyst surveys, the fundamental problem - addressing demand volatility - is still the most challenging pressure on CSCOs. Aberdeen’s “CSCO Planning Strategies for 2017: Addressing Demand Volatility” report rated it their #1 challenge: above customer fulfillment demands, rising supply chain costs, and all other pressures (see Figure 1 below). Analysts like Gartner and Supply Chain Insights have published the same conclusion. But despite best efforts, improving forecast accuracy (or reducing forecast error) at many companies has hit a ceiling. Aberdeen reports that forecast accuracy has stalled in the past four years. Gartner analyst Steve Steutermann reports a similar finding that forecast errors remain unchanged or are getting worse. He says, “Using a traditional, 30-day lag period and a mean absolute percent error (MAPE) unit/location forecasting measure, Gartner often hears from companies with forecast accuracy percentages in the 50s and 60s, with far fewer reporting forecast accuracy in the 70s.” So you might therefore assume that CSCOs would continue to prioritize demand forecasting. But according to the Aberdeen research note, Best-in-Class CSCO Priorities Targeted for Improvement in 2017, it turns out that’s only true at average and low-performing companies, likely those companies still addressing fundamental flaws or shortcomings in their forecasting process. Instead CSCOs at best-in-class companies are signaling that demand planning is still important, but they can only get so far by focusing on their demand forecast. In a Forbes.com commentary, Lora Cecere spells it out this way; “While improving forecasting sounds like the right answer, and companies need forecasting capabilities, what I see working today is not as simplistic as improving forecasting. Instead, it requires a holistic look at inventory. A mistake that I see companies making over and over again is improving forecasting, but not improving inventory levels because of the lack of a holistic focus. After the implementation of a forecasting system, and getting basic functionality in place and providing a reliable forecast with minimal bias and accuracy, there is only minimal value in continuing to refine the forecast. Most companies do not have a forecasting problem. Instead, they do not know how to use the forecast to drive inventory strategies.” She concludes, “Does better forecasting improve inventory? I don't think so anymore.” So according to Aberdeen, what priorities are best-in-class (the top 20%) companies turning their attention to? Many have concluded that the best defense is not the pipe dream of a perfect forecast but rather the gritty realism of optimized inventory. The study says best-in-class companies are 60% more likely (than average and low-performing companies) to focus on inventory optimization (IO), because of its ability to deal directly with unpredictable demand and optimize service levels. Today’s end customer decides “where, when, and how their orders are fulfilled,” Aberdeen says. “This puts traditional channels in a constant state of flux, with a higher percentage of demand constantly shifting, affecting the actual vs. planned inventory needed to support sales for any channel.” Companies are using inventory optimization solutions that detail demand variability down to the SKU level to figure out what inventory is needed to meet specific target service levels. They model their inventory—and continuously refine those models as demand swings. “This is no longer an annual, or even quarterly process, but rather an ongoing, on-demand one, used to adjust inventory as changes occur,” says the report. Demand sensing is another priority, Aberdeen says. “Using demand sensing to better forecast demand by location, in conjunction with inventory optimization to adjust inventory targets as channels shift, will enable the best-in-class to target improvements and maintain a competitive edge.” Sensing demand is also underpins a pull-based replenishment approach. A pull strategy—combined with inventory optimization focused on safety buffers to meet customer service goals—helps CSCOs counteract demand volatility. Aberdeen reports that demand analytics are also favored by best-in-class companies—68% of these firms have analytics programs in place to analyze changes to demand volatility and unearth emerging patterns to better manage service and inventory deployment. Demand segmentation models combined with analytics track sales channels and spotlight variations for dynamically optimizing inventory on the fly. The consensus around the demand volatility remains, but there are new ways of thinking now about how to defend against it.