Demand variability is a problem even with fast moving items (shown on left),
but especially with slow moving items (shown on right) and promoted goods.
According to the new 2016 Gartner-SC Digest Supply Chain survey, forecast accuracy and demand variability are the top barriers preventing their companies from reaching their supply chain goals, according to the highest percentage of respondents.
A new distinction between “micro” and “macro” supply chain agility helps explain why. For the first time, the report differentiates between “micro agility”—the ability of a supply chain to react nimbly and successfully to sudden changes, disruptions, and opportunities caused by things like forecast error, production snafus, late shipments, weather problems, and unexpected demand—and “macro agility”—the ability to align the supply chain with new company strategies and initiatives.
A significant number of companies surveyed reported bigger challenges around short-term micro agility than longer-term “macro” supply chain agility. More than 15% stated they had low micro agility— while only 5% claimed they had very high agility.
Survey respondents said the underlying reason micro-agility is a problem is the lack of visibility into real-time data. They say this lack of visibility was their biggest agility barrier.
Most demand forecasting systems don’t account for the latest inputs and data. They focus on historical trends that can be months old. Hence, they often produce disappointing results and significant forecast errors because their standard models can’t identify recent trends. This inability to model the underlying causes of demand variability leads to poor trade promotion execution and less successful new product launches.
To get a better handle on real-time data, many firms are turning to technology. In the survey, enhanced analytic capabilities was the top-ranked technology-driven supply chain planning improvement in terms of expected benefits.
An advanced analytics approach uses more intelligent software to improve forecasting accuracy and better manage demand variability—sensing and analyzing demand at the customer source, and extracting insight from that downstream demand data to improve near-term forecasts. Advanced analytics go beyond traditional forecasting methods to an approach where companies “will increasingly be able to predict potential problems or opportunities,” the study says.
Advanced demand analytics are enabled by things like pattern recognition, predictive analytics, and machine learning—the ability of the software to learn from past forecasting successes and errors to improve its approach. In today's world, historical sales are old news. Noha Tohamy of Gartner says companies today must factor in multiple drivers, and unravel their complex interdependencies, to predict demand. “With machine learning, these connections can be discovered by analyzing unique patterns within the data,” she says.
A defining characteristic of advanced analytics is their ability to reveal these interrelationships, previously unapparent to the user. And there is plenty of “new” out there—from proliferating sales channels, to social media, to multiple staging options for order fulfillment.
The report concludes, “The potential power of advanced analytics solutions will perhaps trigger an inflection point in supply chain software.” Advanced analytics “offers new approaches to supply chain decision support, potentially upending, for example, traditional methods used for forecasting.”