Organizations are often caught by surprise when commodity prices increase, causing significant financial distress. The better an organization is able to anticipate potential cost increases, the more likely they are able to implement safeguard mechanisms to improve supply chain control and reduce organizational risk. How an organization plans for these events and mitigates supply chain pricing risk can be categorized in three ways: ambivalence, conjecture driven, or data driven.
Ambivalence
An organization that is ambivalent to commodity pricing typically has little to no understanding of how future supply chain fluctuations will affect profitability. The inability to properly project costs leaves the organization in a reactive mode to vendor’s inability to fulfill shipments or soaring commodity prices.
An absent mechanism is a hallmark of startups and tightly held private companies.
Startups have typically not yet matured to a point where supply chain risk is well understood. Tightly held private companies may operate reactively and mitigate potential risk through reactionary procurement measures and leveraging personal relationships. Most organizations will push past this phase and adopt a conjecture driven model with time.
Conjecture Driven
In this model, there is a forward-looking focus to ascertain supply chain risks. The organization is likely keeping vendor scorecards, paying attention to world events, and paying at least some attention to industry trends. However, future commodity projections are likely to be ad-hoc and personality driven. These commodity projections may be based on conjectures of leading economists and simplified projection models built well before the advent of machine learning and artificial intelligence.
Different levels of adoption exist across the conjecture driven model. The model may be loosely relied upon in times of crises or may be well baked into the organization’s policies. Nonetheless, a sub-optimal approach to quantifying risk reduces the organization’s ability to deal with supply chain disruption.
Data Driven
In true data driven model, organizations are proactively planning for commodity fluctuations and adjusting revenue and cost projections accordingly. The advent of artificial intelligence to accumulate, adjust, and weigh large data sets allows for more accurate projections. Deep learning and kernel methods provide the complexity needed for commodity projections and return better results than traditional models. A true data driven model uses modern models that rely on thousands and even millions of data points to forecast commodity prices long into the future.
The data driven model does not mean that traditional modes of vendor evaluation and supply chain risk assessment are put aside. Rather, vendor evaluations and practical wisdom are taken as another evaluation point.
Barriers to a Data Driven Model
An effective implementation of the data driven model requires the following: a strong explanation of data driven benefits, the ability to anticipate organizational resistance, and proper scoping and commitment to installation.
In a study by Vantage, it was found that 95% of the resistance to a data driven model is cultural as employees worry about their expertise being supplanted. Thus, consistent and honest explanations about the benefit of supply chain technology is the single largest factor in a data driven model’s success. Implementation costs are frequently underestimated, so a solid of manpower plan and unwavering financial commitment is key.
When the data driven model is implemented, organizations will have an advantage over competitors as well as the ability to better absorb supply chain shocks. The contrast is empirically proven. 76% of executives from top-performing organizations cited data collection as essential, compared with only 42% from companies that fall behind their peers in performance.
Travis Ziebro
Travis Ziebro provides software design, development, and implementation services to the manufacturing and distribution industries through his consultancy, Momenta Tech. He also runs the popular engineering resource website, punchlistzero.com.
