As a data scientist on Supply Chain Optimization projects for our customers, it’s my job to understand the value and costs that our customers’ organization puts on different aspects of their business. This supply chain planning and optimization includes everything from planned sales and sales of balance code or surplus production, to inventory holding costs, to transportation costs of either moving supply to the processing plant or finished goods to forward warehouses. Although some decisions may be limited, for example there may only be so much holding capacity at a warehouse, I will be trying to properly model all of the potential costs and revenues from production and demand fulfillment plans. The “optimal” plan from my perspective should be to maximize total margin having fully identified all of these above costs and revenues.
However, each of these sets of costs and revenues that our customers face generally live in different departments in their organization. Often in any given organization, there are separate teams for supply/purchasing, production, transportation, inventory tracking, and sales. There may be other functions that I haven’t included, and sometimes there may be multiple segments of each of these teams. For example, there may be both a domestic and export sales team, there may be different production teams for different plants, and there may be many different cross-functional teams as well. It is also very common for these teams to be judged and incentivized differently. The sales team may be getting their bonus from successful bookings of sales and service level fulfillment, while the inventory holding team may be getting their annual raises based on how much they can keep down those warehousing costs.
Of course, while each department is focused on its own purview, they all have effects on each other. Sales bookings inform production requirements and transportation needs, which then inform inventory holding requirements, etc. One major worry is that in the end, you may have an organization where different departments are being measured at cross purposes. If the service level isn’t fulfilled because of lack of inventory, who is taking the hit in their bonus for that? If it ends up costing more to ship goods from a central warehouse, but there’s a risk pooling inventory cost reduction by keeping the overall safety stock lower, is the transportation team going to be unfairly judged?
When a company like Austin Data Labs comes in to build a company-wide optimization model, we may end up recommending some reasonably large changes to the organizations’ standard processes. We may even end up identifying certain bottlenecks, where allowing certain constraints to be opened (increased holding capacity, decreased safety stock requirements, etc.) would generate a substantial increase to margin. To achieve these more optimal, increased profit results, we need the full organization pointing in the same direction. If there are particular deals that they’ve made in the past, but our optimized recommendations suggest that those deals will not drive substantial profit (or may even decrease profit), will the sales team be able to get behind dropping that contract? If our model suggests an increased safety stock volume to ensure a greater guarantee of service level fulfillment, will the inventory team be willing to make that change, and will they not be penalized for doing so on their yearly reviews? Are there other organizational politics in place that might cause certain teams to “rebel” against this company-wide change?
Updating your supply chain planning with optimization models can be very valuable to an organization, and can often cause teams to rethink some of their incoming assumptions about how they do their jobs. We look at our customers holistically, and do our best to make sure that there are no costs or benefits that we’re missing when we’re implementing our models. That means partnering with companies that are willing to be open minded, to see where the data leads them, and to incentivize and reward every member of their teams for following the recommended plan, even if it goes against that particular team’s function. Otherwise, you may end up with someone making sure that they hit their bonus target, but losing the overall company a substantial amount of money because of all the other interrelated effects.