Managing a construction site is an extremely complex task. Construction site managers have to balance the pressure to comply with schedule, quality, specifications and budget targets. They have to interact with the head office, the client, subcontractors, employees, etc. Losses due to stockout of construction material and consumables are a concern because of the impact on profitability and schedule.
What the client really wants
Many suppliers of commoditized products offer construction companies prices that are only marginally different. These price differences are definitely considered in purchasing decisions, but what a construction site manager really wants is the assurance that his or her team on site does not stop. In most of cases, consumption or use of these materials follow a predictable pattern (consider fuel for example) but nobody on site is working with complex inventory forecasting models. They don’t have the analytical tools to predict consumption accurately. And even if they had the skills and the time, it may not be cost effective to focus much of their valuable time in a single supply, each with their specific supply chain considerations. Could outsourcing order generation be a solution?
An option is vendor-managed inventory (VMI). In a VMI model, the buyer of a product provides a supplier (vendor) real time information, like stock, consumption and sales of that product, and the supplier takes full responsibility for maintaining contractually agreed inventory levels at the buyer's consumption location. VMI has been around for a long time, but with mixed success.
A careful balance
For the model to make sense, it has to reduce out of stock loss of profit (client’s unserved sales or production stoppages) and/or reduce the carrying cost of the inventory (financing, damage, obsolescence), in excess to the cost of implementing the solution. Transferring the responsibility from client to supplier has to be analyzed on a case by case basis, carefully evaluating this balance.
Should companies reconsider VMI again? The improvement in forecasting and optimization algorithms for inventory and routing, and the decrease in sensor price has unleashed an enormous potential not only in cost reduction, but also improving the service quality and the management of risk, a key element of the VMI agreement, as a good Sloan Fellow friend with ample experience in this field reminded me today.
Take fuel distribution. Using dynamic optimization of loading, scheduling and routing and real time inventory acquisition and forecast, last mile delivery can be dynamically optimized considering multiple time horizons.
Short run: (intra-day) optimization of routes and load assignments
Medium run (days to months): load balancing (shaving demand peaks through anticipation of demand), reduction of OOS events, increase of delivery quantity and reduction of redirects and no-fits, human resources planning, etc.
Long run: supply network design, including investment decisions in fixed assets.
Last mile, strategic again
Much of the fuel distribution was outsourced by the oil companies long time ago, considering the last part of the supply chain non-strategic. Fragmentation may have resulted in suboptimal operation, leading to inefficiencies and low service levels. Based on what we have discussed here, it seems that significant value can be added if the information and decision-making systems of the last mile are centrally handled again. Should commodity majors reconsider their boundary in the downstream last mile?