Outsourcing manufacturing to specialists who can provide unique value and drive down costs is now the business norm. While companies may be going about it with more caution and consideration than perhaps in the early 2000s, the manufacturing outsourcing model itself is not in question. Brand owners are quick to discover that the first critical impact of outsourcing is the loss of visibility of detailed supply and demand data that they once held in their own ERP systems. Without this visibility they cannot make information-based supply management decisions. While getting access to detailed supplier supply and demand data is usually their first focus, it soon becomes apparent that tools that can allow them to view, analyze, and manipulate this data are also a very high priority. The following are ten critical data issues to be considered in structuring the supply chain data model for maximum utility.
Capture Core Data From Partners
More and more frequently, outsourcing agreements include provisions for data sharing, which address both content and frequency of data feeds. Working closely with suppliers to understand the details of this data is a critical component of the collaboration process.
1. Part numbering issues
Part numbering schemes will usually vary across supply chain nodes. How will common or equivalent part numbers be established for global planning and netting? Cross-referencing supplier parts to brand owner part numbers is the most common approach. However, it may also be necessary to consider parts in groups for netting purposes. For example, some suppliers may identify parts at the revision level, while others do not, or planning may be done at a product line level, not the detailed part level. System support for use of alternate parts and/or aggregation of supply and demand across multiple parts may be a critical requirement.
2. Core supply node master data and supply-demand details
Parts, bill of materials (BOMs), on hand inventories, local order policies, priorities, scrap, and yield factors are usually required for each supply chain node, as well as all active demand and firm supply records. It should be possible to match the local planning behavior of the outsourced supplier reasonably closely. This becomes particularly important when there are multiple levels in the supply chain. Incorrect planning at one level can radically skew requirements for downstream suppliers. Semiconductor manufacturing provides one example here, where wafer fabrication, assembly, and test may all be handled by separate suppliers with potentially different lot sizing, lead times, and yield factors at each stage. Without matching each supplier’s planning policies it would be impossible to accurately plan for product availability.
3. Supplier constraints, both capacity and material based
Suppliers may have shared constraint information, and sourcing may be constraint-based. For example, a supplier may commit to producing a fixed quantity per week of a specific part, or grouping of parts, or he may commit to a fixed percentage of available hours on a particular manufacturing line. If known, these constraints should be reflected in sourcing rules and should be adjustable by the planner for simulation.
4. In transit quantities, lead times
At any given point in time, a significant portion of the existing inventory may be in transit between supply nodes. Clearly, synchronizing the timing of data collection for all nodes, including in-transit quantities is critical, if data collection is not real-time. In practice, getting the in-transit data right is often one of biggest data hurdles and a clear understanding of how in-transit quantities relate to supplier commitments is mandatory. For example, are current supplier commitments net of in-transit shipments, or should commitments be decremented by in-transits? Lead time information is normally available by supplier, but, when modeling the full supply chain, transit time between partners becomes more and more important to model cumulative lead time. Click here to check out part 2.