Should-Costs across a Supply Chain

We recently completed at “Should-Cost” project for a large industrial transport equipment manufacturer. The challenge: estimate manufacturing should-cost analysis based on a single, 600 part BOM and some drawings. We completed the Analysis of Material Costs, vertically integrated Should-Cost (excl. supplier margins) and estimated Procurement Price (incl. supplier margins) in under 4 weeks. We achieved a good degree of predictive cost accuracy (+/- 5-12% variation) and a full view of environmental impacts and regulations for every part of the BOM and its related production processes … at design stage.




Should-cost analysis was developed to help procurement determine fair and reasonable pricing from suppliers. Today the method is embedded in both public and industrial procurement practices and used by best in class outsourcing companies like Apple, Cisco or GE. Should-cost analysis examines what a product should cost based on materials, labor, overhead, and profit margin.




Traditional product costing methods, such as


  • Cost structure analysis (materials, manufacturing process cost simulations)
  • Activity-based costing (incl. process and overhead allocations)
  • Investment and life cycle cost assessments (ROI, TCO, LCC)
  • Strategic sourcing (quotes from selected suppliers)


are effective, proven methods for product cost analysis. However, they also have shortcomings. They are typically


  • costly – a significant amount of work is done by consultants/experts, using proprietary tools
  • difficult to scale – each study done in isolation of the other and only scales with manpower
  • non-dynamic – commodity price changes are not automatically reflected in models
  • one-dimensional – cost impacts like environment costs, regulatory non-compliance, recycling rates, administrative burdens, etc. are excluded.
  • non-collaborative – promoting adversarial relationships with suppliers


How can those shortcomings be addressed, while maintaining cost control?




Networked Should-Costing (“NSC”) addresses those gaps and greatly simplifies current should-cost analysis. To start, it only takes the following Data Inputs:


  • Bill of Materials (such as in the picture above). Line items must include weights and part descriptions that allow an understanding of the materials used (such as a particular steel grade).
  • 3D drawing of part (on PDF). This will serve mostly for reference to understand shapes and associated machine operations.
  • Assumptions (or actuals, if available) on
    • manufacturing locations. This will indicate direct labor rates.
    • tooling complexity. This will indicate direct machine costs allocation.
    • company size. This will indicate overhead rates.


Networked Should-Costing puts emphasis on detailed material cost analysis across the fully supply network, while working with simple assumptions on non-material costs.  The data on production flows is dynamic and connected (for example to online commodity prices and forecasts). Therefore Networked Should-Costing understands all the material flows and processes that have led up to a given assembly.  In contrast, “strategic sourcing” or typical Should-Costing uses static, offline data to analyse a BOM “as-is”. In NSC, because the parts in the BOM can be linked to production chains that have already been modeled, cost calculation speed is taken down to seconds vs. offline calculations that typically take weeks to complete.


Why does this work?


Material costs across the supply network are more determinant of the cost-price, than a very high degree of granularity on labor, tooling and administrative overheads. Take two industry examples:


  • Electronics: Printed circuit board assemblies (PCBA’s) cost structure is typically 60%-80% materials, with labor (5-10%), overhead (3-5%), and profit (5-10%) added. It is not efficient to fully analyze the fabrication cost of every component.  Requiring suppliers to submit quotations with disclosed material costs will not be met with enthusiasm. Instead, accurately networking the material costs across the PCBA supply chain will yield should-costs with a high forecast accuracy of 90-95%.


  • Metals, Mining, Manufacturing: Bill of Material (BOM) costs are  predominantly driven by metal costs, machining & tools, transport, supplier margin, and various types of labor. A networked should-cost analysis provides cost projections of any particular line item “live”, for current or future prices. It uses data about movements in each underlying cost driver across the network and determines what the item ‘should cost’ at any given point in the network or in time. These models can be incredibly powerful both to identify and quantify savings opportunities through fact-based supplier negotiations.



Non-material costs may not be the bulk of the costs, but they remain significant. They can surely be modeled in fine detail, but let’s assume that any price negotiation can take 15% off of a purchase price. With Material costs determining 70-80% of product costs, one could be off by as much as 30% on non-material costs and still be 90% on target. If non-material costs can be determined by simple, basic assumptions at design stage, product costs can be understood on the fly (i.e. while designing a part) with a high degree of confidence before they get designed-in. The already networked data eliminates the need for “strategic sourcing” to find out what anything should cost. By doing so design cycles can be  shortened  by weeks or months.





  • Reduced cost and time of analysis in early design stage: Up to 85% of costs of manufacturing are from raw-materials, energy use and infrastructure. These costs add up through the value chain. NC can provide up-to-date cost information about these constituents, real time, throughout the entire value chain. The result is a value chain look at how costs build up to the purchased goods. When used in early design stage cost analysis, the approach results in instant cost estimations and potential supplier network discovery which forms the basis of pricing conversations with suppliers.


  • Lower information requirements: The approach can be used before CAD designs have been finalized. Starting with even a rough BOM + material information, NSC can cover upwards of 85% of the product with approximations. All other information e.g. compliance, environment, cost models etc. are enriched automatically by the system.


  • Scale: once models are created, they are automatically reused in subsequent analyses, across teams like environment, compliance, material spec etc. Through a live model of the product and its value-chain, changes are dynamically reflected in all derivative models. This allows for models to be enriched by team members,  staff or ext. consultants in real-time. That way the company gets smarter, faster.


  • Multiple dimensions and team collaboration: Cost, compliance, alternative materials or processes, environmental impacts etc. can all be linked and displayed throughout an organization  and its supply network. With that, any other team benefits from the data. i.e. for procurement to use in supplier discussions. NSC allows summing up costs or any other cumulative impact (i.e. CO2, material usage, toxicity etc.) through any point in a value chain.


  • Supplier Collaboration: non-confidential parts and data can be shared with suppliers (and vice versa). This provides the power to unlock aggregated performance improvements. Modifications are reflected in real-time in the models.


  • Mapping and real time information: all of the above can geo-mapped. Great for insight (like local costs or impacts) and additional features, like real-time alerts for disruptive events (i.e. floods, regulations) that are specific to your supply chain.