Someone just asked "*M**y business produces more than 100 items. Each item has different ingredients and has a different selling price. Modeling it in Sysdea required submodel for each product then the submodels can be linked to total product cost and total sales in the main model. this means I have to have 100 submodels. Is there a different technique to model this case*?"

The same question would arise for trying to model individually large numbers of customers, staff or capacity units.

If the exact detail is important, then you would indeed have to use a model for every item [as you would have to do with spreadsheet or any other analysis]. But there are three ways to simplify ...

- There is usually a distribution of products, with a most-important fraction accounting for most of the revenue [a classic 80:20] feature - see class 5.2 and chapter 5 of
*Strategy Dynamics Essentials* at strategydynamics.com/msd2012/class5/ but substitute 'products' for 'customers' ... this means you can model, for example, just the most important 20 products plus an 'average other product' and in the parent model multiply this average other product by 80 to get the total. - Such businesses often organise their products in product-groups - you could model the average product in each group, then multiply by the number of products in each group and sum all groups.
- Still simpler is to treat the whole product range as a single population with average characteristics [
*average raw materials cost, average sales, average selling price ...*] and use the "attribute" approach, also explained in chapter 5 of *Essentials* and class 5 of both the Strategy Dynamics and Business Modeling courses ... Although this loses all item-level detail, it is still possible to model changes arising from [a] introducing a new product , [b] dropping a product, or [c] changing the average characteristic of all products (*e.g. raising all prices by X%*)