Part 1- Decoupling Points in Food Manufacturing

What are the best practices for determining whether an item should be make-to-order (MTO) or make-to-stock (MTS)?  Is it better to hold finished goods inventory to react to customer demand or is it better to hold packaging and raw materials with a flexible production schedule to keep working capital down?

It my experience, food manufacturers typically produce in large batches to keep production costs down and optimize plant utilization.  Most items are MTS to assure there is adequate inventory to meet customer service delivery requirements.  The only items that are typically MTO are specialized items for export or military customers, due to shelf life requirements or intermittent demand. But perhaps more than just these difficult items should be MTO.

Background

Most of the prevailing wisdom on MTO vs. MTS focuses on demand.  Items that are low-volume and difficult to forecast should be MTO while high-volume items with reliable forecasts should be MTS. This is a rather simplistic approach that does not incorporate all the relevant factors. 

Academic Literature on the decision of MTO vs. MTS dates from 1965 when W. Popp recommended a decision based on costs.  Many subsequent papers assumed that lower demand items are produced MTO and higher demand items are produced MTS.  More recent research has considered the impact on the MTO/MTS decision of factors such as market competition and manufacturing time diversity.  The most comprehensive treatment I have found that incorporates all the above is a PhD thesis by C.A. Soman from 2005 titled “Make-to-order and make-to-stock in food processing industries.”

Factors to Consider

Various sources have different lists of factors to consider for the MTO/MTS decision.  Potential factors for consideration include: service considerations, production lead-time and reliability, demand variability, production cost, setup cost, holding costs, backorder costs, capacity utilization, and working capital restrictions.  

Service considerations: How long is the typical customer willing to wait before seeking an alternative source?  If customer will not accept a lead time that fits production cycles, the item must be MTS.  Is there a competitive advantage to a shorter lead time that would capture additional business or allow an increase in price? MTS may also be preferred when there is opportunity to gain additional business by having stock available for unexpected orders.

Production lead-time and reliability: if production lead-time is longer than customer tolerance for delivery lead time then it is not possible to use MTO.  If production output is unreliable it also jeopardizes on-time delivery of MTO orders, forcing an item to MTS.

Demand Variability: As demand variability increases, safety stock increases, resulting in higher risk of obsolescence.  As probability of obsolescence increases, MTO becomes more acceptable.

Production Cost: The greater the value-added during production of the item the more acceptable MTO is to reduce working capital by holding inventory in raw materials instead of finished goods.

Setup Cost: The primary cost driver of MTO is the increased number of setups (or changeovers) necessary to produce items whenever needed in a quantity to match orders rather than producing to ELS when inventory is depleted.  As setup costs increase, MTS becomes more acceptable.

Holding costs: The primary cost driver of MTS is the cost of holding inventory, which includes variable storage costs, cost of capital, and inventory shrinkage or obsolescence. As holding costs increase, MTO becomes more acceptable.   However, production of too many MTO items may require an increase in the inventory of MTS items to achieve the equivalent service levels.  

Backorder costs: As backorder costs increase, safety stock increases and MTO becomes more acceptable. 

Production Capacity: Not holding inventory for MTO items means an increase in changeovers, which consumes capacity and may increase production lead-time.  If production capacity is constrained, this may result in increasing inventory of the MTS items.  Thus, there can be a complex trade-off between decreasing inventory of some items and increasing the cycle and safety stock of other items.

Storage Capacity: If storage capacity is limited, the stock of MTS items may be constrained. More likely there would be incremental cost for additional storage and handling that factors into the decision of how much to produce. In either case, there might be more frequent changeovers for MTS which might make MTO more favorable for some items.

Working Capital Restrictions: If there is an overall limit on working capital imposed for other business reasons, then the total projected inventory value of raw materials, WIP, and finished goods must be considered. 

Decoupling Points

The MTO/MTS decision is part of a larger decision process about where to stock inventory and how much to keep. As mentioned previously, there are typically two primary reasons for holding inventory:

1. To assure service delivery requirements
2. To provide cost savings

Holding inventory helps assure delivery requirements because it enables a business to provide products or parts in less time than it takes to make or buy them. The technical term for that is decoupling—defined in the APICS Dictionary as “creating independence between supply and use of material.”

A properly sized inventory buffer provides a decoupling point.  Decoupling isolates events that impact one part of the system from affecting parts of the system on the other side of the decoupling point. In terms of the MTS / MTO decision, decoupling creates independence between variability in customer orders and variability in upstream supply. 

Figure 1 shows the decoupling point for a typical MTS item. Inventory is held of completed finished goods and customer orders are shipped from this inventory. Production pushes product into inventory based on a schedule derived from inventory targets and shipping pulls as needed to fill customer orders. Having inventory at this point in the process protects production from variability in customer demand and it also protects shipping from variability in production supply.

There are several possible decoupling points in a food manufacturing and distribution system.  Some of the possibilities include:

• Forward distribution centers with full product mix to supply a local market,
• Centralized distribution centers to supply national demand,
• Finished Goods warehouses adjacent to production plants,
• Intermediate storage of semi-finished goods or Work-in-process (WIP), or,
• Raw materials and packaging stored at production plant to enable scheduling flexibility.

The decision for whether to hold inventory at any of these decoupling points, and how much to hold, is an item-by-item decision. 

The typical food production plant consists of one or more continuous process steps (baking, frying, enrobing, etc.). Processing feeds one or more packing lines where food is packed into discrete consumer packages and/or bulk packages for foodservice. At a minimum, there are two possible decoupling points: finished goods and raw materials.  There may also be possible decoupling points for intermediate components or WIP. 

Figure 2 shows additional decoupling points that may be possible in a typical food plant. If there are intermediate components that are used in multiple finished products, working capital and risk of obsolescence can often be reduced by holding inventory of intermediate work in process (WIP) items and deferring final packaging into finished product.  To do so, Inventory and planning systems must be able to maintain inventory and lot traceability of WIP items.  Packaging lines must have the flexibility to pack WIP into finished products as orders are received

For a more detailed discussion on decoupling points, see Chapter 6 of Demand Driven Material Requirements Planning (DDMRP) by Ptak and Smith.

MTO/MTS Decision Process

If production is inconsistent or production lead-time exceeds customer order lead-time, inventory of finished goods or WIP may be necessary to assure service delivery requirements.  Otherwise, inventory of a specific item is only desirable if it is less expensive to hold inventory than to incur the production disruption of setting up for multiple products whenever orders come in.  There may also be constraints that come into play on total capacity and total inventory of all items at a plant or in a production network. 

Therefore, the MTS/MTO decision is sequential.  If service considerations require item to be in stock it is MTO.  Otherwise, the decision is made on cost optimization, subject to constraints such as plant capacity, storage capacity, and working capital limitations.

The sequential decision process for MTS/MTO may be formulated like this:

1. delivery service requirement analysis;
2. demand and cost analysis;
3. capacity analysis to assure feasibility of MTO/MTS classification; and;
4. analysis of constraints on working capital or storage capacity.

Conclusion

My next blog will present a specific heuristic for working through these four steps to determine which items in a plant should be MTO and which should be MTS.  Meanwhile, please comment below and let me know what factors your organization has considered for the MTO/MTS decision. I leave you with these questions to consider for the MTO/MTS decision (most of these need to be evaluated item-by-item):

• How long is the typical customer willing to wait before seeking an alternative source? 
• Is there a competitive advantage having inventory available for unexpected demand that would capture additional business or allow an increase in price?
• Does lead-time have an impact on the price customers are willing to pay?
• Does production deliver product reliably and on schedule?
• Is customer demand predictable and reliable?
• How does typical packaging changeover time compare to process changeover time?
• Are there intermediate or WIP components that are used in multiple finished products?
• Is there opportunity to hold intermediate WIP items for extended periods of time?
• Are backorder costs significant enough to merit consideration?
• Does obsolescence occur enough to merit special consideration?
• Are there unique long lead-time ingredients or packaging that must be held in inventory to produce any item within customer order lead-time?
• What are the constraints on the total system (plant capacity, storage capacity, working capital, other)?

Sources

Ptak, C. and Smith, C.: 2016, Demand Driven Material Requirements Planning (DDMRP), Industrial Press, Inc., New York.

Rajagopalan, S.: 2002, Make-to-Order or Make-to-Stock: Model and Application, Management Science, 48(2), 241-256.

Soman, C. A.: 2005, Make-to-order and make-to-stock in food processing industries, Labryrint Publications, The Netherlands.

Zaerpour, N., Rabbani M., Gharehgozli, A.H. and Tavakkoli-Moghaddam, R.: 2008, Make-to-order or make-to-stock decision by a novel hybrid approach, Advanced Engineering Informatics, 22(2), 186-201.