The choices made concerning these drivers and related operational processes determine the
responsiveness and efficiency of the supply chain. So let us look at the operational processes that
make use of inventory, transportation, facilities and information. As we saw before, the traditional
view on logistics management in a supply chain is the ‘cycle view’. In this view, chain processes
in a supply chain are divided into a series of cycles, each performed at the interface between
two successive stages of a supply chain (see Figure 2). A cycle view of the supply chain clearly
defines the business processes and activities involved and the owners of each process and relative
roles and responsibilities. Furthermore, because of inventory being held between the cycles, the
main processes are decoupled to a certain extent. This implies that each process can function
independently and is not hindered by ‘problems’ in other processes. In other words, each actor
in the supply chain manages its own processes without coordination with chain partners. This
opposes the ‘Just-In-Time’ (JIT) philosophy of inventory management, which states among other
things that the decoupling of activities by inventories should be eliminated, since it hinders supply
chain visibility and supports the sub-optimization of the supply chain.
Reducing inventories is beneficial from the point of view of visibility and reducing costs.
Considering the nature of products in agrifood businesses, it is also beneficial for quality
reasons. These businesses have to deal with specific characteristics of the product and processes
(see also Table 2), such as perishability, with resulting timing and conditioning constraints for
storage, long delivery lead times, sometimes products are supplied from far away or one has
to wait until products are harvested, uncertainty and variability in product quantity and quality,
bulkiness of goods flows, and so on. Traditional inventory management principles, mostly
developed for the manufacturing industries, do not necessarily apply. The main challenge for
agrifood businesses is to match (often uncertain) supply and (often uncertain) demand, taking
care of specific requirements regarding product quality and safety. Furthermore, one has to
keep in mind that excessive inventories in supply chains tie up capital that could otherwise be
used in productive investments. This is an area of special concern for developing countries.
Due to increasing consumer demand variability and uncertainty resulting in an increased
demand for capacity-flexibility and thus reduction of inventories, the ‘push/pull’ view of supply chains is gaining more interest (Figure 6). This view aims at eliminating as much stock
as possible in the supply chain and focuses on the extent to which customer orders penetrate
or may penetrate the logistics system. By eliminating stock at the retailer and wholesaler,
they are risking less by having the wrong products in stock. So when a customer arrives and
demands a product, the retailer will order the product at the processor, resulting in a delivery
lead time. For example, think of buying a car that is manufactured according to the customer’s
expectations (with or without sunroof, radio and so on). This type of customization is now
also used in food supply chains; where is inventory kept, what stages in the supply chain are
so flexible that they can first order and then deliver the product; the starting point is not at the
individual consumer, but at the retailer. Due to product proliferation, shelf space has decreased
enormously in retail outlets. This results in a request for frequent deliveries and short lead
times. Some products are packed in customer specific packaging materials; this packing process
can be done at the wholesaler, processor or producer. The less handling in the supply chain,
the lower the costs and quality losses. The idea is to minimize inventory levels in the supply
chain. If the producer keeps stock, the processor keeps stock and the retailer keeps stock, this
results in a long product throughput time, high costs and possible quality decay. By eliminating
as much inventory as possible in the supply chain, costs are minimized, quality is optimized and
service can be maximized since the right product is delivered in the right quality and quantity,
at the right time, at the right place One of the concepts that goes into the ‘push/pull’ view is the Customer Order Decoupling
Point (CODP) – also referred to as the Demand Penetration Point (DPP); this point separates
the part of the supply chain whose management decisions are governed by customer orders
(pull process) from the part of the supply chain where production plans are made based on
forecasted demand of consumers and/or forecasted orders from partners downstream in the chain (push process). Downstream of the CODP- that is towards the customer - the material
flow is directly controlled by customer orders and the focus is on customer responsiveness (lead
time and flexibility). Upstream towards suppliers, the material flow is controlled by forecasting
and planning and the focus is on efficiency (usually employing large lot sizes) and taking into
account inherent properties of material flows and production capacities and resources. It must
be determined where the decoupling point should be for each product-market combination or
product group in the company.
Hoekstra and Romme (1992) distinguish five possible positions of a decoupling point
(DP) as depicted in Figure 7. These range from having inventory in all stages of the supply
chain and delivering customer orders from stock (DP 1), to having practically no inventory in
the supply chain and starting to assemble (make) a product when the order comes in (DP 5).
Hoekstra and Romme regard the CODP as important for several reasons:
• It separates order-driven activities from forecast-driven activities.
• It is the place where ‘independent demand’ is converted into ‘dependent demand’.
• It generally coincides with the last major stock point in the goods flow.
• It creates the opportunity for upstream activities to optimize independently from
irregularities in market demand (in contrast to the JIT concept in which inventories are
seen as ‘blocking the view on problems’).
• It separates two areas in which the nature of decision-making is very different: upstream
from the CODP the focus is on planning and efficiency, downstream the focus is on the
acceptance of orders and lead time management.
It is clear that in food supply chains, less DPs are possible due to long production
throughput times. Figure 8 presents an overview of DPs that are possible in the supply
chain of perishable products; in this case flowers and potted plants. In the first two designs
(DP1 and DP2) all products are delivered to the customers from local or regional stock – no
customization activities are performed. In design 3, potted plants are customized (that is
value-adding activities performed to make the plants customer specific) at the auction, trader
or hub and successively delivered to the market outlets. Finally, in design 4 the grower has a
direct relationship with the final customer and harvest, packs and delivers its products (via
traders or transporters) to customer outlets; the auction can be bypassed in this network
design. The concept is useful to determine if processors should produce a large volume
of end-products and put them on stock or try to minimize inventory levels by assembling/
packing products to order.
There are several elements exerting an upstream or downstream influence on the
position of the CODP (Figure 9); in other words where to keep product inventory. It is a
balancing process between the delivery time requested by the customer, the throughput time in
purchasing, production and distribution and the expected customer service of an organization.
If the requested delivery lead time is very short, stock should be kept close to the market. On
the other hand, if the delivery lead time is relatively long, stock could be kept upstream in the
supply chain (towards the processors) taking advantage of centralized inventory management.
Other factors, such as whether the products are universal or specific, also play a role in this
trade-off process.
• Perishability of end-products
• Divergent product flows
• High demand uncertainty
• Customer specific products