A ‘one-size-fits-all’ approach to logistics is not appropriate in most instances. Some form of supply chain segmentation is therefore necessary in order to satisfy the various service and cost needs of the marketplace. This raises the question of exactly how supply chain segmentation should be undertaken.
Product Segmentation
It may be necessary to have different supply chains because of the very nature of the products. For example, when delivering to petrol stations the fuel may be delivered in large road tankers whereas the food and other items for the petrol forecourt shop would need to be delivered in clean, enclosed vans or trucks. Such product characteristics are oft en an important basis for supply chain segmentation.
Examples include:
- Size: The size of the product (or the total order) may determine whether it is best suited to be delivered via a palletized delivery network or by parcel carrier or by post. Another example is where large items, such as beds or sofas, require two people to unload and, in such instances, it is oft en found more cost effective to set up a separate network just for large bulky items, rather than having two people in a vehicle for all deliveries.
- Temperature regime: There are three main temperature regimes for food products, namely frozen (about –18 to –25°C), chilled (about +2 to +8°C) and ambient (normal outside) temperatures. These oft en form the basis for segmented supply chains, although it is quite common to find chilled and ambient goods combined. In fact, it is possible to combine all three in a single supply chain that comprises multi-temperature warehouses and compartmentalized vehicles.
- Bulk: Some products are well suited to bulk handling (eg liquids, powders and granules) and therefore require specialist storage, handling and transport facilities.
- Hazard: Hazardous goods may require a separate supply chain so that all the necessary safety measures can be implemented.
- Contamination: Even where goods are not hazardous, they may be able to contaminate other products (eg by their smell).
- Pilferable goods: Certain goods may be the target of opportunistic or planned robberies and therefore require greater security. An obvious example of this is where armored vehicles are used for bank note and bullion deliveries.
- Value: The value of goods may be important for segmentation purposes as this affects how costly it is to hold inventory in the supply chain. For example, goods that are low in value may be held at multiple locations close to the customers, whereas high value goods may be centralized so as to reduce safety stocks.
This concept can give rise to a segmentation basis, whereby high throughput but low value density products (ie a low value compared to their weight or cube) may be dispersed geographically, whilst a low throughput but high value density product may be held at a single Global Distribution Centre and air-freighted from there around the world. A typical example of the former product type is photocopy paper and the latter is high-value electronic parts.
Demand and supply segmentation
In addition to the physical characteristics of the goods, there may be a distinction between whether the goods are ‘functional’ or ‘innovative’ in nature, as noted by Fisher in 1997. Thus, functional goods may have a steady demand and require a cost efficient supply chain. On the other hand, innovative products may be new to the market, may be quite unpredictable in terms of demand and therefore require a much more responsive supply chain. This type of distinction between products with predictable and unpredictable demand is often associated with the lean and agile concepts respectively.
Demand is, however, only one side of the supply chain. The nature of supply also needs to be taken into account. An important factor on the supply side is the length of the supplier lead time – from the time of placing orders on the supplier up to the time of physically receiving the goods.
Under this segmentation framework, lean supply chain principles can be applied where there is predictable demand. In the case of long lead times, the sourcing, production, storage and movement of goods can be planned in advance in the most cost effective manner. Where lead times are short, then quick response and continuous replenishment policies can be adopted so that goods are supplied on a ‘just-in-time’ basis at the last possible moment, again keeping inventories and waste to a minimum.
However, if demand is unpredictable, agile policies can only be fully adopted where supplier lead times are short. In this circumstance, supply can fl ex to meet the rapidly changing demands of the marketplace, and again inventories can be kept low. However, where supplier lead times are long, then this is likely to lead to either an oversupply of goods (leading to high inventories) or an under supply (leading to lost sales).
The geographic location of supply is obviously a very important factor in supply chain design. Separate supply chains will be needed, for example, to bring goods from the Far East to European markets, rather than from local European suppliers. In fact, the decision as to where to source is oft en part of the supply chain design process. For example, goods with predictable demand may be sourced from low cost suppliers in distant parts of the world (ie part of a ‘lean’ approach) whereas goods with unpredictable demand may be sourced locally where lead times are generally much shorter and therefore supply can easily be changed to meet fluctuating levels of demand (ie part of an ‘agile’ approach).
Marketing segmentation
Segmentation has been adopted in marketing for many decades. It is used for demand creation purposes and it has long been recognized that different classifications of customer require different marketing approaches. As it is the customer that supply chains are trying to satisfy it would be sensible to examine whether these segmentation frameworks are relevant.
There are many categorizations of marketing segments but one such classification is as follows:
- Geographic: The location of the customer, eg by continent, country, region or urban/rural.
- Demographic: Populations are oft en broken down into categories according to such factors as age, gender, income, home/car ownership, employment and ethnic origin.
- Psychographic: This form of segmentation is concerned with the interests, activities and opinions of consumers, and is oft en related to lifestyles.
- Behaviouristic: This relates to how consumers behave, in terms of, for example, how frequently they buy certain products and whether they remain loyal to particular brands.
- Firmographic: In the case of industrial customers, then a common form of segmentation is by such factors as turnover, number of employees, and industry sector.
Combined segmentation frameworks
Most segmentation policies involve some combinations of the various frameworks described above. For example, one that has been proposed by Childerhouse, Aitken and Towill, has been named ‘dwv3’ with the key factors being as
- Duration: This refers to the length and stage of the product life cycle and may be related to Fisher’s ‘innovative’ and ‘functional’ product segments.
- Window: This is the time window for delivery or the delivery lead time that is required.
- Volume: This relates to the Pareto volume classification, ie whether the products are fast or slow moving.
- Variety: This relates to the product range, particularly in terms of the number of individual SKUs (eg colours, forms, sizes, etc).
Variability: This relates to demand variability and unpredictability.