Improving responsiveness in food supply chains
What happens when demand becomes very difficult to plan and the risk of wastage and stockouts increases? This chapter begins by considering the role of agility in supply chains today and compares it with lean thinking, addressing these challenges by reviewing the supply chains for two products – bacalao (salted fish) and bagged salads. The bacalao supply chain allows a comparison of agile and lean supply chains, while the Vitacress bagged salad case shows the principles of coping with rapid response and freshness integrated with 6-week growing cycles. The chapter ends with a summary of the nature of agile supply chains in terms of logistics focus, partnerships, key measures, process focus and logistics planning
In Chapter 9, we saw how lean principles can be applied to food supply chains, where the emphasis is on the elimination of waste, leading to productivity and high quality. Agility is concerned with improving responsiveness to end-customer demand and mastering market turbulence (van Hoek and Harrison, 2001). This policy can be especially relevant when demand fluctuations take place under very short timeframes (hours, minutes), and/ or where product lifecycles are relatively short. Both of these conditions – especially when accompanied by wide product ranges – make it difficult to forecast demand accurately.
Smoothed and levelled schedules – advocated by the lean thinking mindset – help to achieve high productivity and a stable inbound supply chain. But the downside is that such schedules often erode alignment between market needs and supply capabilities, leading to stocks of finished products that are not what the end customer wants. ‘Lean production’ in the trim and final assembly plant can lead to very ‘un-lean’ stocks of vehicles parked in disused airfields around Europe. Agility aims to forge a close alignment between needs and capabilities and is a particularly effective strategy under conditions of rapid demand change, high product variety and short product life cycles. The payoff may be that on-shelf availability (OSA) is improved without recourse to high levels of stock write-offs. The limitations may be that the supply chain is not as efficient in terms of resource utilisation and that manufacturing and distribution costs may be increased as a result. Overall, the aim is to improve value to the end customer while improving shareholder value at the same time (Harrison et al., 2007). The concept of the agile supply chain is an important perspective from which to propose how supply chains of the future will evolve.
This chapter begins with a comparison between lean and agile thinking. The comparison is then developed by means of a case study on the Norwegian bacalao (salted fish) supply chains serving the Dominican Republic and Portugal, respectively. While the basic commodity is similar, the market priorities are strikingly different, resulting in quite distinctive logistics strategies. We then turn to an example from the fresh food supply chain (bagged salads) to illustrate the current challenges and the promising practices that are being incorporated into future supply chain design.
The ‘agile supply chain’ is an essentially practical approach to organising logistics capabilities in order to be able to meet end-customer demand instantaneously. It is concerned with moving from supply chains that are structured around a focal company and its operating guidelines (e.g. ‘Ford production system’) to supply chains that are responsive to the needs of the end customer. The important change in mindset is to organise logistics from end-customer demand back upstream (‘outside in’), as opposed to pushing product/service offerings into the market (‘inside out’). Four key principles are involved, illustrated in Figure 10.1 (Harrison and van Hoek, 2008):
Fig. 10.1 Four principles of the agile supply chain (after Harrison and van Hoek, 2008).
1. The agile supply chain is customer responsive. Most firms are forecast driven because they have inadequate real time data and would be unable to respond fast enough to real time demand even if they had access to it. So they are forced to make forecasts based on past sales history or shipments and to convert these forecasts into sales and operations plans. Breakthroughs of the last decade, in terms of such advances as efficient consumer response (ECR) and the use of information technology to capture demand data from point-of-sale or point-of-use, are transforming the capability to listen to the voice of the end customer. In turn, this creates the need to respond to real time demand data.
2. The supply chain should be viewed as a network of partners who have a common goal to collaborate together in order to respond to end-customer needs. While individual partners may compete for market share in different networks, they collaborate with other members of each network in which they operate. Competitive advantage derives from focusing the efforts of the network of partners. In turn, this creates the need to coordinate the efforts of all partners through consistent objectives.
3. End-customer needs are met by aligning the business processes of network partners. Stand-alone process objectives that appear to make sense internally for a given firm, such as low manufacturing costs, may not make sense in the context of the network. Low manufacturing costs may mean large batch sizes that create high inventories and long lead times for other partners in the network. Significant challenges remain in order to coordinate the diversity of business processes that characterise food networks, for example, dairy farm, milk cooperative, cheese manufacturer, warehousing, distribution, back of store, front of store. In turn, this creates the need for partners to share common data, both internally (within a firm) as well as externally (between firms).
4. Network partners need to create a virtual supply chain in which key planning and control data are shared between partners. Virtual supply chains are information based rather than inventory based. While conventional logistics systems seek to identify optimal quantities of inventory and their spatial location, electronic point of sale (EPOS) and the internet have enabled network partners to act upon real-time demand data. The opportunity is to develop practices which avoid the distorted and noisy picture that evolves when orders are transmitted from one business process to another in an uncoordinated way.
Table 10.1 compares characteristics of lean and agile supply to show the differences in emphasis between the two.
|Logistics focus||Eliminate waste||Customers and markets|
|Partnerships||Long-term, stable||Fluid clusters|
|Key measures||Output measures like productivity and cost||Measure capabilities and focus on customer satisfaction|
|Process focus||Work standardisation, conformance to standards||Focus on operator self-management to maximise autonomy|
|Logistics planning||Stable, fixed periods
Smoothed, levelled schedules
We will now illustrate these characteristics in practice by means of our first food supply chain case, that of Norwegian bacalao (Jahre and Refsland Fougner, 2005).
Bacalao has been produced in Norway since about 1640. It is salted fish that has also been dried, traditionally in the open air on rocks, today in a drier. It can be kept refrigerated for several years and is said to improve over time. It has developed a strong position in the food cultures of many Latin countries, such as Brazil, the Dominican Republic and Portugal, where consumers often follow the Catholic tradition of eating more fish on Fridays and in the run-up to Easter. Marketing over many years has created the association with Norway as ‘the land of bacalao’, or ‘bacalhau da Noruega’ as it is called. It is a matter of great pride among consumers to master a variety of recipes for serving bacalao.
The overall supply chain is illustrated in Fig. 10.2. It takes at least 4 weeks to make the end product. The best fish is wild and taken by line, but trawled fish is also good, while nets give the lowest quality because the fish can be dead for a while before being hauled up. Today, the fish is increasingly farmed as well. The raw material is the major cost item – prices are set by the Råfiskelaget (the Norwegian raw fish association) and can vary a lot, for example from NOK 26/kg to NOK 15/kg within a year. Electricity and insurance are the other two major cost items. The fish is slaughtered (and bled on boat for the best quality), then matured in salt for 2–3 weeks. After salting it is dried, sorted, packed and distributed. There are no reliable ways of measuring salt and water content, so manual methods of touching and feeling the fish during each stage are used to ensure consistent quality and weight.
Bacalao is mainly produced from cod, which is preferred by Portuguese customers. But consumers in the Dominican Republic prefer pollock, which is a darker-fleshed fish that is more abundant in the North Atlantic. Cod is up to three times more expensive than pollock. The Norwegian fish industry is highly fragmented, with many small-scale fish farmers, fishermen and producers. Marketing activities are coordinated by the Norwegian Seafood Export Council.
Consumers are very quality conscious when buying bacalao. Quality is determined by colour, texture and firmness, water content and size. Portuguese consumers prefer smaller cod around 2.5 kg, while consumers in the Dominican Republic are less concerned with size, but prefer pollock. Note that ‘quality’ here refers to grade of fish rather than to conformance quality; both grades are fit for purpose in the markets they serve.
Company Noruega (CN) company has 150 employees and built its bacalao production facility in 1997 in the port of Ålesund, which has one of the largest harbours in Norway and one of the most modern fishing fleets in Europe. The company focuses on volume in order to benefit from economies of scale. Production is stabilised throughout the year by ensuring a stable supply of fish through sourcing a combination of frozen and fresh fish, creating a buffer of some 3–4 months supply. The company only trades in full truckloads, which are distributed via Hamburg or Rotterdam. Product is sold under the generic brand name of ‘Bacalao de Noruega’ in standard transport packaging. While CN serves most Latin markets, 80% of its sales go to the Dominican Republic as pollock bacalao. This market is relatively stable throughout the year, which matches CN’s stable production policy. CN is experimenting with pollock farming further to improve supply reliability.
Company Superior (CS) is also based in the Ålesund area and accounts for 15–20% of Norwegian bacalao exports to Portugal. Only cod bacalao is exported to this market, which commands a 10–15% price premium over other Norwegian bacalao. The product is popular with consumers, which creates a strong relationship with the single supermarket chain that sells it. Fish are sold whole, with a CS tag showing guarantee of origin from fresh Norwegian cod, which was an idea that came from its supermarket customer. This ensures that CS bacalao stands out from other offerings. Joint marketing campaigns are funded by both CS and its supermarket customer and include TV promotions. Only fresh cod is used in bacalao superior, caught by the coastal fleet in small boats. Supply is heavily dependent during the winter on quotas that are permitted in the famous Lofoten fishing field in the far north. CS buys from three fresh cod suppliers and from 15–20 suppliers of salt fish. Processing follows traditional routes, but some technology has been introduced into cutting and drying. The finished product is transported to Portugal in 22 tonne truck loads three times per week. The finished product is stored in Lisbon at the customer’s warehouse.
Table 10.2 summarises some of the major differences between these two products.
CN accepts more variation in its raw material source to enable continuous supply. This applies to type of fish as well as where and how it is caught. Farming and a healthy stock of frozen fish help to reduce further supply variations. On the other hand, CS seeks the best quality with minimum variation. The only inbound stock that is permitted is small quantities of salted cod.
While the raw materials and end product have many similarities, there are substantial differences in inbound and outbound logistics as well as processing and distribution strategies. These differences are fundamental to the need to support the brand (raising consumer expectations) by means of logistics strategy (meeting consumer expectations). We can conclude as follows:
• Two fundamentally different inbound strategies: CN focuses on secure, continuous supply and accepts greater variation in terms of type of fish, where and how caught, so farming is encouraged. They buffer and store extensively. CS goes for consistently high quality by not accepting much by way of variation: size, line catching and location are all important requirements. They do not store fresh fish or use frozen fish.
• Internally consistent marketing and logistics: CN matches the low price, continuous availability marketing mix by means of efficient sourcing and continuous availability and of ‘lean’ production and distribution methods. This enables high and consistent production volumes supported by a flexible product mix. There is less to go wrong in terms of supply, but the generic nature of the product mitigates against better margins or customer loyalty. CS matches the high price, seasonal availability marketing mix by means of highly selective sourcing and by focused factory production that is seasonal and relatively inefficient. Production is possible only when high quality, line-caught fresh fish are available. Limited and sporadic availability mean that the product has to reassert itself following supply interruptions, so the marketing pull must be consistent and strong. Traceability through tagging reinforces the superior quality image in consumers’ minds, supported by joint marketing with the major retail customer.
The way that the two supply chains have evolved illustrates the tradeoffs at stake – more of one thing means less of another. The CS supply chain has become focused on top quality (grade) product, but at relatively high cost and sporadic availability. The CN supply chain has become focused on the opposite – low cost and continuous availability, but at average quality (grade). While we have used the terms ‘lean’ and ‘agile’ to typify these opposing mindsets, they are really only labels to describe the logistics responses to two very different market situations.
Vitacress Salads was formed in 1951 by UK entrepreneur Malcolm Isaac, and has become the world’s leading producer of watercress. It specialises in top end, highly perishable, washed and ready to eat watercress and babyleaf salads. It was acquired by RAR Group, a private Portuguese company, in the summer of 2008. There are 600 employees in the UK and 1000 in total. Today, the company is one of the UK’s largest growers, packers and distributors of high quality salads with 15% of the market. Vitacress has a 40% market share in Portugal and is developing its market position in Spain. The company ‘has a commitment to total integration, growing a proportion of its seed requirements, drilling and cultivating the seeds, harvesting and transporting its crops and … processing, packing and distributing its produce’ (see http://www.vitacress.com/uk/about_vitacress.htm).
The reasoning behind the need for integration is that the shelf life is very short: the product has a shelf life of 5–6 days from packing and retailer requirements are such that no packed stock can be held by Vitacress. The order lead time is also very short – six hours prior to receipt of goods in the retail customer’s DC.
The inbound strategy is climate driven. In the summer, 95% of the product is sourced from the UK, while in the winter, 100% comes from abroad. Vitacress sources from its own farms in the UK, Portugal and Spain. Close relationships with partners in Kenya, Italy and the USA (Florida, Arizona and California) enable additional quantities to be purchased, along with sourcing from spot markets in Italy and France. Transport is by road from Spain and Portugal and by air from Kenya and the USA. Given the very short customer order lead time, winter harvesting must be carried out before receiving the customer’s next sales forecast. This creates serious issues in planning the supply chain, as Fig. 10.3 illustrates.
The supplier has to make decisions 6–8 weeks (depending on crop and temperature) before the customer’s order is received. So the whole of the period from 1st March to 24th April in this example is based on forecasts which have to be made far in advance of orders.
Forecast detail only becomes available during the final week (days 1 to 7 in Fig. 10.3). By using the same source of weather forecasts as its major customers, Vitacress is able to take an educated guess at the day 5 customer forecast in time for its ‘day 4’ planning meeting on day 3. Sales forecasts are only 80% accurate at best, so Vitacress must be prepared for some significant changes when the orders are made on day 6. Orders that are made by 07.00 am on day 6 are received in store 36 hours later. The cycle can be tightened up during the UK summer period, when it is possible to harvest at 05.30 am and to have an order in the customer depot on the same day. The vertically integrated nature of the Vitacress inbound supply chain makes this far more practicable; most competitors do not source from their own farms, so could not be as precise about when their raw materials were harvested.
As if long growing periods and short order lead times were not sufficiently challenging, demand variability is also relatively high. Demand during peak weeks is about 35% higher than average and within each week demand for the peak day is about 40% higher than the average. Because of the drive for freshness, less than 24 hours of stock is held to buffer any changes in demand or supply and 98% of factory output is made against customer orders. Post-farm waste is another control figure, to be kept within ‘low single figures’.
What techniques does a company like Vitacress have available to cope with the need for ever shorter throughput times? We can speculate what these might be under a number of generic headings, some of which have been informed by the descriptions above:
• Work to improve forecast accuracy: the long growing season and the need for very fast response to customer orders means taking ‘bets’ on raw materials at a relatively long range. Devices that help improve the accuracy of these ‘bets’ all help to reduce risk. Seasonal planning and promotional planning are two, and working from the same weather forecasting models as customer is a third.
• Maintain flexible supply capabilities: by ensuring flexibility in regard to growing times, by retaining ownership of sources of supply and by selected strategic supply partnerships. If it is not available in Europe, then fly it in from Kenya or the USA.
• Keep the product formulation flexible: an element of flexibility can be designed into product specifications. Examples of this would include products where the description is not specific to a particular leaf and also where, at times of high sourcing risk, the customer agrees to a substitute leaf being included in the recipe.
• Speed up the planning process: there must be no delay between new information (for example an order from customer) and response. So there is no room for bureaucracy or for ponderous planning and control systems. Decision making needs to be delegated and local. The aim should be to get closer to the real time situation, and to put in place systems that facilitate this objective. This could include following EPOS sales at major customers for each Stock Keeping Unit (SKU) each hour.
• Speed up the response process: by allocating new orders in ‘waves’ each shift and by reducing the time between waves. It’s not about fire fighting; it’s about a smooth, continuous response to an ever-changing market place.
• Recognise the risks and plan counter measures: a high-speed operation is exposed to what is happening immediately and must respond accordingly. Most risks, such as the impact of the weather on product availability and quality, are known, even if their timing is not.
• Improve coordination between all steps in the supply chain: the vision is for the end customer to create demand in each store and for that demand to be seamlessly transferred to each process in the supply chain. There is no room for lack of understanding of the ‘big picture’ at any stage.
• Behavioural capabilities must be developed in parallel: the above technical changes demand new time-constrained mindsets in all employees, both within the focal firm and in its supply chain partners. There needs to be a formal link between technical change and behavioural change.
The above techniques imply that there can often be a trade-off between service levels (OTIF), cost and quality. This trade-off can change from one product to another, from one shift to another. It can therefore help operating personnel to visualise what is wanted by being explicit. Figure 10.4 is designed to help in this visualisation process.
If the priority ‘ball’ is situated next to quality, for example, this dictates that quality is the top priority for this formulation for this customer. That could mean that cost suffers, for example, it may be necessary to scrap current (non-conforming) stock and to fly in replenishments from the USA. The delay could mean overtime in packing. It could also mean that there are delays in shipment and that service levels suffer. In order to make up for lost time, it may be necessary to courier the delivery to the customer at a premium price.
These case studies illustrate different stages in the route to agility. Let us compare the CS bacalao and Vitacress supply chains, using the characteristics listed in Table 10.3.
|Characteristic||CS bacalao supply chain||Vitacress supply chain|
|Logistics focus||Top quality product to premium market in Portugal
Productivity and capacity utilisation are not top priorities
|Top end product to markets in UK and Portugal
Productivity and capacity utilisation are not top priorities
|Partnerships||Raw material supply from selected reliable sources of fresh or salted fish
Quality not compromised by freezing or farming
Close partnership with retail customer
|Total integration, but fluidity provided by well-established sources of backup supply to fill in the gaps. Occasional use of spot markets.
Increasingly close partnerships with major customers
|Key measures||Product traceability, conformance to uncompromising raw material and process quality standards||Focus on customer satisfaction (OTIF and quality), but help employees to recognise the trade-offs|
|Process focus||Focused factory for processing
Subjective controls (touch and feel), but automated drying
|Focus on operator self-management
Help operators to cope with dynamic environment
|Logistics planning||Process when conforming product available
Stocks of finished product held by customer
|Very fast reaction speeds to customer orders, supported by flexible supply sources
Minimum inbound and outbound stocks
A transition can be perceived between the two cases. CS bacalao has developed a responsive supply chain that is geared to processing fresh line-caught cod quickly when it is available. Logistics planning is helped by use of salted cod from selected sources and by the ability to store finished product at the customer, but facility and labour utilisation are both subject to peaks and troughs as a result of uneven product availability. So there is still plenty of waste (in the form of finished product inventory and inconsistent productivity). Vitacress has moved further along the experience curve towards coping with market and supply uncertainty. It has been developing capabilities for very quickly responding to customer orders by developing relevant technical and human systems. This will also mean making compromises in terms of waste, but Vitacress is in the business of growing and processing a highly perishable product. These exigencies are driving Vitacress to become a truly agile company, able to respond to constantly changing customer demand by mastering uncertainty.
I am grateful to Ann-Karin Refsland Fougner, former PhD student at BI Norwegian School of Management and Marianne Jahre, Professor of Lund University and BI Norwegian School of Management for providing the basis for the Norwegian bacalao case.