Lessons from KFC's chicken shortage: responsiveness versus cost efficiency Wednesday, 7 March 2018
Last month’s interesting supply chain story, when Kentucky Fried Chicken (KFC) failed to deliver essential supplies of chicken to 750 of its 900 UK outlets, made it onto nearly every headline and news bulletin. The company was forced to close nearly two thirds of its branches around the UK, resulting in angry customers, and ridicule directed at the company trending on social media. Here’s what supply chain research can teach us about the reasons behind this spectacular supply chain breakdown.
The short version is simple enough: the fast food chain recently changed its logistics partner from Bidwest Logistics to DHL. This new partner operates on a totally different supply chain model, which was apparently not responsive enough to cope with KFC’s demands.
So, what’s the difference between the two models? New logistics partner DHL has a single warehouse in Rugby, in the UK Midlands. This warehouse is big enough and has enough capacity to service every region in the UK. This supply chain model is referred to as a centralised supply chain. Bidwest Logistics, on the other hand, was operating from six warehouses with moderate capacities situated around the country. Bidwest’s structure is known as a decentralised supply chain. It also has some facilities in other regions, each serving mostly regional customers.
Capacity versus lead time
Essentially, what we see in KFC’s recent decision is an attempt to navigate a trade-off faced by every manufacturer when building a new construction site or while choosing a logistics partner: that between capacity and delivery lead times. To maximise profits, you can’t have both, all of the time. In this case, obviously, short lead times lost out and hungry people were left without their buckets of chicken. Let’s look in more detail at the underlying trade-off.
Centralised means more for less
Centralised supply chains offer two advantages compared to decentralised supply chains. First, a centralised model helps to cut operational expenses significantly due to economies of scale and reduced fixed overhead costs, such as rent, insurance, and administration. As a result, centralised supply chains typically have higher storage and production capacities than the decentralised ones.
The second advantage is that the single massive warehouse in a centralised model does not need to be close to the most important – mostly urban – markets, so it can be built on a relatively cheap site, which drives down the cost of acquiring the land. Building additional capacity means expanding just one location, so that too is relatively cheap in the centralised model.
But while a centralised strategy may reduce operational costs, being located further away from the company’s market bases adversely affects its responsiveness. This may become a problem when outlets have to be replenished frequently – as is the case with fresh produce – and when demand fluctuates rapidly.
Working in a decentralised model, with warehouses located close to markets, allows companies to reduce the lead times and to be more responsive to volatile customer demands. This strategy is effective for companies whose profitability and survival depend on meeting highly volatile customer demands.
The downside of this approach is that the cost of land per square metre is expected to be higher. Fast-growing companies that operate with a decentralised strategy face higher costs for additional warehouse capacity. They might have a hard time meeting the additional demand generated by the growth of their businesses.
We analysed this trade-off in a paper I recently published with Prof. Ralf W. Seifert from the Operations Management department at IMD. It was published in Production and Operations Management. We developed a framework that can help companies such as KFC that don’t have ambitious growth targets to decide the most effective supply chain strategies.
In short, our results showed that the strategic decision whether or not to centralise distribution capacity should be based on the company’s profit margins and uncertainty of demand. We have identified two ideal situations.
- Centralised supply chains work best for companies that have low profit margins and low demand uncertainty. In this case, having a high-capacity warehouse achieves economies of scale. Delivery lead times can be long, because customer demand is predictable
- Decentralised supply chains are most effective when customer demand is uncertain and profit margins are high. This way, the higher margins can make up for higher costs of operating out of more and decentralised warehouses.
The problem with KFC
This framework can explain what went wrong with KFC. The company has a business model that operates with very low margins. What’s more, the food industry deals with intrinsically high uncertainties of demand, just like retail, fashion and sportswear. Here, decentralised supply chains can help to reduce mismatches between supply and demand, and improve customer fulfilment. Food and beverage company Nestlé, for example, operates successfully in 47 US states with 25 production facilities and 43 distribution centres for exactly this reason.
But by moving away from a decentralised system – probably under pressure to reduce costs – KFC introduced a big risk of being unable to cope with customer demand, or being left with costly surplus stock in its supply chain. This decision, to centralise its logistics, would be a logical one only if KFC’s outlets didn’t need frequent deliveries, and their demands were regular and very stable. Both of these conditions are simply not possible in this industry.
What else could they do?
What else could KFC have done? Our study shows that KFC could have reduced the high risk that it introduced to its supply chain in two ways. First, they could have continued to work with Bidwest Logistics and maintained the decentralised warehouses to keep responsiveness up. Second, to make this profitable, KFC could have improved its margins by increasing the breadth of its product offerings and charging higher prices for the new menus. This way it could have achieved a far better match between its business model and the design of its supply chains.
Read the paper: Biçer, I. and Seifert, R.W., 2017. Optimal dynamic order scheduling under capacity constraints given demand‐forecast evolution. Production and Operations Management, 26(12), pp.2266-2286