Supply chain: Planning

Hugh Williams of Hughenden Consulting explains why implementing a single supply chain planning process is a crucial move in the race for competitive advantage.


I attended a conference several months ago, for which all of the delegates had to answer a questionnaire on their areas of interest beforehand. Unusual? Not really. The part that I found most interesting, however, was that 50 per cent or so had ticked the box titled ‘supply chain planning’. I was intrigued because no one had explained what supply chain planning was, or what it covered. How did anyone know whether to be interested? I suppose the words are topical jargon, but what did the delegates think it meant? My discussions with many people over the three days illustrated to me that there is a whole range of different ideas.

I have since spent a great deal of time defining the term in such a way that any industry sector and any size of company can easily understand it. The place to start was to define a supply chain. Largely, everyone is in agreement with the theory. Take the food chain as an illustration. We, the consumer, order a meal in a restaurant. In truth this is demand for a variety of items, but let’s just take one. Our order puts demand on the kitchen, then the butcher. He orders his raw material from the abattoir, who is supplied by the farmer, who in turn is being supplied with animal feed. This is a supply chain.

So if we now ask who plans the supply chain, it is rare that anyone controls the whole chain (the extended chain). Each link in the chain controls its own area, and influences adjacent links, up- and down-stream. One day there may be someone who plans the entire chain, for now we should look at supply chain planning as being the planning of as much of the chain as we can directly influence.

This is in itself an interesting notion. We can all probably think of examples of a complex supply chain. Equally we can think of a simple one. A complex supply chain might be represented by a large multinational company, with several primary manufacturing sites throughout Europe, say. Two or three other factories of its own feed it with intermediates and some raw materials. Products can be made in more than one location and can be distributed to perhaps 15 or 20 national distribution centres.  These in turn may send products to a larger network of regional distribution centres.

A simple supply chain may be described as a small privately owned single site manufacturing unit, receiving materials from a host of suppliers. It may produce and ship directly to customers on a make-to-order basis—no warehouse, no distribution network.

Does supply chain planning mean the same thing to both companies? Certainly not in terms of scope, but some of the base functions of supply chain planning may apply to both. The base functions are demand management and forecasting, inventory planning and sourcing, planning and scheduling, and materials management.

Perhaps we can use an analogy to explain supply chain planning further. Take, for example, a game of chess, or any board game and compare it to the task of planning.

In chess, we all know how the pieces move, that there are two colours, black and white, which start opposite each other. At any point in the game, we know where each piece can move and where they all are. This is all information on how to play the game; the question is, how do you win at chess?

There are two ways to play chess. One way to play is like most of us amateurs. We have a clear objective but we do not have a clear plan on how we are going to achieve it, so we take small steps. We respond to the other player’s moves, we react and maybe we are thinking one or two moves ahead. Grandmasters, on the other hand, also know the objective, but they have a bagful of strategies and tactics which they will use to get there. They even have special names for those. The Grandmasters are probably forecasting their opponent’s moves several moves ahead and monitoring to spot patterns in play.

So how is this analogy useful? In our organisations, we too know how to play the game. We know where the pieces are and how they move. For example, we know the orders placed on us and the delivery destination. We know where our warehouses are, how many pallet spaces we have, how much stock is in them. We know about our machines, people, production rates, and stock. We know our suppliers’ lead times and prices. We know a great deal of information in addition. This is information on how to play the game—this is the set of instructions we read when we first open up a new board game. But whenever we visit companies, we hear words like firefighting, shortages, stock-outs, expeditors, unplanned overtime, and probably similar words. This sounds like the way amateurs play chess, not the way the Grandmasters play.

So knowing how to play the game is important but it is not enough. We need to understand how to win the game. Supply chain planning is a mechanism for deciding how to win the game.

Every company has strategies on how it will win the game. Typically these will cover which products it will make and sell, which markets it will sell to, and how it will distribute its products to those markets. A company will decide whether it is going to supply from stock or make to order, and so it will have decided whether to respond in either ‘push’ or ‘pull’ mode. It will also have decided on capacity implications. One is always hearing about plant rationalisation stories. Whatever a company chooses to do, the strategies will determine which techniques and tools are relevant. Let us briefly mention some of those.

Almost every company with whom we work is about to start, or is in the middle of, some kind of project to improve what they are doing. Sales forecasting and demand management is a common one, often explained by the sentence, “If we could just improve our forecast of what the customer wants...”. There is a great movement currently in projects on co-managed or vendor managed inventories, as there is with efficient customer response (ECR). Many companies are looking at responsive, finite capacity based planning and scheduling. All of these, and many others, are tactics which companies are using to support their chosen strategies, and each company will require a different blend of these tactics in supply chain planning.

One of the most common tactics being chosen by companies revolves around computer systems, and a trawl around web sites will reveal the wealth of applications available. They range from forecasting to inventory optimisation and deployment, forecasting and planning systems, planning and scheduling systems, sales and operations planning systems, transport management tools and of course a whole variety of MRP/ERP systems (management/enterprise resources planning). The systems available today fall into two categories. There are those which are referred to as decision support systems, and those known as enterprise wide transaction systems. The demarcation lines, however, can get a bit blurred: it is possible to find, say, forecasting or perhaps scheduling modules in some of the transaction based systems. This is where our chess analogy comes in again, to help us understand the differences.

In our analogy, we concluded that what we need to do is to learn how to win the game. Knowing how all the pieces move any better will not help us in this. MRP/ERP systems are called transaction systems: they place major emphasis on the base of data that they hold. They cater for the order intake, the movement of inventories, shop floor data, where the products are, etc. This is information on how to play the game, not how to win the game.

Many companies today are reinvesting in their MRP/ERP systems. For the most part they are upgrading them or replacing them, and there are some sound reasons for doing so. Consolidated financial management is clearly a major one. Achieving commonality of systems across a multi-site group of companies is another good reason. There is a growing belief, however, that those companies who are investing in these new systems believe that they are going to get a mechanism for winning the game. I think they are wrong. Let me give an example.

A company we visited recently had searched the market for MRP/ERP suppliers, produced a shortlist and then visited in excess of 10 different reference sites. What they saw were good references, but what they also came away with was a clear message that while each reference had implemented a new system, they were still working in the same way as they had before. In other words, they knew how to play the game better, with newer technology, but they had made no progress on how to win the game.

Decision support systems are specifically designed to be told how to play the game by other systems. They are not databases, nor transaction systems. Their whole purpose is to help organisations make better decisions, by applying rules, algorithms and linear programming along with other mathematical techniques. They allow the user to evaluate different scenarios in order to decide how to win the game.

The scope of what they provide ranges from supplier to supplier. There are a handful of companies whose marketplace is the large multinational company, and for whom they will provide what might be described as ‘end-to-end’ supply chain planning. There are others, at the other extreme, whose product is defined as one of the disciplines within supply chain planning, perhaps forecasting, planning or scheduling, where they are looking to supply the smaller to medium sized company and maybe only within their own national frontiers. Each approach is valid and each has a definable marketplace. Interestingly, there are also examples of decision support suppliers teaming up together to provide a wider range of the supply chain planning functions, together.

The confusion lies in whether the transaction-based MRP/ERP systems really do have decision support capabilities within them. Unfortunately the answer is not as clear-cut as ‘yes’ or ‘no’ as a small number of suppliers are taking steps towards this. The vast majority, however, do not, even the largest and most popular systems. The reason we know this is that MRP/ERP vendors are forging relationships of various sorts with the specialist decision support providers. There are examples of ERP vendors acquiring decision support vendors. There are companies working on technology bridges. There are others who have standard interfaces to the larger systems. If both sets of companies were providing the same solution, this kind of activity would not be necessary.

Systems are important, but they are only one part of the tool set that a company will use. Winning the game does not mean just putting a system, or systems, in place. It means changing the way the process of supply chain planning is carried out and addressing all of the issues that this will throw up. 

What we find most interesting is that companies today carry out all of the supply chain planning functions in a departmental fashion. Forecasters forecast, planners plan, and so on. Very few companies have a department for carrying out ‘end-to-end’ supply chain planning. This is gradually changing with the knowledge and understanding of what is possible today; however, just putting in ‘best of breed’ systems to carry out each function is absolutely no guarantee of improving the process.

How often do we hear of companies who forecast monthly and release the forecast once a month to Planning, who then need several days to put the plan together before they in turn give it to the factory? Who schedules the factory? Is it a planner or someone on the shopfloor? What are the influences that lead them to their final decisions on what to do next? Is it Production, who want nice long runs with no changeovers, or is it Customer Service, who want it to be delivered when the customer wants, with scant regard for efficiency in the factory? How long does it take in your company for a change in the forecast to work down to the shopfloor? Minutes, days, weeks, or months even? Do any of the individual functions have any appreciation, understanding or even regard for the others?

To change the way we work will not be easy. Companies will need to understand what measurement systems influence the way their people work. Giving an incentive to increase forecast accuracy may have a serious negative impact on projected inventory and consequent sales. In supply chain planning we are talking about the optimisation of the global supply chain, and therefore, potentially the sub-optimisation of some of the individual elements of the supply chain. Tell that to the factory manager who is measured on volume output. It is equally true that we do not have to change everything in one go, or systemise everything at once. 

Using computer-based decision support systems will give us the ability to evaluate quickly decisions that are extremely difficult to evaluate in the current manual way. They do not in themselves take time out of the overall ‘end-to-end’ process of supply chain planning. Having a good forecasting system does not mean that the shopfloor will respond any quicker to changes. Therefore what we need to do is to design very carefully the way we want our planning process to work, in an integrated way. We shall need to understand how all of the individuals can work together in one single process and what currently exists to prevent that.

The supply chain planning decision support systems that we choose will need to be integrated into the host business system, so that any movement of data between systems is completely seamless. Now that we have a world of technology on our PCs and we can just move data and objects from one system to another with no problem, our expectations of ‘interfaces’ between systems have changed. It is no longer acceptable for the user to be involved other than by a few keystrokes.

In summary, therefore, we can see that our business systems are the guardians of real-time changes, for example in inventories, orders, or production activities. They tell us how to play the game and where all the pieces are. Decision support systems allow us to take that knowledge and try out different scenarios, to help us arrive at the way we are going towin the game.

However, systems are only part of the solution. The construction of a single supply chain planning process is what will give our organisations the speed of response and a real competitive advantage. It must be a continuous evolution.  

Hugh Williams is managing director of Hughenden Consulting, a specialist consultancy focused on the processes of supply chain planning.