A manufacturer of metal parts hires a management consultant to help stimulate sales. The consultant at once suggests the manufacturer prioritise production of its top twenty (20) best-selling items.
The manufacturer thus makes one month’s worth of stock of each of the twenty (20) top-selling items. Three (3) months later, the stock is hardly selling. Meanwhile, customers complain that they haven’t received their shipments of items that are not on the top twenty (20) best-seller list. Pending orders is equivalent to one (1) month’s average sales and the manufacturer simply has no stock to serve the orders.
The management consultant had analysed the manufacturer’s sales history and listed the manufacturer’s top selling items based on their average sales value over the previous year. The top twenty (20) items constituted 80% of the manufacturer’s sales that year. It therefore seemed logical to have on stock those twenty (20) items. It was easy to see that the top twenty (20) items have a high demand history.
The manufacturer hired a supply chain engineer (SCE) to do something about the pending orders and out-of-stock problems. The SCE analysed the manufacturer’s operations and observed that the manufacturer produced 1,800 individual items or stock-keeping units (SKU’s) in that same period of twelve (12) months. Most of the customer orders the manufacturer received, however, were delivered late and many others were cancelled due to out-of-stock.
The SCE noticed that the management consultant based his recommendation to produce the top twenty (20) selling items on the following analysis:
The SCE broke down the daily histories of the top selling 20 items and saw that each item had an erratic demand behaviour, in which for one (1) item, it looked like this:
Not one of the top twenty (20) items was selling at close to the overall average quantity at any day or even any week throughout the twelve (12) months surveyed. Each item would experience very high demand in one or few orders but hardly would any item be selling close to average every day or every week. The variance between average demand and each day’s demand over a year was very large.
The manufacturer sold more than 1,800 unique items over a one (1) year period and most of each item’s sales were limited to one or two orders sometime during that same period. Some items did have frequent daily sales but they were in small quantities. The management consultant’s list of top twenty (20) did sell up to 80% of annual revenue but the manufacturer was losing potential sales from unserved orders of other items.
The management consultant thought that producing and having stock of the top twenty (20) best-selling items would bring higher sales as based on historical numbers. The consultant, however, didn’t see that customers didn’t need the said items every day. A few customers with big projects bought large quantities of the top twenty (20) items in one or few orders. Other smaller scale customers ordered much fewer pieces of metal products at any one time and for certain items, more frequently. The consultant didn’t realize that the manufacturer’s items were not needed every day, or even every year. Customers only bought for projects or for maintenance needs; items were only needed periodically.
Further studies by the SCE showed that some customers ordered each of the top twenty (20) items only once. It would be a different customer ordering for a large quantity. There was no uniform demand pattern. Customers buying plenty of an item were probably buying for one-time projects. Customers buying smaller quantities were buying for fewer requirements.
And because they were for projects, customers would have unique specifications for the items they needed. A customer’s order of an item was often different from that of another. Some customers would want better finish on an item; other customers would deem the item’s finish as is as all right. Even if basic specifications were consistent, it was commonplace for the manufacturer to do additional work on an item as per a customer’s request.
The manufacturer therefore was really customising items more than making the same items over and over. Sales orders very often had instructions for how products would be finished, cut, and packed. Some customers required very tight specs, others did not. Some customers wanted their items cut to certain sizes. Some customers wanted more stringent packaging; some were satisfied without any packaging at all.
The manufacturer’s order fulfilment system did not take into account these frequent instructions. The information system had on file more than 10,000 items and it was found that many of the items were similar to each other. In other words, every time a customer order was received, it asked for an item that was made before but with slightly different specifications. The accounting and IT groups were constantly entering “new” items into the information system.
The SCE therefore suggested that the manufacturer re-develop its customer service strategy. The SCE suggested the manufacturer refocus the order fulfilment system from one that sells based on a fixed inventory of items to one that is based on customisation. Instead of having a system like a grocery store, the system should be like a machine shop—i.e., only make an item when there’s an order. The SCE also recommended that the manufacturer only keep stock of needed raw materials, not finished items.
A large metal manufacturer a few kilometres away was actually doing that kind of thing. His inventories of finished goods were limited to stocks that are about to be shipped. He only kept at most a month’s worth of raw materials (he thought that already was too much) and he had no backlog of pending orders. Every item that was made had its own unique identity unless it was a repeat order to the same customer.
The SCE proposed a system in which the manufacturer’s sales representative would prepare quotations for customer inquiries. When a customer is interested in an item, the sales representative would quote not only price and quantity but also confirm specifications and schedule of deliveries. The sales representative would coordinate with a joint sales and supply chain support team that would translate customer inquiries into a quoted proposal for the customer. The quoted proposal becomes a sales order upon negotiation and agreement between customer and sales rep.
The supply chain team would keep stock of raw materials which happen to only number to less than twenty-five (25) items or stock-keeping units (SKU’s). The stocking strategy would be independent of actual demand but would take into account large spikes as in when a customer conveys interest for a very large order. Again, the sales and supply chain support team would ask the sales representative to negotiate delivery schedules to take into account the manufacturer’s capabilities to buy raw materials and produce the needed item.
How demand is fulfilled varies from industry to industry, enterprise to enterprise. One should study demand based on customer behaviour, not on overall totals or averages.
One should also tailor the supply and fulfilment of demand to the needs of customers. At the same time, one should always be aware of the system’s capabilities. Customers may be always right but the enterprise is not one with unlimited power. There has to be communication and collaboration via negotiation and mutually beneficial agreements that would address price, terms, and supply.
There has to be a right way to serve an order. Not for management, not for consultants. But for customers.