One of the questions for any production system is if the product is produced on stock before the customer order (make-to-stock, MTS), or only on demand after the customer order (make-to-order, MTO). In many cases this is an easy decision. Custom-made items are always make-to-order, since you cannot start before you know what the product will be. Everything else does have exceptions. Let me dig deeper into the decision tree on deciding which items to produce on order, and which ones for stock. This is a short series of blog posts, and the first one looks at the key aspect (but not the only one) in deciding between make-to-order and make-to-stock.
Introduction
In make-to-order, production starts only after a customer order. The customer always has to wait on the completed product. In make-to-stock, production starts before a customer orders the item, in the hope that the customer will actually order such an item later on. In this case, the customer (ideally) does not have to wait for the product. Overall, customers never want to wait for products. If they have to, they would like to wait as little as possible. However, it is not always possible to produce make-to-stock, and even if it is, it may not be economically sensible.
This leads to an often overlooked but important distinction between make-to-order and make-to-stock: how you optimize the production system. With make-to-order, the customer must wait, but wants to wait as little as possible. Hence, the customer is interested in a short lead time between the order and the delivery. This can be achieved by having lots of capacity available, but the manufacturer does not want his machines and workers to sit around idle waiting for an order. Hence, for make-to-order the production system is optimized for a trade-off between the lead time and the utilization.
With make-to-stock, the customer wants its item right away, and hence is interested in a high material availability or delivery performance. This can be achieved by having lots of inventory, but the manufacturer does not want lots of expensive inventory waiting around for a customer. Hence, for make-to-stock the production system is optimized for a trade-off between the material availability and the inventory quantity. These are quite different goals, but it is quite possible to combine this in the same production system. Especially if the make-to-order jobs are no more than 30% of the workload, you can prioritize the make-to-order jobs over the make-to-stock products, and achieve a faster lead time for make-to-order at the cost of an only slightly larger inventory of the make-to-stock parts. I have written quite a bit on this in my book All About Pull Production.
Overall, you have to decide for all of your products if you want them to be produced make-to-order or make-to-stock. This applies not only to the final products, but also to the subcomponents and materials. Do you produce your subcomponents make-to-order or make-to-stock? Even when purchasing your own supplies, you can similarly decide between purchase-to-order (PTO) where you buy material only after you need it for a product ordered by a customer, or purchase-to-stock (PTS) where you buy material beforehand, hoping that you can sell it later as part of a customer order. Overall, this is a question of cost-benefit. Make-to-stock has higher inventory-related costs, but make-to-order may miss sales, customers, or even have penalty payments for being late.
On a side note, it is common to create custom products using (at least some, usually mostly) make-to-order parts. Your end product may be produced only based on a customer order, but you produce it by building variants using the many parts you have on stock. Any custom-ordered supplies will extend the lead time significantly.
Quantity?
The first criteria everybody thinks of when deciding between make-to-order and make-to-stock is the quantity produced. The more of an item is required, the easier it is to produce it to stock. This is often a good approach, but not flawless. Let me give you a counterexample.
A company produces coffee cups. Cup A sells around 10,000 items per year. Cup B sells around 5000 per year, and Cup C sells much less with 1000 items per year. At a first glance, Cup A would be a very good candidate for make-to-stock. Often, this is correct. However, if Cup A happens to be, for example, a one-off order by a single corporate customer that wants to give all of his employees a motivational cup for an anniversary, then this will be a one-off order that can be produced only after the customer orders the cups and finalizes the design. Hence, even though Cup A has a larger quantity, it would have to be produced make-to-order.
Cup B could be a Christmas-themed cup, which sells only between Halloween (when Christmas traditionally starts in retail) and actual Christmas. The total quantity is smaller, but it is make-to-stock, since it is sold to individual customers that just happen to desire a Christmas-themed cup. However, the demand does have quite some fluctuation, and outside of the Christmas season it will not sell at all. Hence, it is probably best to produce the item to stock before Christmas, but not to produce it at all or maybe produce it only make-to-order if someone really wants a Christmas mug in March.
Finally, Cup C may be a long-selling cup that is popular as a birthday gift for people when you don’t really know what to give (i.e., “Best Dad Ever”). This product has a quite stable demand. Hence, Cup C may be make-to-stock.
Quantity & Fluctuations!
Overall, while the quantity is relevant, on its own it is not enough information. A stable and high demand is the best candidate for make-to-stock. A small and highly fluctuating demand may be more of a candidate for make-to-order. In its extreme, a product that is produced only once ever (i.e., a custom-made item) is the prime candidate for make-to-order. Overall, a combination of the quantity and the fluctuations are the most important parameters for deciding if you produce make-to-order or make-to-stock. However, it is not the only parameter, and there are additional factors that influence your decision. But before I go into additional parameters, I first want to look at how to measure quantity and variation in my next post. Now, go out, look at your production volume and fluctuation, and organize your industry.
I fully agree with you, Christoph. I would however a few comments here.
Make-to-stock (MTS) production is certainly advantageous to customers because order-to-delivery time is only due to packing and shipping.
As you indicated, make-to-order (MTO) production is inevitable in custom manufacturing. In many high-mix, low-volume (HMLV) manufacturing units (like machine shops), which make a large variety of products, demand for some products is infrequent, uneven and highly unpredictable. Normally, such products are made against received orders because it is not economical to make them in advance and store them for a long, uncertain time in anticipation of demand.
As you mentioned, capacity of such HMLV unit may sometimes be very much underutilized due to major fluctuations in demand, In such case, the unit may make some of those products to stock as a trade-off between inventory cost and potential reduction in total sales.
I see another scenario. Make-to-stock production is driven by inventory policies for final goods. Manager of final goods inventory serves as an internal customer. The policies generate internal customer orders for products based on the existing inventory levels and estimated production lead times for products. In such scenario with frequent but erratic and unpredictable demand for MTS products, I consider the entire production as MTO type.
Hello Prasad, and thanks for commenting. Interesting view. Yes, make-to-stock does have internal orders. However, if we see these as make-to-order, then there would be no make-to-stock ever. For me, this is still make-to-stock, although due to the higher fluctuations you need more inventory, and a better ordering structure would be beneficial. If you produce make-to-stock the customer still sometimes has to wait for material, namely when you run out of stock. But the goal is that he does not have to wait, whereas in make-to-order a waiting time is inevitable.
Thank you Christopher for bring up this important topic.
There is a possibility in your example to follow a Make-to-Finish (MTF) production policy. Since all the cups are of the same size, the cups can be produced as (MTS) and stored in a supermarket, and when an order arrives cups are pulled to a printing process. So, here we have two pull loops, one is for the cups and the other one is for the printing. However, the quantity of the Kanbans has to be determined using the formula described in your book.
Hello Mohammad, definitely true. I picked the cup example for easy understanding, but of course one possible approach is to have blank cups on inventory, and print on demand.