Postpone differentiation consists of delaying the point of product or process differentiation in order to supply semi-finished products rather than finished products.
It is by looking more closely at the evolution of the automotive sector that we understand what is postpone differentiation. Let us take up the model of Henri Ford advocating, at the beginning of the 20th century, mass production and his model of the Ford T. In a position of strength in a favorable economic context, the whole sector copia this model of operation (stock, maximum capacity…).
It will be that much later, from the years 1950, that the model was challenged by the arrival of a more diversified demand. It will be Alfred Sloan, then CEO of General Motors, who brought a real innovation in flow management and design : postpone differentiation. At that time, it translates into the setting up of platforms common to several car models.
In a market that always demands more diversity, postpone differentiation becomes an important issue. It allows to increase customer satisfaction, to reduce inventory levels, to improve the service rate and simplicity in the scheduling and management of the production.
Postpone differentiation consists of delaying the point of product or process differentiation in order to supply semi-finished products rather than finished products. The goal is to produce a maximum of standard elements and to push back as long as possible the point where each product is different from the others and needs to be identified as such1.
It consists of manufacturing semi-finished products as long as possible. There are different types of differentiation : Labelling, packaging, assembly and manufacturing differentiation2.
The following diagram illustrates the general principle of postpone differentiation. Upstream, we will most often produce in a strategy Make to Stock to power a Supermarket of standard semi-finished products. Downstream, we find a flow of Assemble To Order, where we produce to order what the end customer wants.
The interest is obvious. By delaying the customer order point, This allows us to:
- Reduce the costs of in-progress stocks
- Improve our service rate
- Standardize our processes
- Reduce the cost of our processes
- Reduce serial change times over the entire value stream
1-Understanding the customer’s needs
The first step is to map all the ” options ” of our product range. We’re going to get the models and the bills.
2-Create the different semi-finished products
From this data, we will define the standard elements and the optional elements according to the 4 types of differentiations.
We are a company that manufactures gas cylinders. We manufacture about 12 000 products per day for a finite product reference infinity:
In fact, from the same carcass, a piston/tube sub-assembly, it can be inflated to any value.
We create the following table representing the number of possible alternatives at each stage of our process
|Phase 3||Phase Four|
Total number of references
Stem : 4
Tube : 5
Piston : 200
Carcass : 500
Painting + marking: As far as there is reference
Packing : 20
It is quite easy to see that the majority of the diversity is done at the level of inflation. The number of alternatives is infinite, because it depends directly on the customer’s need. In addition, most of the other combinations are carried out at the level of the piston and carcass assembly.
3-Designing the Flow
Once we have “carved out” Our product line, we will redesign the stream using VSM.
Taking our example, we get the VSM Next (it is voluntary to have only the essential elements to facilitate the reading).
In view of the situation, we redécoupons the flow in 3 levels of differentiation, and we rebalance the operations. Indeed, we defined in the previous step 4 different families of operations with varying degrees of diversity. This is the last family, the closest to the customer, that generates the most combinations. Our order point will then be the last stock before the customer, green on the drawing. It is from him that we will ” shoot ” the production upstream.
For companies with a wide variety of products, we will find automatic lines dedicated to high volume products and manual lines for others.
In the light of our new flow, we reconçoit the processes according to the following rules:
- Integrate and/or remove operations.
- Increase the possibilities of diversity of equipment.
In taking our case, before, we have at the end of the line:
- For high volume Products: an automatic machine carrying out the screwing and packing operations.
- For low volume Products: 3 manual lines performing the same operations.
Now the automatic machine will have to collect the features of paint, marking and inflation in addition to the other operations.
1 – H. L. Lee, C. S. Tang (1997) – Modelling the costs and benefits of delayed product differentiation
2 – W. Zinn (1990) – Shoul we assemble products before an order is received?
M. Freynesset, A. Mair, K. Shimizu, G. Volpato (2000) – What productive model? Trajectories and industrial models of global car manufacturers
O. Bruel (2005) – How to achieve the productivity/responsiveness compromise
B. Agard, D. Vinck (2004) – Modeling and simulation for the analysis and optimization of industrial and logistical systems
P. Médan, A. Gittens (2003) – Production Management