Understanding Supply Chain cost drivers.
Various techniques have been used to gain transparency of these costs and try to understand the true costs of serving a product to a customer. For example, Activity Based Costing has been advocated by many; however most of the implementations I have witnessed been characterised by three main routes to failure:
· “Paralysis of analysis” where no useful solution is achieved because the costs are over analysed.
· The “It is not my cost” attitude, where the internal politics of the organisation or supply chain resulting in arguments about overhead allocation and subsequently results in the whole costing system falling into disrepute.
· The “We will get an answer one day” attitude, where the project takes so long that eventually it is forgotten about when a new initiative (or accountant) arrives in the organisation!
A pragmatic approach and more common sense approach to gaining transparency of supply chain costs is called Cost-to-serve. Cost to Serve enables an organisation to gain cost visibility on which to base strategic decisions. Cost-to-serve is a cross supply chain method of looking at the process-based costs, which does not descend into excessive detail. It supplements the general ledger thus relating information back to the cost base of the business. This information can then be used to produce benchmarks for the business thus directing attention to the strategic issues of customer and product variety, supply chain design and commercial and service terms.
The basic principal of the approach is the recognition that different product families and different channels to market have different cost drivers within the supply chain. For example, the number of orders or order lines drives the costs of order management and administration. The cube (how big the box is!), routing and frequency of delivery drive the transportation costs. For local delivery the drop density and order size are critical whereas for direct delivery the cube and delivery frequency become the primary cost drivers. Large customers may order in bulk, in comparison to small customers who order small orders, frequently with many items. Therefore it is dangerous to allocate costs equally to all products and/or customers. Cost-to-serve links cost information on customer groups, product groups and key characteristics (e.g. cube, returns, speed of delivery) thus giving the visibility of the true logistics cost involved in a transaction.
It has only becoming feasible to do this level of analysis with the advent of cheap fast computing power and effective data warehousing. Cloud computing and “in-memory processing” are also enabling analysis to run faster and more effectively. The size of the data files being down loaded for analysis are large, it is not uncommon for order files of six million lines to be downloaded for analysis. To generate a cost-to-serve analysis information from many sources is required. For example, product master files including product groups, manufacturing costs, and physical characteristics such as dimensions and weight, need to be related to customer master files that include customer references, group and geography. This then needs to be linked to historical order files that include information on customers, products ordered and number requested. Once obtained the data can be analysed for key cost driving criteria through relational database and subsequently re-compiled at a summary level into a spread sheet that can be easily utilised within the business. An important part of the methodology is linking the results of the analysis to the total cost by function of the business. This ensures correct interpreted of the costs and that the model is capturing and linking the various cost drivers correctly.
The application of the Cost-to-serve method in a number of organisations has given a number of interesting insights into supply chains. It has highlighted major differences in profit contribution, the relative profitability of the channels and products can vary enormously, one organisation identified one seemingly profitable product was actually making a significant loss once the real cost of transportation, warehousing and order entry were taken into account. Routing, stock cover and physical characteristics are key cost drivers, generally managers pay little attention to the dimensions and weight of a product. For slower moving items the key question is “how can cost penalties be reduced?” Cost-to-serve can be used to identify alternative ways to handle and deliver products.
You may think that this type of modelling is very time consuming, however experience so far shows that for most organisations the level of analysis required can be achieved in a matter of weeks. More detailed models do not enable more sophisticated conclusions. Advocates of the approach propose a general rule of “simpler is better and faster”. A key thing to consider is that for a supply chain professional to make decisions they do not need information to be 100% accurate. 80% is generally good enough. Knowing that a product or customer performs badly in comparison to other products or customers is the information that is required; the answer does not need to be produced to 4 decimal places of accuracy! So be careful if your accounts team want an answer that is 100% accurate, the analysis will probably take far too long and by the time you get the answer everything will have changed and the answer will be of no use! So simple, fast and reasonably accurate models which enable decisions to be made should be our goal.
The main benefit of Cost-to-serve would seem to be that it gives organisations a totally new perspective. This can then be used, not just for day-to-day management, but also for innovating the supply chain for the needs of tomorrow. So simple, fast and reasonably accurate models which enable decisions to be made should be the goal.
We include a session which discusses Cost to Serve and Supply Chain Finance on the "Cranfield Supply Chain Management Programme" 5 day course. It is also an part of the MSc Programmes in Logistics & Supply Chain Management. In my opionion all Supply Chain Professionals need to understand this approach.
Research at Cranfield is also undertaken in this important area.
Braithwaite, A and Samakh, E. “The Cost-to-serve Method” The international journal of logistics management, Volume 9, Number 1, 1998 pp.69-84.
An accessible paper giving an overview of the philosophy behind Cost-to-serve, it also includes a worked example of the approach.