Supply Chain Management (SCM)

By Estelle Vermorel, January 2020

Multiple steps, network effects, variations and options: above all, Supply Chain Management is about handling complexity from a general perspective rather than a local one. It could also be defined as a science of trade-offs, as choices that need to be made, often related to costs vs. service level or product customization.


Basic definition

Abstract graph illustrating the complexity of a supply chain
SCM (Supply Chain Management) is traditionally defined as the flow of goods and services encompassing everything (processes, systems, people) from the production of a product with raw materials to its final delivery and consumption. It typically includes stages such as production, shipment, distribution and possibly return systems.
Supply chains only exist insofar as there is a network, meaning the interconnection of several elements to provide the ability to operate profitably on a large scale. In contrast, one cannot really talk about supply chain for, say, subsistence farming - although goods are definitely handled. Therefore, supply chains entail a real complexity, and SCM is the attempt to control and orchestrate such a complexity in the most profitable manner.

Handling the complexity of multiple options

In particular, this complexity is due to the myriad of options that come with a supply chain network. One should not confuse logistics and supply chain. As an example, the first can entail how to handle a truck delivery (making sure the truck operates properly, the driver is there on time, etc.), while the latter will be more a matter of how many trucks should be on the roads to deliver the goods. This requires the answer to a question - “How many trucks?” - and a choice, which in turns opens more choices: what should be put in the trucks? Which products, colors, sizes? And accordingly what should be produced? And so on and so forth.

SCM implies handling optionality on every level, for the product itself (variations, colors, sizes), but also for the suppliers (overseas vs. local, one supplier or many), the teams/departments (internal or outsourced - more specialization, but possibly less control), the systems (one central system or several specialized tools), etc. Naturally with variations comes the question of the cost. Usually more variations imply higher costs. As an anecdote, it is well known that black was the only color the Model T Ford came in in the 20’s, with the famous policy “any color as long as it’s black”. What is perhaps less known is that, initially, the car was available in gray, green, blue and red from 1908 to 1913, but it came with higher costs. Varying the options for a product means smaller batches, less economies of scale, and ultimately it’s a tradeoff between a more attractive product, appealing to a more varied audience, and the related costs. Volumes vs. price per unit. The same goes for pretty much every question in SCM: more suppliers with easier to reach MOQs (Minimum Order Quantity), MOVs (Minimum Order Value), price breaks vs. several suppliers, maybe with more flexibility on lead times, more local? One warehouse vs. several, more expensive but with more coverage geographically?

The science of economic trade-offs

Abstract photo illustrating conflicting concerns in supply chain
As a consequence, SCM is the science of economic trade-offs. It implies a detailed understanding of the costs attached to pretty much every decision, every step of the way, and finding a strategy to reach the best balance to generate the best ROI (Return on Investment) on a general level, as opposed to a narrow optimization of just one step, which could have an adverse effect for the whole. Therefore, the goal of SCM cannot be measured by a single indicator. It would be untrue or way too simplistic to say that SCM should reduce stock outs (what if the goal is to control stock outs for certain products to lower inventory costs and favor other faster rotating products? Or to voluntarily create a notion of very small series, quickly disappearing if not bought now for hard luxury), or increase service levels. There is no single dimensional way to monitor a supply chain.

As opposed to the management of operations (i.e. making sure raw materials are transformed properly at the production stage), SCM is a multidimensional problem that requires proper quantitative analysis of the network. In this regard, we can sometimes see that SCM - the management of the business side of supply chain - is opposed to supply chain engineering - the mathematical side -, when it should be one and the same. Business and quantitative analysis should mix properly to make the best of both worlds. Ignoring the quantitative aspect would reduce management decisions to gut feeling, while ignoring the business would simply produce meaningless numbers.

A growing complexity

The complexity of supply chain networks has risen steadily since the 90’s, with globalization, but not only. Where goods used to be bought in centralized superstores, ecommerce has now risen to the task, and with it whole new stages in the SCM, such as last-mile logistics. Where cars only came in black, there are now more and more variations for every product, with the choice of options and customization sometimes being left to the customer (for some brands you can even choose the color of the sole, base, laces, border…). The products can now be delivered directly to the consumer, with specific 2-hour windows. Even for retailers, multi-channels have opened more possibilities, with whole new brands of systems to handle them such as MOMs (Multichannel Order Management solutions). More choices, more complexity, and to stay abreast in the market, there is no choice but to learn how to navigate.

Lokad’s take

Lokad is a transversal actor whose goal is precisely handling the complexity and providing the quantitative analyses whilst monitoring your need to navigate between choices to find the right trade-offs. We gather the data from multiple sources, we analyze your constraints and business specificities across departments and with the help of state-of-the-art technologies such as Machine Learning or Differentiable Programming, we try to consider every possible scenario, feasible decisions, and their economic impact on your supply chain as a whole. This is in fact the credo of the Quantitative Supply Chain.