Manuel Davy, President and Founder of Vekia.
In recent years, many leading textile brands have accelerated their international expansion. In a time of increasing pressure on selling prices induced by the digitization of trade, retailers will seek to increase their turnover and margin in new areas of mass consumption.
The Middle East, the United States, South America, as well as countries such as India, China and Kazakhstan represent very fast growth markets and strategic development regions. Even with aspirations to expand, it’s a complex process to implement and master. Along with these new regions come some difficulties. Among these and perhaps the most important challenge is the ability to evolve their logistics organization, especially when adapting and managing stock to an international context.
Two Major Issues: Supply Lead Time and Local Specificities
Companies in pursuit of global expansion are confronted with the physical distance between their headquarters and their warehouses from these new markets. In addition to increasing logistics costs (such as transport and customs), the distance causes delays and makes piloting replenishments more complex. It’s imperative that these expanding retailers recognize the local specific adaptions to be made consumption particulars (such as colours, sizes and styles). This results in more references and stocks to be managed.
Local specificities have a heavy impact on demand and consumption peaks. Retailers must respond to seasonal episodes when high demand is concentrated in short time periods.
In the Middle East, the period surrounding Ramadan brings more foot traffic than usual to textile stores. Global stock is mostly sent at the same time and creates a high level of clogging at customs and potential shut down. As a result, retailers are likely to anticipate this peak period to align their supply management, while taking into account of the seasonal delay of deliveries.
The celebrations specific to certain regions of the world are also opportunities to offer promotions and incentives to consume (such as Black Friday in the United States and Singles Day in China). These holidays further emphasize the importance to anticipate precisely.
When a brand expands internationally, a fair demand forecast without overestimating or undervaluing it becomes a complex and raises important themes. The lack of precision can quickly eliminate any hope in developing the company margin. Supply and demand management is often carried out at central headquarters, which can be thousands of miles away. The piloting is delicate without local information or the use of advanced tools.
These points can be handled by the organization, specifically the correspondents attached to the supply chain organization, as well as the main territories. Technology has an important role and can easily bridge the gap of geographical distance between headquarters and correspondents.
Machine Learning at the Service of Advanced International Logistics
The new world of trade and distribution is now governed by “Big Data”, which is often defined around the V4S Model:
- Volume (for the constantly increasing amount of information)
- Variety (because the broad nature of the data).
- Veracity (because data is always partially erroneous)
- Value (because data is a great added value for the company)
- Speed (because data is collected, analysed and evaluated in real time)
These data allow a more accurate and more thorough knowledge of consumption habits to better anticipate them, provided they have the right tools to analyse them.
A few weeks ago Korean prodigy, Lee Sedol’s world-renowned victory in the “Google AlphaGo” game, named him best player in the world. This illustrates the immense progress of AI and its capacity to analyse situations of great complexity in a controlled time (to the game of go, the number of combinations to be explored with each shot amounts to 10170).
Managing a Supply Chain internationally means managing a number of factors and combinations. The AI contributions made of have become indispensable. Anticipating timely shipments is one of the keys to the success of international development. The capability of forecasting demand as closely as possible while taking into account all local characteristics and constraints is imperative. It is also necessary to define the shipments and orders to be made each day by integrating all the specific complexities and real world constraints such as production capacity, transport, preparation, reception, cash management and coherence of supply.
Fine-tuning the flow in near-real time with maximum decision-making support is precisely what Machine Learning is capable of.