With the advent of big data analytics, there have been some bold moves in supply chain optimization. Not too long ago, Amazon announced that it can read its customers like a book.
In December, Amazon patented a system that it hopes will dissuade customers from ever entering a physical shop again. It’s being dubbed ‘anticipatory shipping’; and the media has been quite enamoured with the prospect of a company delivering goods before the customer has even realized they wanted them. The Wall Street Journal was among the first to drop the story, and opened with the line ‘Amazon.com knows you so well it wants to ship your next package before you order it.’
It sounds nigh on impossible, right?
Maybe not as much as it first seems. The International Business Times wondered whether it was all just another marketing ploy but, although Amazon only issued a patent as opposed to acting on this idea, it is an ability that is already being enjoyed by many of my customers. And there is no Big Data involved with regard to the established definition (there are indications that Amazon might be using Big Data only for certain scenarios). This is because optimized warehouse network planning, using advanced analytics and operations research, is already a component of INFORM’s add*ONE software. It tackles one of the biggest problems in inventory management: making sure stock is readily available for the consumer, but not overstocking or overcrowding warehouses. Go too far either way, and big money can be lost.
Network planning uses forecasts based on historic dispatches to determine demand in local distribution centers and forwards this information upstream through the network. This way, demand information is bundled at the starting point of the distribution network – the central warehouse. Based on the demand information from the local nodes, the central warehouse is able to hold just the right amount of inventory, avoid local shortfalls in stock and keep delivery times short across the network.
It is one of those things: at first you think it’s not quite possible, but then you discover it is, and it becomes the lifeblood of your business. Normally, central warehouses must maintain sufficient stock levels in order to be prepared for a sudden surge in demand, coming as a domino effect from further down the distribution line. However, this is just what I call ‘safety stock’. A large part of it is never used, eventually finding its way to the proverbial waste paper basket. The enterprise then has to deal with growing costs as the capital represented by this stock accumulates. With daily, self-adapting forecasts, safety stock levels can be accurately re-calculated in order to make certain that the right stocks are in the right place, at the right time, across the entire network of warehouses. add*ONE recommends optimal stock levels and movements from multiple sources of supply to achieve minimum cost and maximum service.
You can learn more about add*ONE’s ability to give product predictions here, as well as download a case study of one of our customers, SFS intec, and find out how they have optimized their network planning process.
So this kind of pattern-based, predictive behaviour is not only possible, but it’s already in existence. As true as the results have been for many companies using add*ONE, this potential seems to have surprised a lot of people. They are amazed that ‘this decrease in delivery time has the potential to fulfill a very strong want – the never-ending consumer desire for instant gratification,’ but what is ‘potential’ for Amazon customers is already a reality for add*ONE end users (though I’m pretty sure, they are already using conventional network planning right now). And let’s not forget that Amazon is only at the patent-filing stage! There are no concrete plans for this service to roll out in the near future.
At the end of last year, I read an article that quoted Richard Powell, managing director and co-founder of supply chain consultancy Crimson & Co, as saying the term ‘big data’ had become widespread in 2013, “but most supply chains are yet to really understand what they need to do about this”. I think, there is a lot of potential in drawing information from Big Data for certain supply chain processes, but it is still very risky if you don’t have money to burn and want to run a tight inventory strategy.
It’s neat that they’re able to do this right down at the consumer level though, where there is typically not enough data to generate reliable forecasts. As you say, this is just a patent for the time being, but it will be interesting to see if it makes it into production.
Comments are closed.