Advances in technology and innovation make the future we saw, just a few years ago, in some cinematographic films, not as a fantasy any more but now becoming part of our daily life.
It comes to mind “Minority Report,” “The Fifth Element”, “2001: A Space Odyssey” and “Robocop.”
HAL 9000, the computer that we saw in “2001: A Space Odyssey” could interpret the feelings of the characters and was able to predict when they would turn it off. We do not know if Amazon was inspired by the film when creating “anticipatory shipping”, but this algorithm is something similar because it can predict the demand of users for the purpose of sending products before being purchased and they can be delivered in an hour.
In 2012, Amazon presented a patent for a delivery system designed to reduce delivery time, by predicting what users will buy before they were actually buying it and shipping the products to their homes even before the sale has been made: “Method and System for Anticipatory Shipping Package“.
Today, the Amazon commitment with users and buyers who choose the Premium service marketplace is the one or two-day delivery. Although there are already some locations where the deliveries can be made in less than 24 hours. The aim of the e-commerce giant is to minimize delivery times and in order to make this possible, the key are the so-called predictive models.
The idea is aiming to ensure that online shopping can be as immediate, as to buy goods on a physical location, in which you can pay and take the product at the same time.
How is this possible?
Big Data plays a key role in this. Thanks to stored data in the order history of each customer on Amazon, the online sales giant can predict how many new books of Dan Brown’s latest novel, how many Samsung 7 or how many drones will be sold in a given area: neighborhood, city, etc.
This model simulates the reality by analyzing the customer behavior in the past and even though it is not exact, it allows to approach the number of orders that will be carried out. Given this estimate, Amazon sends a variety of products to its distribution centers and even trucks assigned to an exact area. From the user finally ‘clicking’ on the button, until the arrival of the purchase at home, the shipping can take just a few hours or less.
We cannot ignore that carrying packages from one place to another, has a cost. Even for Amazon. But if there is anything that characterized them in the past, is having an efficient logistics that allows them to make a difference and their excitement to innovate.
What happens if their predictive model is wrong?
This part also is under control through a system that is able to calculate whether it is worth (speaking of costs, of course) to return the product to the plant or if it is better to send discounts and promotions to customers in the area where the products are (always a lower value than the cost of returning the product to the plant). The idea is not so far-fetched. It allows offering great deals to customers and making them happy, as well as the liquid stocks are reducing costs and maintaining a complex logistics machine running.
Amazon is very sure of their prediction algorithm and its reverse logistics capabilities, which in the US is now already sending products to certain customers without them even done “the click”. It is already in the last phase, where they will be able to know the purchase of a product even before we know it. Really incredible and disturbing, they know more about us than ourselves.
Implementing the model worldwide where customers are counted by millions, with hundreds of distribution centers to control all across the planet, and thousands (maybe more) suppliers is not an easy task. But it is undoubtedly a brilliant step to the future and a unique way to differentiate them from the competition.
Alan Key said:
❝The best way to predict the future is to invent it❞.