The predictive logistics system patented by
Amazon in December 2013 is a preemptive attempt in pointing to the future of
logistics. But is this system truly
feasible? So, let’s look into the
futuristic vision of Amazon’s new logistics system. There you will find a new role of warehousing
capitalizing on the so-called big data.
What can be done by the predictive logistics
system?
The predictive logistics system is highly
indicative of the near future vision of logistics, and it makes the following
tasks and functions possible by analyzing enormous customer data (that is big
data).
It permits a pinpoint prediction of the
customer wants in terms of goods and services.
It permits a prediction of the purchasing
schedule of the customer for certain merchandise.
It permits a prediction of additional item(s)
simultaneously bought with the merchandise.
The patent application submitted by Amazon
for the patent registration describes that utilization and analysis of big data
enables highly accurate prediction of the customer needs and wants, and Amazon
preemptively ships and delivers the merchandise to be ordered and purchased
soon by the customer. In order to realize this epoch-making logistics, Amazon
is now building the necessary infrastructure for its own unique system.
Prior to looking into the mechanism of the
new logistics system, let’s see how Amazon analyzes big data to precisely
predict the impending customer order.
What is the six-point data analysis for
Amazon’s prediction of the merchandise wanted by the customer for purchase?
Amazon indicates that in its predictive
logistics system , the analysis of the following 6points can permit the
prediction of the customer’s future behavior.
(1) The
track record of the customer’s orders in the past.
(2) Data
analysis of the customer’s merchandise and key word searches on the Amazon
site.
(3) Analysis
of the wish/want lists prepared in the past
(4) Analysis
of the past data in the shopping carts.
(5) Data
analysis of past cancellation and merchandise returns.
(6) Analysis
on the time spent when the customer’s cursor stopped for particular category
and merchandise.
In addition, incorporating analysis of other
data such as the time zone that the customer availed of, the interval of
orders, interests in relevant merchandise, and the custoemr’s personal data,
Amazon comes up with a predication of the future ordering and purchasing
behavior of the customer.
Why then did Amazon come with such a
predictive logistics system?
This predictive logistics system of Amazon
was invented from the needs to enable the same day shipment and delivery. In America with such an enormous expanse of
its land, same-day shipment and delivery are almost impossible under the
current capacity of logistics.
A few-days time lag for shipment and delivery
after having received the order from the customer is inevitable due to the long
distance in geography. An interesting
data shows a correlation between the cancellation or merchandise return between
the days spent from the order placed by the customer and the arrival of the
delivery in Ecommerce. Thus, for
Ecommerce traders, same-day shipment and delivery is a big challenge for better
sales performance and reducing cancellation or returns.
A the present, Amazon’s logistics process is
as follows;
(1) The
order placed by the customer
(2) Putting
of the address label on the merchandise ordered inside its warehouse.
(3) Loading
merchandise on a truck chartered from a delivery or transport company.
(4) Actual
delivery
(5) Arrival
of the shipment several days after the merchandise ordered.
If shifted to the predictive logistics process,
(1) Predictive
shipment
(2) Delivery
of the merchandise to the local or regional hub closest to the residence of the
customer, and holding it there until the actual order to be placed
(3) Actual
order generated by the customer
(4) Delivery
of the ordered merchandise to the customer
(5) Same-day
arrival of the shipment
If prediction is made for
the order to be placed by the customer, the same-day delivery becomes possible
at all times. And the reduction in cancellation and merchandise return rates
can contribute to the maximization of profits.
But in order to realize this, one needs to retool and overhaul his
logistics system from A to Z.
What is the nitty-gritty of Amazon’s unique
predictive logistics system?
The same delivery on the customer’s order is
not feasible if shipping directly from the warehouse to the customer. Instead,
Amazon preemptively transfers the shipment to the local or regional hub
geographically closest to the customer, and once the actual order gets generated,
the merchandise is immediately shipped from the hub. For this purpose, Amazon locates local or
regional distribution hubs or centers across the nation. The warehouse in those
hubs is not designed for sorting the packages of the shipment, but rather for
temporarily storing the pre-ordered merchandise. Another important aspect of this is the
coordination of one such hub to another in a different but neighboring region.
If an order placed by the customer in particular hub, and no stock is available
at that time there, you can have the same merchandise reserved for another
customer in an adjacent hub transfer to him, and significantly reduce the time
for delivery. The concept of regional hubs and distribution centers help enable
the same-day delivery and enhance the customer satisfaction.
The main function of distribution hub is
stand-by warehousing.
It is often becoming the case with many major
distribution firms also in Japan with a network of regional warehouses across
the country. These regional warehouses play the role of distribution center in
particular region where merchandise is consolidated, sorted and loaded on the
truck for delivery to each and every address of the customer. But Amazon
advances a step farther than that by positioning those regional warehouses as
stand-by distribution centers to temporarily reserve the shipment of
merchandise predicted to be purchased by the customer by utilizing the analysis
of big data gleaned from the past behavioral patterns of the customer. Some
might challenge this kind of logistics arrangement by asserting that rather
than resorting to the practice of predictive logistics, you might as well just
build regional distribution centers for the fast moving items. However, there
is no assurance how fast those supposedly fast moving items are actually
ordered, and chances are that there would be a lot of backlog sitting and
collecting dust in the warehouse. There is no guarantee for even such fast
moving items to be purchased and
shipped out in a timely manner in a particular region. What is sold well
on a national average does not necessarily reflect the situation of any
particular regional market. While big data analysis of the customer’s past
behavioral patters and personality characteristics permit highly accurate
anticipation for the near-future action to be taken by the customer. You can
safely say to the extent that it is in the cards that the customer is scheduled
to order particular merchandise for sure, since such analytics gives a fairly
precise prediction of the customer’s purchase decision any time soon. Overflowing stock of those supposedly fast
moving items demand additional manpower for maintenance, administration and
sorting inside the warehouse, and it will push up logistics cost. On the other
hand, in Amazon’s hub warehouses, only those merchandise anticipated to be
ordered and shipped out very soon are reserved there, cycle time of the
inventory is kept brief and wastage is avoided to the maximum. Amazon’s
anticipatory logistics system is thus changing the function of warehouse
management as well. That is, temporary stand-by reservation of the merchandise
anticipated to be ordered and shipped out soon for sure. It is a significant
paradigm shift in the world of logistics.
In the future of EC business, the
marketer capable of consolidating, analyzing and utilizing big data for
warehousing and distribution is likely to prevail, and those left behind this
kind of new trend will lose out and get perished in fierce completion.
No comments:
Post a Comment