Go to: Section menu | Main menu | Footer | Top of Page



Guessing the future thanks to big data

Amazon orders will be sent to us before we buy the product. Or almost

  • Key4Communications

We have already spoken on occasion of big data - how companies can manage all the information they collect from their customers to offer them a customized service. These tools are becoming increasingly important due to the digital trail we leave on the internet. The best known examples are Netflix, Ebay and Amazon, where we find a different home page every time we enter, customized to our preferences.

The management of these data has many possibilities. And not just in the case of large enterprises - an SME can identify its best customers and their preferences, and make personalized offers. Big data also allows us to find patterns in less active customers and adjust our services to encourage them to buy more. Moreover, with the right tools we can even predict who will buy what by when.

In fact, Amazon has launched a service called "anticipatory shipping" that shows how precise can be the predictions based on its learning algorithm.

Obviously, we will not receive at home that book we were just thinking of buying. The company will first send the packets to the geographic area of potential clients, specifying the exact address once the customer places the purchase (which can be done even on the road). The goal is to minimize delivery times, which are one of the main complaints of online shoppers.
And to do all this, Amazon bases its predictions on the products we buy, we consult, we search, as well as those we put in the shopping cart and then abandon, and also what other consumers with similar profiles do. We are not Amazon, but we can also work with some or all of this information in our companies, although we seldom do.
Of course and after the drones marketing gimmick, it is understandable to question if this is serious, even though the company has already patented the system in the United States.

For starters, these tools are not perfect - for example, if I use a single profile on a online shopping site to buy products for my entire family, when entering the web I can find recommendations based not only on my preferences, but also on those of my partner or on the purchases I may have done for my parents.

It is also a logistically very complex system: how many products can be delivered to each Amazon store, and how long can they remain there? Maybe Amazon is convinced that I will buy the only Martin Amis book I still do not own, but the company may not know that just yesterday I bought it in another library.

It is quite likely, though, that this service is coordinated with the tests being done for same day delivery, only available to premium customers.

The lesson is that this way of working is already possible. That is, we can anticipate orders with increasingly high accuracy and security, if we work with the data we already have.

  • La Tafanera
  • delicious
  • Facebook
  • Twitter
  • LinkedIn

Go to: Section menu | Main menu | Footer | Top of Page