From as early on as 2012, Amazon have been using data analysis to anticipate customer needs before even the customer figures out what he or she wants. Through analysation of behaviour, data specialists can forestall what the customer will buy before arriving at the ‘buy now’ check out.
By this point, Amazon are already transferring the products deemed likely to be bought by the customers from distribution centres to local hubs in a bid to be closer to the customer’s home address. It is Amazon’s goal to deliver these products to the local hubs an hour or less before the customer even realises what they are about to purchase and of course before hitting the ‘confirm order’ button.
Some of you may be reading this and thinking Amazon is totally jumping the gun here but in fact, this is only just the ‘baby’ steps in this field. In the future with a fully functioning data analysis field up and running, Amazon aims to deliver the products to the home of the customer before the purchase has even been made. Worried you will receive the wrong stuff? Worry not! According to Marketing Professor Mark Galloway, Amazon will complete your order with a separate box for things that the customer doesn’t want to keep. He believes that returns are a fundamental part of this development process and is even considering allowing the customers to keep their unwanted items as a bonus scheme! Even though Amazon are aware that they could very well get the order wrong and customers will, from time to time, receive unwanted items, they are convinced that within these deliveries will be something that the customer will like. (of course, reflecting their personalised data analysis) Albeit this is a risky business but nevertheless, a cheaper option than the entire returns process.
Some may think this all sounds very efficient and of course heading in the direction of technology that we, as society, expect to be heading. Others, however, could also be thinking that this could be classed as an invasion of privacy and may, for accountable reasons, not be on board. The US store Target for example has recently come under fire for sending baby coupons to teenage girls. Target was testing a system based on customer purchasing history against 25 suggestive products relating to pregnancy and new-borns. Teenagers buying extra skin lotion and vitamin supplements must mean their expecting a child, no? Well, one father of a teenage girl was in fact caught red-faced after he stormed to Target and demanded an explanation on top of an apology as to why his daughter was receiving such baby coupons when it later turned out that she was actually pregnant! How could Target know more about his child than him? Yes of course, this could have just been a fluke and doesn’t necessarily determine the accuracy of future data analysis but nevertheless this is a strong indictor as to the direction we expect from Me-tail.
In the next decades we envisage the datafication of business processing to improve purchasing, logistics and general forecasting. Only in the later stages can we expect that the use of data will offer opportunities to extend forecasting to individuals. Koos Nuijten, a lecturer in Computer Sciences at the University of Leiden states, ‘Datafication will first lead to more efficient processes and only later to new and improved business models. All the rest for the moment is still in the hype stage.’
All that being said, the predictive power is data is already being used to its optimum level for most companies internally. To give an example, Dutch Albert Heijn supermarket chain uses customer data to restock its stores automatically. The supermarket’s built-in replenishment process continually monitors the purchasing behaviour of its customers and predicts what they will buy in their next visit, hence making sure this stock is readily available on the shelves.