An experiment involving bananas, detergents, Gouda cheese, supermarket shoppers, smartphones and a web-based app has been playing out in the Netherlands over the past few months, and the results could be hugely important for the future of traditional brick-and-mortar shops as they try to fight back against the increasing might of online retailers.
As part of the trial, supermarket clients scan the articles using their mobile phones before placing the goods into their shopping baskets. A click sounds and the article pops up on the screen.
Until recently such self-scanning of products was only possible using special equipment installed at supermarket check-outs. Now the Dutch discount supermarket chain Albert Heijn is testing the smartphone-based scanning technology with a view to rolling it out in its stores.
More than 200 stores are taking part in the experiment which involves customers making use of free in-store wi-fi to connect with the shopping center.
Customers only have to enter their personal discount card details once and the data is thereafter stored on their phone, without the need for them to find their card each time they go to the store.
“The success factor for online traders is that they know so much about their customers”
The experiment underway at Albert Heijn stores is primarily about convenience, speed and automation but is also about the surveying of customer behavior, data analysis and how to respond to the pressure from online retailers – whose warehouse-based business models means that, compared with traditional shops, they have lower overheads because they do not need to pay for costly retail space.
Another massive advantage that online retailers have enjoyed is access to a wide range of data about their clients. They can tell which region a customer is accessing their internet site from, which operating system they use and even what content their customers share on social media.
Analytics software tells them how long a customer has been viewing a product on their website, which other products the client also liked and what he bought last time he visited the site. Meanwhile, intelligent algorithms use this data to make product recommendations, to highlight special offers and make personalized offers.
By contrast, the managers of supermarkets or high street fashion shops often know next to nothing about their customers, only getting to know their customers’ preferences when it is too late: at the checkout when the client is about to leave the store.
And even this small amount of information is often unusable as the same customer might return to the shop a few days or weeks later unrecognized and unknown.
Marco Atzberger of the EHI Retail Institute in Cologne believes that the small amount of data about their customers traditional shops have is one of their biggest commercial disadvantages.
“The success factor for online traders is that they know so much about their customers,” he said.
“The shop owner gets to know the customer better and customers benefit from shopping lists, personalized offers and product suggestions”
While many traditional retailers have ignored this for years, they now have to catch up with the online retail trade. In Germany, internet retail sales has boomed in size from €2.3 billion in 2000 to an estimated €44 billion ($45.7 billion) in 2016.
Online retailers such as Amazon are constantly expanding their activities. The Seattle-based company, for instance, is now also targeting grocery shopping.
It is because of that pressure that traditional shop-based retailers are increasingly trying to imitate the sales techniques of their online rivals. This provides them with valuable insights into how customers behave while shopping, such as which items a customer likes to buy at the same time or when a customer interrupts his shopping as he walks through the store.
Antonio Krüger of the German Research Centre for Artificial Intelligence believes it will only be a matter of time before traditional shops adopt the same techniques as the large online stores.
The technology already exists, Mr Krüger said, for a mobile phone app to suggest other products to a customer buying celery and leeks in a supermarket’s vegetable section, such as, for example, a meat-based stock with which to make a soup at home. Alternatively, the app could offer the customer a reduced price on a six-pack of beer.
Such possibilities are an advantage for both shoppers and shop owners, Mr. Krüger believes.
“The shop owner gets to know the customer better and customers benefit from shopping lists, personalized offers and product suggestions,” Mr. Krüger said.
However, most retailers in Germany are still conservative when it comes to the use of intelligent data analysis and individual discounting systems. Self-scanning systems such as the one in the Netherlands, for example, are available in the German supermarket chain Globus – but not using a web-based app.
In eight Globus stores, customers can book their articles themselves with a scanning device and pay at a payment station.
“Couponing, shopping lists or product proposals are not available on the devices,” says Malte Wolters of Globus. “We regard the self-scanning system as a customer service rather than a source of customer data.”
Dynamic pricing – where prices change on certain days or at certain times – is the current buzz word.
German supermarkets are, however, keen to use new technology to improve their stock ordering to ensure shelves remain full.
Many supermarkets have more than 20,000 product lines and it is difficult to always order the optimal quantity to keep up with demand.
Until recently branch managers estimated their orders based on statistical evaluations and years of experience. But more and more supermarket chains are now relying on automatic decision-making systems based on artificial intelligence.
The Blue Yonder company in Karlsruhe is one of the leading European providers of such services and its software is used by German retailers such as Kaufland, Globus and Real.
“The stationary trade has for a long time been able to survive using the traditional techniques,” says Blue Yonder managing director Uwe Weiss. “But the new methods are simply faster and more accurate.”
Blue Yonder’s software systems use machine learning to learn from previous mistakes and to make reliable forecasts.
Such software allows retailers to identify simple relationships, such as mothers buying less bread and cheese during school holidays.
The software can also be used to predict how a special offer on Gouda cheese, for example, is likely to impact on the sale for Edam cheese and thus help the retailer order the correct amounts of both products to meet demand.
But Blue Yonder does not wish to confine itself to predicting customer behavior but also wants to control customers’ purchasing decisions. When it comes to this, dynamic pricing – where prices change on certain days or at certain times – is the current buzz word.
Mr. Krüger also believes that dynamic pricing is the way forward for retailers, not in the form of constantly changing prices but through individual special offers. “Anyone who does not use the new techniques will experience falling sales,” he said.
Thomas Schmelzer reports for Handelsblatt and WirtschaftsWoche from Düsseldorf. To contact the author: email@example.com.