Online shops predict what customers want to buy
E-commerce platforms have brought a veritable Aladdin’s cave of products right into our homes. As a consumer, you no longer need to go to a shop, instead, the shop will come to you.
Digital platforms such as these amass a vast amount of data from their customers, which they analyse to enhance the customer experience. Once customers are happy with an online shop, they begin to trust the seller and will likely frequent the shop again.
The world’s largest online marketplaces, for example, Amazon or Alibaba, have revolutionised competition and the rules of traditional trade. Information management and artificial intelligence make shopping more and more addictive, but an astute customer can also reap the benefits of online shopping.
From multichannel to omnichannel
Back in the 1990s, shops managed each of their sales channels separately. The brick-and-mortar shop operated as one entity, with an online shop as another separate entity and telesales as a third. At that time, shops obtained consumer information mainly from receipt data. With the advent of loyalty cards, more detailed customer information began to be gathered.
‘This multichannel approach was followed in the 2010s by an omnichannel one, where the customer decides where they do business. From the consumer's point of view, the shop forms one entity, in which they may first visit a bricks-and-mortar shop to physically see the product, which they then order online. Transactions are independent of time and place,’ explains Lasse Mitronen, Professor of Practice at the Aalto University School of Business.
He has researched global marketplaces and the platform economy in several different projects in collaboration with Mikko Hänninen. Hänninen's current position is as an assistant professor of commerce at the University of Nottingham in the UK.
‘A traditional supermarket or hypermarket offers 25,000 food-related items and 30,000 other products. Amazon has 650 million products. The difference in product ranges is inconceivably large,’ says Mikko Hänninen.
Online shops brought customer information to a new level
While traditional commerce consists of a single sale, the revenue generation model of e-commerce consists of selling goods, payment processing, advertising space for the media, and value-added services such as Amazon’s digital content. They are all based in one way or another on the data economy.
When expanding its operations, an online shop collects an ever-increasing amount of information from its customers, which they must analyse reliably. On the other hand, the customer's trust must also be able to be maintained.
Each click a consumer makes while surfing the internet is tracked and analysed by a digital robot. Its learning algorithms then turn data into a marketing tool. The more data available, the more reliable the predictions of consumer behaviour are.
‘Machine learning and automation are used to screen the collected data. Data in itself is not worth anything unless we can use it to make interesting and useful observations,’ says Hänninen.
Digital giants like Alibaba, Alphabet, Amazon, and Facebook make the best workable models for predicting consumer behaviour. They have hired the best IT specialists in the world for the job.
‘Even with small test samples, 80% of the data can be used to predict how well the models will work. The models are then duplicated into a huge database to screen customer groups and their behaviour,’ says Mitronen.
The companies’ algorithms detect correlations in their data that are useful for marketing automation.
‘The models help create different customer segments and predict whether or not you will become a good customer in the short term or over a five-year period. As a result, the importance of digital marketing has also increased.’
Every click is tracked
An online shop such as Amazon collects a mass of data from its customers that identifies each consumer’s movements, clicks, and time spent on each product. This digital big brother knows your dreams and helps them come true – by buying products.
Whether a consumer benefits or loses depends on their awareness of such tactics.
Your everyday life can be improved when an online shop knows your taste in music or what movies you have watched and on which websites you shop. Thus, artificial intelligence also serves the consumer.
But consumers are often concerned about the use and possible manipulation of their data. UN Secretary-General António Guterres has also addressed this concern about how user data is being sold for advertising and to sway public opinion.
‘It can be harmful to a consumer if their information is used for a different purpose than originally authorised. In response to such concerns, the European Union passed the General Data Protection Regulation (GDPR). The GDPR’s purpose is to guide organisations’ actions and help prevent the misuse of consumer data. Online consumers are also concerned with credit card security and identity theft,’ Mitronen explains.
‘The excellence of e-commerce is justified by the fact that it removes trade barriers and makes shopping easier. But we should remember that algorithms work well and correctly only as long as their purpose is also for good from an ethical point of view,’ adds Hänninen.
One example is your health. Suppose you have osteoarthritis and have done an online for information on this condition. In that case, the algorithm will save that information, and then your email may be flooded with ads for products to help those suffering from knee conditions – some of which you might find useful.
The United States deemed it ethically unacceptable that a young woman’s secret pregnancy was revealed to her family by baby product ads created using algorithms.
Algorithms can also lead you down the wrong path entirely. One interesting case occurred in the UK when a person making a homemade bomb ordered materials such as steel wire from Amazon. Thanks to algorithm-generated ads, subsequent customers who also bought steel wire were shown special offers for acid, which is also used in bomb-making.
Companies have started to demand strong identification from their customers to improve their position and avoid risks. Various security breaches and crimes have fueled this development.
‘Gradually, companies like Amazon and Alibaba may join the telecommunications business, gaining access to internet and phone usage data. And next, they may look to expand into the healthcare sector.’
Amazon already owns, for example, the PillPack online pharmacy and sells prescription drugs, among other things.
‘Also, Amazon Care, a digital healthcare service, is in its pilot phase. All the big technology giants have these experiments underway.’
They are also developing new service packages around pharmacies.
‘We can measure consumer confidence by whether a consumer is willing to provide access to their healthcare information,’ says Mitronen.
Text: Helena Raunio
Photo: Jaakko Kahilaniemi
This article is published in the Aalto University Magazine issue 28, May 2021 (facsimile copy on issuu.com).