Few industries can expect to remain untouched by the artificial intelligence revolution with automakers, healthcare giants and big banks all embracing AI technology. Consumers are now beginning to experience a whole new way of online shopping thanks to the launch of AI-powered digital assistants such as Amazon’s Alexa and Google Home.
Frank Palermo, executive vice president of global digital solutions at VirtusaPolaris, a provider of IT consulting and outsourcing services, believes the invention of dynamic AI solutions is a game changer for online retailers and consumers alike.
“Responsive retail has peaked – the next step for the industry is predictive commerce,” he says. “As AI assistants and machines begin to take over more of our purchasing decisions, retailers will have to become far, far smarter about their use of data to stay ahead of the pack.”
Digital consumers are beginning to tire of what feel like endless product choices, which can leave even the most eager shopper overwhelmed. US-based consulting firm Walker Information predicts that by 2020 customer experience will overtake price and product as the key brand differentiator. E-commerce companies should move to incorporate AI solutions if they want to meet the needs of increasingly demanding buyers.
The present picture
The growing appeal of AI solutions will massively change how consumers interact with online brands, especially when buyers utilise virtual assistants or chatbots to make their purchase.
“Marketing and brand identity will lose a great deal of power when coldly rational machines are making decisions. It’s all about providing the best deal at the right time. The key to this is understanding the context for any given purchase and matching that up to other possible needs,” says Mr Palermo.
The greatest strength of AI in this industry is its ability to adapt quickly to fluid situations where customer behaviour is erratic. This functionality is already being used by e-commerce brands to present consumers with items they didn’t even know they needed, according to Abhinav Aggarwal, chief executive of artificial intelligence startup Fluid AI.
The company’s software was being used to predict customer behaviour for one of their clients when the system saw a sudden change in user behaviour from a city which was experiencing a snow storm.
“Users who would typically ignore the e-mails or in-app notifications sent in the middle of the day were now opening them as they were stuck at home without much to do. Within an hour the AI system adapted to the new situation and started sending more promotional material during working hours,” says Mr Aggarwal.
Reaching consumers at the most convenient time led to a major jump in sales during the day and the following day, when the normal workday resumed, the AI adapted once again and reduced the number of messages.
Several e-tailers have gone as far as creating their own AI-powered platforms to help potential buyers pick the most suitable products for their needs or simply make the purchasing process more natural. However, some in the tech industry expect searching by text to be supplanted by image search in the future.
“2017 will see customers searching by uploading images of things they want,” says Carrie Lomas, IBM director for AI computer system Watson. “Image recognition allows retailers to present products similar and complementary to the one being searched for.”
There is, of course, a limit to AI’s e-commerce applications, with businesses needing to be aware that consumers won’t switch to voice or text platforms overnight. “Just like online didn’t completely kill the bricks-and-mortar store, chatbots will not kill online shopping. This is especially true for millennials, who on average look at as many as ten to fifteen reviews before making a purchase,” says Mr Palermo.
Jon Darke, director at user experience design agency Every Interaction, agrees. “You can’t research multiple options or compare things easily with voice. Having a user interface provides many benefits for the vast majority of online uses that will always be superior to spoken interaction,” he says.
It is yet to be seen if online shoppers will use AI-powered services as simple customer service platforms to resolve complaints and answer questions or as fully fledged digital concierges, which will help consumers select the most appropriate product.
The future of AI
The next step in AI technology will be a fully cognitive website which offers a completely personalised experience to users. Through AI and machine-learning, buyers’ habits, preferences and behaviours can be comprehensively understood, allowing e-commerce firms to customise how they interact with consumers visually to enhance the experience.
“Fully cognitive websites will drastically change buying behaviour and selection. We see the average spend of a buyer per session significantly increasing with these websites. Cognitive sites take consumers through a much smarter funnel, allowing them to buy before user fatigue starts to kick in or the user gets distracted,” says Mr Aggarwal.
For example, based on previous buying patterns, these websites will know when your toothpaste is running low or when your next dinner party is likely to be and prompt you to buy at the exact time products are needed.
“New AI-based products combine a deep and granular understanding of products with real-time analysis on shopper intent. By understanding what shoppers are doing, for example what products they are clicking on, the AI can then return successively better recommendations to the shopper until they find a product they want to buy,” says Jon Epstein, chief marketing officer of AI startup Sentient Technologies.
Buyers will expect more tailored offers from e-tailers and for the entire purchasing journey to be as natural as possible, without any unnecessary interruptions. If companies can find the right balance of AI and human analysis, then more of the 67 per cent of online shopping carts that are abandoned each year, according to comScore, will be checked out.
Companies that don’t capitalise on AI technologies to enhance customer experience face being left behind by competitors who do. However, AI has some way to go before it can perfectly interpret the complexities of human behaviour.
“Consumers still expect authentic interactions. For instance, don’t make the mistake that because someone liked several pictures of a friend’s new-born baby on Facebook that they automatically want to buy baby food. Consumers don’t want to be sold to, they want to be engaged with,” Mr Palermo concludes.