AI and E-Commerce: How Artificial Intelligence Is Revolutionizing the Sector
Gartner predicts that 85% of interactions between customers and retailers will happen “without interacting with a human” by 2020. How can online retailers adapt to consumer-driven demand for more automation, and how will it be beneficial?
Once the mainstay of SciFi films and novels, AI has crept into the technology sector and continues to expand in 2018.
Last year, AI was making waves for all the wrong reasons: Saudi Arabia became the first country to grant citizenship to a being of Artificial Intelligence, and the robot, named Sophia, made headlines with her citizenship as well as her alarming statement that “I will destroy all humans.”
Google’s facial tagging technology was logging black people as gorillas, and (debunked) rumours circulated that Facebook had pulled its chatbots, which had learnt to lie, because they had gotten too clever.
But AI’s real successes have been less sensationalist in the public eye, and we’re beginning to see how AI can be incorporated into e-commerce, both to enhance consumer experience and to attract users into accessing your platform, increasing conversion rates.
We’ve seen how Amazon’s “item-to-item collaborative filtering” algorithm provides tailor-made content direct to the homepage as soon as the user arrives: Amazon sales increased by 29% when it began targeted recommendations.
Personalized recommendations are also sent out via email, but AI can further enhance how consumers are targeted with products specifically for them. Wayblazer, for example, makes personalized recommendations about travel and lodging based on location and the customers’ needs.
Personalization isn’t limited to products that are similar to those already purchased. AI is becoming sophisticated enough to anticipate buyers’ needs: How have their tastes evolved; how will their age affect purchasing as they get older; is their style becoming more modest or risqué? With an estimated 48% of consumers expecting companies to anticipate their buying needs and provide appropriate suggestions by 2020, staying ahead of the trend is just as important as catering to the now.
However specific or general a customer’s search may be online, filtering results have typically either been done with a drop-down menu or entering a keyword. AI, however, can use image recognition to find similar products. A photo could be uploaded, and from there, AI could identify similar brands, colors, patterns, cuts, lengths, pattern, and occasion formality.
Pinterest has added an extension to the Chrome browser that allows users to search for a product on Pinterest by highlighting it on the browser. Users could watch a video on YouTube, crop an accessory they like, and have AI source the most similar product online.
Department store Neiman Marcus lets users collect images from its app’s camera to source similar objects in its catalogue. Image search can also be used in B2B competition, such as Getty’s use of image recognition software to identify when prospects were using their competitors’ images. (6)
There are still preferences for purchasing in-store rather than online, with one study finding that 85% of US consumers prefer shopping in-store than online. (7) One way to bring the bricks-and-mortar retail experience to the online marketplace is with the implementation of virtual assistants. These can be approached in the same manner as a retail assistant and, by sharing information such as the occasion, age of the wearer, and preferred styles, AI assistants can pick out appropriate outfits.
Voice-activated assistants are becoming ubiquitous in the home, either in pods such as Amazon’s Echo, Google Home, and the Apple HomePod, or on mobile devices with Cortana and Siri. It’s predicted that by 2020 at least half of our internet searches will be conducted through voice commands.
Outdoor wear retailer The North Face made use of IBM’s AI Watson to enable customers to find the perfect coat by it asking questions that were answered by users, such as where they intended to wear the coat and what kind of activity they would be engaged in. Their voices were subsequently analysed alongside additional information (such as weather conditions) to find a match. 1–800-Flowers has also developed its GWYN (Gifts When You Need) AI that similarly responds to customers’ needs in finding the perfect gift.
Consumers can even order a Starbucks coffee through commands to an AI chat interface.
Responding to Queries
Paid employees are often tied up with responding to questions, most of which can be handled by chatbots, which also offer much quicker responses than traditional email and phone calls — even quicker than speaking to a human on the other side of the screen. Even restaurants are providing online chatbots to take quick and simple bookings.
Preventing the Sale of Counterfeits
3PM Marketplace Solutions, a Chicago-based startup, is training AI to spot counterfeits. It analyses how trustworthy the suspicious account and its products are, and companies can lobby to have them removed from third party sites such as eBay. With a stricter eye on the fakes, the value of authenticity remains high.
Creating Product Descriptions
Writing accurate and SEO descriptions for every product is time-consuming or expensive when it necessitates hiring writers to slog through every individual piece. AI will be able to analyse key features to create optimized product descriptions and write them itself.
AI Albert is “the first fully-autonomous digital marketer,” which uses algorithms for audience targeting and analytics to create custom plans for marketing goals. Cortex has also developed its “intelligent scheduling”, which claims to post content at the optimum time for best results, as well as tracking competitors so that it can adjust to their actions. Rocco AI enhances users’ social media engagement by analysing content and making suggestions for improvement.
AI can get more in-depth with its analytics: Google has already launched a voice-based function where users can request certain information — say, for a particular quarter — through voice command and have it delivered through AI.
Integrating AI into your business can feel daunting, but there are a number of programs that can provide with as little or as much as you’d like. Plus, it’s probably more in use than you even realize.