In 2016, the artificial intelligence and machine learning hype cycle reached fever pitch. Weve been here before; it seems as if revolutionary artificial intelligence has been on the cusp of reality for decades. Its understandable that many in the eCommerce world are skeptical, particularly smaller eCommerce merchants who have to be careful where they invest resources.
Although the benefits of AI / ML are often overblown, its clear that it will make and is already making a real difference. The contrast between this and other AI changes everything hype cycles is significant. The technology in the limited domains relevant to eCommerce has advanced enormously: were better at data analytics on huge scales, were better at natural language processing and pattern recognition, were better at building cheap and scalable infrastructure, and the market has matured: many companies offer products and services that use AI / ML to provide real-world benefits to retailers without requiring a phD in computer science.
Artificial intelligence is the application of automated systems to decision making, the discovery of solutions, and the delivery of insights. That sounds great, but it doesnt mean anything without practical applications, and, in 2017, theres no shortage of eCommerce companies and solution providers who are building those applications.
Gartner says 85% of customer interactions will be managed without a human by 2018. The prognostications of analysts are to be taken with a pinch of salt, but those working in the eCommerce field are hardly less enthusiastic. According to Keenan Baldwin, co-founder of SiteZeus:
Online retailers are scrambling to partner with or adopt new AI technologies to help facilitate customer interaction to try and match and even surpass the typical in-store experience
Lets have a look at two of the ways eCommerce merchants are already using AI to augment the experience of both retailers and their customers.
Visual search
Shopping is often a visual experience: we see something we like and we look for something similar we can own. Until recently, eCommerce search has been, for the most part, text-based. Visual search allows customers to upload an image and find products that match that image. The customer uploads an image of a pair of shoes they like, and the store returns products that match that image.
Pinterest allows its users to upload images which it will then search for product recommendations, suggesting potential purchases based on the products in the image. Cortexica offers similar capabilities that can be integrated with any store or app. Its findSimilar and ExactMatch APIs can process uploaded images and suggest matching products from a catalog.
Personal shoppers
The ideal personal shopper knows all about what you love and what you buy, and, using that information, can suggest products that match your style and budget. Although it may seem ironic that artificial intelligence is the key to a personalized shopping experience, its actually the perfect application for artificially intelligent algorithms that can process large amounts of information from eCommerce stores along with personal data to preemptively offer the right product at the right time.
Mona is an iPhone application that does just that. It learns what the shopper wants, and presents only the products theyll be interested in. From the users perspective, they avoid the hassle of trawling through multiple stores to find the right product, and the eCommerce merchant benefits because theres more chance the shopper will make a purchase.
These two practical examples are just the tip of the iceberg you can bet your bottom dollar that Amazon and other mega-scale eCommerce merchants are plowing billions into eCommerce AI, along with service providers who want to make their mark selling AI services to smaller stores. As the market matures, it will behoove merchants to think about how the eCommerce experience might be enhanced by artificial intelligence.
