Leading ecommerce companies keep increasing their investments in AI. These five trends shed the light on the future of e-commerce business.
By Darya Yermashkevich
eCommerce is one of the main areas of applying artificial intelligence in business. According to the Cowen and Company IT Survey, over 20% of AI investments flow to online retailing. Today, we will look at the directions of these investments and shed the light on how current trends shape the future of e-commerce.
AI to change the way retailers talk to customers
The success in eCommerce business is defined by how well merchants make shopping convenient, fast, and effective. For this reason, AI mainly entered eCommerce in its frontend part, or the point of contact between shoppers and the web store. In particular, the hottest innovations are focused on improving search and product recommendations.
#1. More accurate and flexible search options
Customers have become intolerant of irrelevant content. According to Janrain, 73% of shoppers expect retailers to improve the accuracy of search suggestions. And this is exactly the point for AI to come onstage. How?
First, Natural Language Processing and Artificial Intelligence can make semantic search engines more precise. Unlike the search focused on keywords, the application of machine learning allows recognizing typos, synonyms, and product relations.
Second, AI offers more search options based on voice and image recognition. Voice search is gaining popularity for its convenience. According to Google, 20% of searches in 2016 involved voice recognition. ComScore predicts that this figure will reach 50% by 2020.
#2. Smarter AI assistants and chatbots
The main drawback of online stores that keeps their brick-and-mortar predecessors alive is the lack of human assistants. Though there are chatbots coded to respond in a particular way, they can hardly replace people who can listen, understand, and talk. AI changes the customer experience game as it allows shoppers to communicate with a web store via text, voice, and pictures and enjoy human-like responses.
Case: The North Face XPS
As an example, take the North Face XPS powered by IBM`s Watson. This recommendation tool asks a shopper questions about location and gender, as well as does additional research to find out the weather conditions in the shopper’s location. As a result, the customer gets several most relevant options to consider. Now the tool is limited in understanding non-trivial answers and typos. Still, XPS shows how online shops will work some years down the line.
Case: Rakuten Fit Me
Until July 2018, another popular shopping assistant powered by AI was Rakuten Fit Me recommendation technology. Rakuten Fit Me asked a shopper about their body shape and age, and then found the items that fit these parameters.
Case: Amazon Alexa
Actually, AI assistants are not only about the search for particular items. Let’s take Amazon Alexa and Echo devices. Alexa is a cloud-based voice service available on Amazon and third-party devices. Thanks to the Skills Kit (a repository of self-service APIs, tools, documentation, and code samples) Alexa is getting smarter every day.
One of the latest Amazon innovations based on Alexa is Echo Look, a hands-free camera and personal stylist that utilizes cloud data storage, machine learning, and data analytics. Echo Look can analyze a person’s outfit and suggest advice based on the latest fashion trends.
Alexa, Google Assistant, Microsoft’s Cortana… Such personal assistants are the future of online apparel shops. As they can not only help a customer find the item they need but also help them understand what exactly they need to look great.
#3. Precise upselling and cross-selling recommendations
Machine learning makes cross-selling and upselling offerings more relevant to customers and thus more effective. Recommendation engines powered by AI can analyze customer data and behavior and anticipate their needs with smart recommendations.
In fact, the algorithms are getting sophisticated enough to understand how non-trivial factors can influence buying behavior (like changes in preferences with age). Algorithms can learn based on previous customer reactions in particular circumstances, determine the most effective tactic and apply it in a similar situation. The early adopters of AI-based recommendations have already seen positive results: Amazon got a 29% increase in sales and Netflix achieved 75% of successful item suggestions.
AI to change the way retailers work backstage
Apart from influencing customer experience directly, AI transforms the backend part of e-commerce, which is marketing and warehouse logistics.
#4. Insights for better marketing and planning
To illustrate the marketing shifts caused by AI, let’s turn to the Rakuten case. Rakuten Institute of Technology has developed an AI algorithm that analyzes 200 million products traded on Rakuten Ichiba and delivers forecasts of the sales volume with high accuracy. Smart predictions of sales help merchants manage logistics costs better and eventually let them strike the balance between supply and demand.
One more application of Rakuten technology is fact-based segmentation of shoppers. Normally, marketers use their wit to identify segmentation parameters (like gender or age). AI can distinguish customer segments leveraging more subtle criteria, like particular product preferences or shopping behavior.
Another AI benefit for eCommerce marketing is the analysis of reviews. Customer feedback is the key to identifying your weak points and improving service. On time reaction to negative and positive reviews helps a company to increase customer loyalty and prevent negative publicity if something goes wrong. Machine learning algorithms can analyze the content of reviews and mine valuable data from page descriptions and reviews.
#5. Logistics based on 100% automation
Finally, online retail business depends a lot on the efficiency of logistics processes. No wonder, AI entered this sphere as well. AI and robots can help to improve the speed and efficiency of warehouse operations, reduce the need for employees and thus cut costs and increase revenue.
Let’s take a look at JD.com. This Beijing-based e-commerce leader has the ambition to become a 100% automated company. For this, they invested in an AI-powered retail research center and entered into several strategic business partnerships. Today their logistics processes are automated to a max possible extent:
- Warehouse robots stack products on shelves, pack, and ship merchandise;
- Delivery drones transport products across China.
The first warehouse that needs no man to function is planned to start operating in November 2018.
Current trends show that for now mostly eCommerce incumbents can enjoy AI implementation in the backend operations. Humanlike digital assistants, robot-operated warehouses, and shipment with drones – it seems like the lines of a fiction novel have become a part of our everyday life. For sure, now the booming AI is too expensive to go outside the circle of market leaders. Still, it’s obvious that the rest of the market will soon start to keep up as the technology will keep advancing.
As an Amasty author, Darya Yermashkevich researches into tech trends and the ways IT can help businesses stay competitive. She shares insights and ideas through business posts on the blog and reputable online resources. After work, Darya loves writing fiction stories, practicing yoga, and hiking in picturesque places.