The different technologies of artificial intelligence: an optimisation lever for e-merchants

The different technologies of artificial intelligence: an optimisation lever for e-merchants

There are different technologies of artificial intelligence. Used correctly, they can be a real lever for optimizing your e-commerce activity. The technologies of artificial intelligence that we will discuss in this article are robotics, customization, intelligence logistics, visual search, chatbot, etc.

1. Personalization and improvement of the customer experience

Personalization of the customer’s journey on the commercial site

Artificial intelligence makes it possible to identify customers according to their characteristics (location, age, gender, etc.). And their behavior on the site (pages visited, time spent per page, number of clicks, products viewed, the application used, etc.). According to the customer’s profile, navigation on the site is personalized in terms of sorting results, product recommendations, etc. Given the intensity of competition in e-commerce and the increased demands of customers. Personalization is becoming a key functionality that the main players of e-commerce have now implemented (private sales, Decathlon, Monoprix, etc.).

Start-ups play an important role in these advances in personalization. For example, Early Birds, the 2017 winner of the FEVAD START ME UP challenge. Offers e-merchants artificial intelligence algorithms for personalizing the customer journey. These algorithms use customer data. As browsing path on the site, centers of interest, Facebook likes, purchase history, etc. To establish their typical profile.

Predictive and targeted marketing with automated management

Beyond product recommendation, artificial intelligence algorithms are also beginning to make it possible to manage marketing promotion campaigns. Capable of predicting the propensity of customers to spend on the merchant site (according to their socio-demographic profile and behavior). They allow precise targeting for the various marketing operations in order to optimize their return on investment.

For example, marketing promotion campaigns are only triggered for customers. Who are characterized as undecided for whom the algorithm predicts that promotion could trigger the act of purchase? Marketing expenses for so-called “hot” customers (who would have bought anyway). Or “cold” customers (who will not proceed to the purchase act anyway) are thus saved.

Here too, start-ups play an important role. Founded in 2013, the start-up Tinyclues offers a solution for targeting marketing campaigns with a 2-step algorithm:
– Analysis of product buyer data
– Analysis of data from thousands of other consumers in the merchant’s customer database to identify those with the same characteristics. This segment is then targeted by the promotional marketing campaign.

Solutions around the personalization of the customer experience are the most common. With more than 70% of operational solutions based on product recommendation and personalized marketing.

2. Robotic and intelligent logistics: logistics and stock management for merchants

The most common and mature application of artificial intelligence in e-commerce today applies to logistics management.

For example, Cdiscount has been using robots in its warehouses since December 2017. These robots have the particularity of moving in 3D and are managed by a control system equipped with artificial intelligence algorithms.

Ocado, the English e-merchant, has taken the solution one step further with its ‘honeycomb’. Which not only speeds up order processing and enables more efficient stock management. But also significantly reduces warehouse space.

This solution allows a location as close as possible to town centers, a key factor in the success of in-line food production. Monoprix is in the process of adopting this solution in France.

In a context where multi-channel commerce is growing. Merchants need to have a unified, reliable, and real-time vision of their stocks in order to optimize them. Combined with Cloud technology, artificial intelligence is also able to manage different “click-and-collect” or “ship-to-store” scenarios. In order to find the best shipping point for each order. Thus reducing storage and transport costs.

3. Visual research

Pattern and image recognition is a technology that is becoming more and more mature with direct applications in e-commerce… Visual recognition brings together several technologies. That enables the machine to understand, via learning by analogy, the different media (images, videos, etc.). Coming from social networks, submitted by users via their smartphone or tablet, etc.

In e-commerce, fashion, decoration, and furnishing brands are taking advantage of this technology. To enable product searches using images instead of words. Thus, the user can point the camera of his smartphone. Or take a picture capture on the web to search for an identical product at the e-merchant. This makes shopping more fluid, fast, and interactive. Thanks to the multiplication of images shared on dedicated social networks (Instagram, Snapchat, Pinterest, etc.). As well as artificial intelligence algorithms for pattern recognition will improve rapidly.

4. Management of marketplaces databases

In order to master the complexity of the products and information they manage. Marketplaces have integrated AI-based solutions. Millions of items are passing through the marketplaces. It is essential for them to control the nature of the products to ensure, among other things, that they comply with French legislation and standards. They must then classify them in the right tree structure so that the consumer can easily find them when searching. Then there is the problem of incomplete or erroneous descriptions transmitted by suppliers. For example, a description: ‘red dress’ without specifying ‘strapless’: an essential search criterion.

The big marketplaces of pure web players have understood that the heart of the competition of the future will be played on artificial intelligence and data. These have started the race for artificial intelligence a long time ago. By collecting huge amounts of data in order to feed algorithms that are ever more data-intensive to ingest in order to be ever more relevant and intelligent. These great pure players are thus equipped with platforms that are e-commerce ever more advanced and intelligent. Then at the same time, they benefit the masses of consumers, their customers, and potential future customers, intelligent wizards (Amazon Alexa, Google Home, Tmall Genie). Only a handful of the marketplaces have the capacity to take up the challenge of artificial intelligence. Through their network flexibility, the talent of their engineers, and their innovative corporate culture.

Application examples :

Alexa is the most visible part of the development of IA at Amazon. One of its most lucrative applications of ATI remains less visible and integrated into its operations through predictive algorithms. For targeting customers with products likely to interest them (35% of sales by one estimate). Amazon also has 100,000 robots across its entire logistics network. has embarked on a partnership strategy around automation. In order to optimize logistics operations, greater speed and efficiency, lower costs, increase revenues. Considering that it has reached the limits of traditional technologies and methods to lower costs and reach its maximum potential.

Alibaba has invested heavily in its voice assistant Tmall Genie. As well as its customer service assistant Chatbot who now handles 95% of customer requests. Alibaba has also invested in its recommendation engine as well as in the operational optimization of delivery routes.

– In France, marketplaces such as Rakuten and CDiscount have developed their own recognition solutions. Enabling them to analyze large volumes of information, unparalleled with the capacity for manual analysis. Thus offering a wider choice of products and more frequent range renewals.

5. Chatbot for managing customer questions and requests (pre- and post-sales) and personal assistants


More and more e-merchants are using a more or less advanced chatbot on their site. Or on social networks (Facebook Messenger is the most widespread) to answer customers’ questions. About products before the sale and to manage various after-sales problems.

Chatbots allow eMerchants to interact with their customers any time and to respond to a wide range of requests and queries almost immediately. It’s an exciting innovation in the eCommerce space. And for eMerchants who need to be instantly and accurately respond to their customers, chatbots are a complete game-changer.

Global eMerchants like H&M, Pizza Hut, Sephora, eBay, Nike, and Duolingo have all introduced chatbots to engage with their customers and to drive sales. Each case is different, and each chatbot is personalized. Like customers, no two chatbots are (or should be) the same.

The most commonly used chatbots are those you’ll find on websites (an embedded chatbot on an online portal). And social media platforms (as wrote previously). They can either be rule-based chatbots, which are basic and only respond to specific commands and cannot react to complex requests. Or chatbots controlled by Artificial Intelligence (AI), which do not have to respond to specific commands only, can understand language. And can become smarter through real-time learning from their ‘conversations’ with users.


Now you are aware of some of the technologies of artificial intelligence. You should try to implement them in your company. They can add significant value to it.

If the subject of artificial intelligence or innovation interests you. Take a look at these articles in this category.

Do you have any questions? Would you like to discuss this subject with us? Contact us or leave a comment. We will be happy to hear from you.

Leave a Reply

Your email address will not be published. Required fields are marked *