E-book – The Digital Salesperson – Reinforcement Learning in E-commerce
Get comprehensive insights into the use of reinforcement learning for personalizing online shops.
Shopping as an experience: What brick-and-mortar retailers have long been good at is now becoming increasingly accessible online thanks to AI (artificial intelligence) software. An online fashion store and a bookseller show how it's done.
Here's what you can expect to find in this blog article:
Why AI software is important for your online store
How to integrate AI software into your online store
How to communicate with your online shoppers using AI software
Best practices: These shops are successfully using AI software
How Outletcity Metzingen achieved over 5% more revenue per session
How Ex Libris generated 10 times higher product email sales in its online shop using AI-supported email campaigns
Conclusion: Optimized KPIs and digital shopping experiences thanks to AI software
Artificial intelligence in e-commerce? It's one of the hottest topics around, but many retailers are struggling to find concrete approaches that deliver measurable success. Outletcity Metzingen and Swiss online bookseller Ex Libris are showing how it's done: The fashion company increased sales per session by over 5% with the help of AI software, and Ex Libris increased its product email sales tenfold with AI-supported email campaigns – all while fully complying with the GDPR and the privacy of every online shopper.
To answer this question, let's take a look at brick-and-mortar retail. Compared to online retail, we find that:
These are restrictions that you, as an online retailer, do not have to deal with. Brick-and-mortar retailers, on the other hand, have the opportunity to establish one-to-one communication with customers and serve them personally and individually – at least before the coronavirus pandemic. When it comes to returning customers, employees on site can draw on their experience to respond to specific needs and requests. This personal connection remains one of the key success factors for brick-and-mortar retailers. With the sheer volume of online shoppers in e-commerce, this was hardly conceivable—until now! AI software makes it possible to establish one-to-one communication with almost every single customer online.
Stay up to date on personalization: Sign up for the epoq newsletter. Register now!
According to an article in ecommerce magazine, the use of artificial intelligence is worthwhile in many respects. According to Gartner (a provider of market research results and analyses), AI will have the greatest impact on all aspects of commerce. The article states, "Analysts predict that the majority of companies using AI for digital commerce will achieve at least a 25 percent improvement in customer satisfaction, revenue, or cost reduction by 2023."¹
In order to communicate personally with every online shopper, you first need to build up a knowledge base for your online shop and send the algorithms involved back to school. You start by filling this knowledge base with your wealth of experience and your product catalog. The basic principle here is: the more structured information, the better.
For tips on how to optimize your product data so that it boosts sales, check out the blog article Product Data – Sales Booster for Your Online Store.
Then the learning process about the behavior of visitors and customers to your online store begins. To do this, your online store is simply connected to the knowledge base with a single line of code. All information about the click and purchase behavior of each individual online shopper is then fed into the knowledge base, processed by the algorithms, and fed back into your online store.
Using modern AI methods such as reinforcement learning, the algorithms ultimately optimize themselves in a closed-loop process and communicate with your customers in an increasingly personalized way. With the help of this technology, you can make AI tangible for your online store.
Example: A customer visits an online store and navigates to the "Blouses" category page, where she uses the faceted navigation to narrow down her selection to pink and plain-colored blouses. She particularly likes a pink blouse, so she clicks on it. Based on the customer's behavior (filter restrictions, clicks, etc.) in the online store, the algorithms learn that the customer is potentially interested in pink blouses. This knowledge can be used, for example, when displaying recommendations. If the customer now visits a product detail page, pink blouses can be displayed as recommendations based on her personal preferences.
In a next step, the algorithms could learn the customer's size, for example, through further actions such as clicks/selections. In this way, the intelligent algorithms can continue to develop through self-learning, allowing communication to become increasingly personalized.
Behind AI software lies a knowledge base with complex algorithms that operate in real-time software services (SaaS) in online shops. They ensure 1:1 communication between customers and online shops, which benefits various stages of the customer journey.
For example, you can:
You therefore communicate using the same shop functionalities as before, but in a more intelligent way. In many cases, this allows you to create 1:1 communication, just like in brick-and-mortar retail, which enhances the shopping experience and generates higher sales for you.
Example: The online shop fahrrad.de uses AI-supported recommendations on its home page. If a customer visits the shop for the first time and therefore has no history, two recommendation widgets with top sellers and new products are displayed on the home page. If the customer then navigates through the shop, clicks through different categories, or views specific products, this click and purchase behavior is recorded and processed. The customer now has a history that can be used to display personalized recommendations. If our example customer now navigates back to the fahrrad.de homepage, they will see a third recommendation widget with "personal recommendations" that match their learned preferences.
Now you've learned a bit about AI for your online store. But what does the practical application of artificial intelligence look like? We'll show you two best practices examples of how well-known stores are successfully using AI software to optimize their KPIs.
Personalization is an important success factor at Outletcity Metzingen for ensuring a consistent shopping experience in the online store. For this reason, Outletcity Metzingen has been using a recommendation system for the personalization of product recommendations for some time now. In order to ensure the best possible display of individual recommendations for each customer, Outletcity Metzingen wanted to use the technology described above and compare the personalization strategy used to date with the previous provider in the form of an A/B test. The result was clear: with personalized recommendations based on AI technology, Outletcity Metzingen was able to increase sales per session by over 5% and record an increased shopping cart value and conversion rate. You can find all the information on this in the case study Previous personalization strategy put to the test.
Stay up to date on personalization: Sign up for the epoq newsletter. Register now!
In order to inform customers 90 days in advance about a new book release, Ex Libris developed the desire to develop and implement an AI-supported process. Only recipients who were potentially interested in the newly released book were to be contacted via email. This AI-supported calculation was implemented automatically via personalized emails. Product email sales increased tenfold as a result. This implementation even won the 2018 Email Award. You can read all further information in the case study Increasing e-commerce sales.

Automated target group identification based on AI for email campaigns from Ex Libris
Artificial intelligence is becoming increasingly important. By using AI software in a targeted manner, you can achieve 1:1 communication with your shop customers, which ensures a positive shopping experience and also optimizes your KPIs. AI software can be used at all touchpoints along the customer journey. Many companies, such as Outletcity Metzingen and Ex Libris, have recognized the potential of artificial intelligence in terms of personalization and are already successfully exploiting it in their own online shops. Thanks to AI software, both retailers have been able to significantly increase their sales. Examples like these show that the potential of artificial intelligence should not be overlooked and that its use is definitely worthwhile.
Sources: ¹ ecommerce magazine (2020)
Learn more about the use of artificial intelligence in e-commerce in our e-book:
Willy-Brandt-Straße 3
76275 Ettlingen
+49 (0)7243 2001-0
You are currently viewing placeholder content from HubSpot. To access the actual content, click the button below. Please note that doing so will result in data being shared with third-party providers.
More InformationYou are currently viewing placeholder content from HubSpot Meetings. To access the actual content, click the button below. Please note that doing so will result in data being shared with third-party providers.
More Information