AI in Ecommerce: Personalization and Beyond

Are you laboring under the belief that AI is a buzzword? It’s much more than that, especially when it comes to improving ecommerce personalization.

Many industry writers still don’t take artificial intelligence seriously, introducing it as ‘the latest buzzword.’ This is surprising because many online retailers are already making use of the AI capabilities built into ecommerce platforms (for example, SAP Hybris) to transform the customer experience and increase the revenue. Ecommerce personalization, customer relationship management, and inventory control are retail elements benefitting from AI today.

Spend on cognitive and AI systems in retailImage source:

With retail industry spending approaching the $2 billion mark, AI clearly deserves recognition as an established technology in ecommerce development rather than a buzzword, but the misnomer may arise from a simple lack of clarity about what artificial intelligence is.

AI Technologies

Artificial intelligence is a collective term used to describe a set of technologies that aim to make computers behave as intelligent agents (like a human or, in some aspects, even smarter). The most promising field of AI today is machine learning - a set of computational techniques aimed at making computers learn by processing huge amounts of data and finding patterns in them. In this article, we’ll find out how machine learning and, more precisely, data mining - the process of digging into massive data sets to unearth trends and patterns with the help of machine learning¾are reshaping ecommerce.

For now, AI in ecommerce is used primarily for customer experience and personalization purposes. However, the scope of possibilities is much broader, and AI is also transforming ecommerce back-office management activities like website analysis and process optimization.

Let’s continue now with a closer look at how AI personalizes the ecommerce customer experience. In some ways, it does so more effectively than any human could ever do.

More Personal than a Person?

To see how AI-based personalization works, let’s look at SAP Hybris and its recommendation engine. It uses data mining to help suppliers and retailers upsell, cross-sell and encourage impulse purchases by online shoppers. With an engine like this behind your ecommerce store, a customer looking at a specific product can be presented with complementary or related products tailored to her individual tastes, in real time.

These recommendations are based not on general predefined rules (although programmed rules do play a part in the process), but on what the engine learns from historical data relating to an individual customer’s behavior and preferences, as captured during past interactions between the customer and your enterprise.

Personalized product recommendationImage Source:

As AI personalization software learns more with every new transaction, recommendations presented to returning customers become increasingly accurate over time. Personal recommendations like this would just not be possible for humans to achieve, even in a brick-and-mortar retail environment with a small product range, let alone when an online catalog comprises hundreds or even thousands of SKUs.

Beyond Personalization: The Future of Customer Experience

But with platforms like SAP Hybris costing anything upward of $50,000, do AI features represent real value for your business, or are they just nice-to-haves, with their importance inflated by hype? If the findings from Gartner’s 2017 digital commerce research are anything to go by, including the prediction that personalization software will lift profits by 15% for ecommerce retailers, artificial intelligence seems to be something way more serious than just a fancy gimmick.

Growth of AI in ecommerceImage Source: Jake Rheude - The Growth of Artificial Intelligence in E-commerce

Of course, there is more to creating a great customer experience than making it personalized, and the good news is that AI can make online shopping more intuitive—and a whole lot more convenient. Let’s take an example of an AI-based conversational agent (or simply chatbot). This is a computer program that uses natural language processing (NLP) techniques to analyze text queries and respond to them in a human-like manner. To achieve this goal, rule-based chatbots use simple sets of ready questions and answers, while AI-based chatbots learn on huge sets of real textual information to respond to rather complex and unexpected questions in the most natural way (although they still have some limitations).


Perhaps a hypothetical scenario, based on what an NLP-enabled chatbot might accomplish, will help to highlight the benefits of artificial intelligence in ecommerce.

The Scenario

A shopper visits your ecommerce store looking for a particular item. Instead of picking through a tag cloud or typing in keywords, the customer pages a virtual sales associate (an NLP-enabled chatbot) and begins a natural typed conversation.

In moments, the chatbot has presented the shopper with a range of available items that closely match the shopper’s description of the item she seeks. Images, prices, descriptions, and consumer product reviews are right there in front of the shopper’s eyes.

With the item placed in the customer’s shopping basket, the chatbot ventures some personalized recommendations for complementary items. The customer adds one of those items to her shopping basket after the bot tells her a personal discount is applicable because this is her tenth store visit this year.

Before finalizing her purchase, the shopper realizes she has a couple of questions about the item she’s buying. The chatbot answers her first question but the second is too complex for an electronic assistant to answer, so the bot seamlessly diverts the customer’s chat session to a human, who resolves the query and transfers the customer back to the chatbot.

Finally, the bot checks out the customer’s order and accepts online payment, and what we almost forgot to mention was that the customer completed the entire process on her phone while riding the train home from work.

Artificial Intelligence, Real Benefits all Round

It’s safe to say that convenience is as much a part of the customer experience as personalization, and AI can offer both in an integrated manner. The customer in the scenario received the following benefits:

  • She didn’t have to browse or search a webshop on the small screen of her smartphone.
  • Her entire shopping experience relied on a text messaging app and was no less convenient than any text conversation.
  • Her online shopping excursion was fast and fuss-free.
  • She had no need to navigate between product and checkout pages.
  • She made a saving on the complementary item she purchased.
  • She was able to make a complex enquiry without becoming frustrated by AI limitations.

In addition, the entire experience was uniquely personalized. Meanwhile, the retailer was able to:

  • Capture the sale and increase its value.
  • Gain some new data about the customer, the products, and the sales channel (mobile).
  • Minimize the need for human effort in providing customer help.

While the last point may not amount to a huge reduction in the individual transaction cost, when multiplied over a year’s worth of customer service/help enquiries, it can make a difference to the number of helpdesk seats that need to be occupied by employees—and that will be a meaningful saving.

AI in the Ecommerce Supply Chain

While your customers are enjoying a futuristic online shopping experience, AI can also help you stay on top of your business and its supply chain. For example, by mining inventory, sales, and purchasing data, and applying machine learning to understand the patterns, an AI solution integrated with your ERP, WMS, and ecommerce software supports the transition from reactive to proactive inventory management.

How can AI boost inventory accuracySource:

By using AI to predict and forecast requirements even more precisely, your company stands to save money and time and improve stock availability by optimizing order patterns. So again, the result is a win-win for your business and customers.

What’s good for your customers is also good for you as a buyer. As artificial intelligence gains adoption in the B2B environment, there will come a time when your company can benefit from AI functions on the supply and demand side.

In fact, there may soon be no need for any manual purchasing processes whatsoever, as another element of AI technology—machine-to-machine communication—increases in sophistication. But that’s a topic for a future article.

Artificial Intelligence in Ecommerce: Beyond the Hype

Whether applied to personalization, customer service, supply chain management, or business improvement, artificial intelligence adds up to real advantages for ecommerce operators. Buzzwords are for ideas and concepts, and hype is what happens when marketers think they see a good thing on the horizon.

AI is no longer a conceptual technology—it is here right now!

AI is in the process of disrupting and transforming ecommerce, and yes, this is just the beginning. You can sit back for a while and see where it will go, or you can get onboard and go with it. Whatever you do though, don’t wave it off as one of the latest buzzwords, because that ship already sailed—and some of your competitors may well be on it.

Content type


It’s simple!

Please wait...