How AI is transforming online shopping this holiday season

If there’s anyone who should be worried about artificial intelligence making their jobs obsolete, it’s elves. From personalized gift-giving recommendations to more targeted marketing campaigns to stronger security measures, AI is stepping in as our secret holiday helper, enhancing holiday shopping from product discovery to checkout.

Though in-store shopping still makes up the majority of total retail spending, online shopping is increasing at a faster pace. On Black Friday, online retail sales grew 14.6%, while in-store sales grew 0.7%, according to preliminary insights from Mastercard SpendingPulse, and the Mastercard Economics Institute expects online retail sales to grow 7.1% year over year during the holiday season. AI may help that growth to continue.

Here are three ways AI is transforming digital commerce.

01
A personal shopper — even when you're the one shopping for gifts

Some of us have the gift for gift-giving; some of us lack the knack. AI-powered personalization tools are already helping e-commerce brands tailor their shopping experiences to our individual preferences. But what about those of our loved ones? A year ago, Dynamic Yield by Mastercard launched Shopping Muse, a personal AI assistant that shoppers can talk to via a chat interface. The assistant engages with consumers through natural language conversation, responding to specific inquiries to come up with the right product bundles, whether you’re shopping for yourself or someone else. For example, one could ask it, “What should I gift my 15-year-old son for the holidays?” and continually refine the results with additional questions.

Using image recognition technology, Shopping Muse can find items visually similar to those recommended, with the ability for shoppers to offer suggestions based on an uploaded picture coming soon. “We’re continuing to push the boundaries of what conversational commerce can achieve, delivering more relevant recommendations per each individual’s unique needs, whether known to the brand or unknown, and making holiday shopping smarter,” says Dynamic Yield CEO Ori Bauer.

02
Responsive shopping experiences — and smarter marketing blitzes

Personalization goes far beyond product recommendations, says Bauer; it makes the overall shopping experience more responsive. For example, if a retail site knows you’re shopping for gifts, you might see what’s called “social proof” prominently featured: customer reviews, testimonials, or even a running tally of how many other shoppers have purchased an item recently or have it in their baskets. This helps help you home in on choices and have more confidence in your ultimate selection. Plus a cart full of products might trigger an alert about shipping deadlines, to ensure your purchases make it in time for the holiday, or reach the threshold for free shipping with items priced to help you get there.

And at a time of year when most shoppers are already oversaturated with ads and emails about deals, some brands are being smarter about connecting with potential customers. The most powerful personalization engines are tailoring email campaign content and recommendations based on a user’s most up-to-date interactions with a brand, whether on a site or within an app. This can be done automatically based on their most recent likes or dislikes, fueled by deep learning models, with creative visuals that can be dynamically adjusted on both a one-to-many and a one-to-one basis, influencing users to click “buy” instead of “unsubscribe.”   

03
Better protection

Artificial intelligence has been the secret sauce in payments security for years, scanning billions of transactions for fraud patterns, spotting data breaches to alert banks before their cardholders fall victim and assessing the validity of online transactions in real time. In 2023 alone, Mastercard’s AI prevented $20 billion in potential fraud losses.

Now, using generative AI techniques, our cards are even safer. The company’s Decision Intelligence Pro, launched earlier this year, scans 1 trillion data points — including account, purchase, merchant and device information — to predict the authenticity of a transaction even faster, boosting fraud detection rates on average by 20% — and as high as 300% in some instances. It also reduces false positives by more than 85%, lowering cardholder frustration by flagging fewer legitimate purchases as fraudulent.

“Unnecessary friction or, even worse, having your card compromised by fraudsters are the last things holiday shoppers want to worry about,” says Johan Gerber, Mastercard’s executive vice president for Security Solutions. “The ask from consumers is simple: They want their card to work without friction or fear of fraud. That is our focus: trust and peace of mind.”