Retailers need to make their products discoverable through conversational AI while giving shoppers confidence in how their data is used, says Lidewij van den Ham, senior vice-president of consumer acquisition and engagement at Mastercard Europe.
As AI assistants offer new ways to research and compare products, she sees an opportunity for retailers to understand more about customers’ needs than traditional keyword searches reveal.
Go deeper with GlobalData
Discover B2B Marketing That Performs
Combine business intelligence and editorial excellence to reach engaged professionals across 36 leading media platforms.
“Instead of typing a few keywords into a search bar and working through a list of results, shoppers can use AI tools to help them find the right product, explaining what they’re trying to achieve and seeking guidance to refine the options based on their individual needs and preferences,” she says.
Large language models, which underpin many conversational AI tools, can interpret the context behind these requests and help shoppers explore suitable options.
Van den Ham gives the example of someone planning a wedding who asks an AI assistant to find a dress suited to a particular style, venue and budget. Recommendations could extend to shoes, accessories, beauty appointments, travel arrangements and gifts.
“For retailers, this presents a significant opportunity,” she says. “These interactions provide a much richer picture of purchase intent than traditional search, and retailers can use this insight to deliver more personalised and relevant recommendations, better understand how shoppers move from consideration to purchase, and engage consumers with their brand and products at the moments when they are most receptive.”
Making products discoverable
For retailers seeking visibility through conversational AI, van den Ham’s advice starts with product information.
“As AI becomes a more prominent part of the shopping journey, the most successful brands will be those that can ensure their products are surfaced in response to real-time customer conversations and behaviours,” she says.
“Clear descriptions, accurate attributes and strong product data help conversational tools determine how and why a product is relevant and when it should be recommended.”
Alongside making their ranges easier for AI systems to interpret, retailers face another challenge: connecting customer information held across their businesses.
Van den Ham says retailers already have substantial data across advertising, customer relationship management, product, loyalty and digital systems.
“The challenge isn’t the collection of data, it’s effectively bringing together these signals and understanding which ones should be acted on to improve the customer experience,” she says.
AI can analyse browsing activity, searches, clicks and additions to baskets, alongside purchase history where available, to identify patterns that may indicate what customers need.
“That allows retailers to respond dynamically rather than relying on static experiences that are identical for every shopper,” she says.
Van den Ham points to Mastercard’s Dynamic Yield personalisation technology as an example. She says it combines real-time behavioural signals with historical performance data to identify relevant recommendations, including complementary products and alternatives to unavailable items.
Building trust in recommendations
The value of personalised recommendations depends partly on whether shoppers understand why they are receiving them.
“The key is to ensure AI feels assistive, not intrusive,” van den Ham says. “Personalisation is most effective when consumers can clearly understand why they’re being shown a particular recommendation.”
An alternative to an out-of-stock product, or an item that complements something already in a basket, offers a clear connection to the shopper’s immediate needs.
“As AI takes on a greater role in commerce, trust will be the differentiator and retailers must ensure personalisation remains transparent and grounded in permissioned data, or risk undermining customer confidence,” she says.
Consumer control becomes particularly relevant when AI moves beyond offering recommendations to making purchases on someone’s behalf.
Van den Ham cites Dentsu’s Checkout Society research, which found that nearly two-thirds of UK consumers were uncomfortable with AI making purchases for them. The finding concerns delegated purchasing rather than AI-assisted product discovery, but illustrates the importance of distinguishing between advice and authority to buy.
“As consumer confidence around new technology grows, retailers must make their customers feel supported through their shopping journeys, while also in control of the process and secure in the knowledge that their data is safe,” she says.
For van den Ham, ethical personalisation rests on transparency, consent and a clear benefit to the shopper.
“Consumers should understand the reason a product, offer or piece of content has been surfaced and feel confident that the experience is working in their interests.”
Connecting AI investment with results
Retailers also need the organisational foundations to translate AI investment into commercial value.
“Despite having access to more customer data than before, many organisations still struggle to create a unified view of the customer and their intentions because valuable insights remain fragmented across different systems,” van den Ham says.
She argues that appropriate processes, governance, resources and operating models must accompany investment in technology. Experienced specialists can help retailers implement systems and identify which data signals are most useful for improving customer experience and conversion.
Measuring results remains a further challenge.
“Demonstrating measurable impact remains one of the biggest challenges for retailers and CMOs,” she says, referring to chief marketing officers.
McKinsey’s State of Marketing Europe 2026 report provides broader context for that investment. It found that 50% of chief marketing officers ranked generative AI-enabled marketing among their three fastest-growing investment areas. The research covered multiple industries across five European countries, rather than retail alone.
Van den Ham argues that delivering relevant recommendations when purchase intent is strongest can help retailers connect personalisation with observable outcomes.
“Hyper-personalised marketing strategies can help address that challenge by delivering recommendations and offers when customer intent is at its highest,” she says.
“This increases the likelihood of consumer engagement with the offering and makes it easier to connect the delivery of a personalised experience to a commercial outcome like a purchase or click.”
Preparing for peak trading
For retailers serving markets where Black Friday, Cyber Monday and Christmas are major trading events, van den Ham’s priority is to turn existing customer insight into relevant shopping experiences.
“Retailers must use AI to make better use of the customer data available to them, and turn insight into relevant interactions, recommendations and experiences as shoppers look to be guided through the product discovery journey,” she says.
“The ability to connect consumers’ needs with the right product or message at the right time will be key to commercial performance ahead of the peak trading period.”
