How AI agents could reshape retail as stores move beyond sales
Per the full note from ING Think, agentic commerce — autonomous AI agents that search, compare and purchase on behalf of consumers — is shifting retail's centre of gravity away from the transaction and toward advice, service and experience, with physical stores recast as omnichannel destinations rather than pure sales channels. The desk's structural read is that the disruption lands hardest in repeat, replacement and standardised functional categories, where agents can exploit consumption patterns to time purchases, while luxury, trend-sensitive and experience-driven goods stay human-led because inspiration and emotion resist automation. ING Research's own segmentation, published under Katinka Jongkind's byline, draws the key distinction not against human shopping but against AI assistants such as ChatGPT and Gemini and against conventional comparison websites — agentic commerce is the step where the purchase executes without the consumer present. This is a slow-burn margin and channel-mix story for retail, not a rates or FX event: no bank forecast, no central-bank pivot and no high-impact calendar item sits behind it, so our coverage pages carry no consensus target, no per-firm spread and nothing to schedule. The tradable content is second-order — AI capex, cloud and semiconductor supply chains, and the currency exposure of cross-border e-commerce platforms — rather than a direct G10 call.
What the desk is arguing
Per the full note , agentic commerce — autonomous AI agents that independently select and purchase products for consumers — is forcing retailers to rethink what a store is for. The desk frames this as the next phase of digital disruption: e-commerce changed where people shop, and agentic commerce changes how, shifting buying from an active task to an automated background process.
The supporting evidence is a taxonomy rather than a data print. ING Research separates agentic commerce from AI assistants (ChatGPT, Gemini) and from comparison websites, and argues the technology's greatest potential sits in repeat and replacement purchases and in functional, standardised products, where an agent can compare price, specification and availability and use consumption patterns to time the buy. Adoption is explicitly expected to be slower for luxury, trend-sensitive and experience-driven goods, because inspiration, emotion and personal preference remain central and consumers are less inclined to delegate.
The alternative read the desk is implicitly rejecting is that agentic commerce is simply a better comparison site — a tool the consumer still sits behind. The note's distinction matters: once the agent executes, the retailer loses the point-of-sale interaction, and store economics have to be justified by advice, service and omnichannel support rather than basket conversion.
How other firms see it
There is no per-firm forecast set on this commentary, so there are no aligned or contrary houses to group — the piece is thematic retail and technology research, not a currency call.
The adjacent lenses worth watching are AI infrastructure spend and semiconductor supply chains, which monetise agentic commerce upstream of the retailer, and the cross-border e-commerce complex where platform margins carry the FX translation. Watch USD/CNY and EUR/USD for the translation effect on listed retail and platform earnings, and the broader AI capex cycle as the demand signal that tells you whether agent adoption is accelerating or stalling.
What the calendar says
No high-impact events are scheduled for this jurisdiction over the next 30 days, so there is no scheduled catalyst to position around. Treat this as a medium-term thematic input rather than an event-driven trade.
Key takeaways
- 01ING Think argues agentic commerce shifts shopping from an active task to an automated background process, changing how consumers buy rather than just where.
- 02The note's key distinction is against AI assistants like ChatGPT and Gemini and against comparison websites — agents execute the purchase, they don't just inform it.
- 03Greatest near-term potential sits in repeat, replacement and functional standardised goods; luxury and experience-driven categories stay human-led on inspiration and emotion.
- 04Physical stores are recast from sales channel to advice, service, experience and omnichannel destination.
- 05This is thematic research with no bank forecast, no central-bank catalyst and no scheduled event — the tradable read is second-order, via AI capex, cloud and semiconductor supply chains and platform FX translation.
Market implications
With no forecast, no level and no calendar event attached, the actionable signal is upstream: watch AI infrastructure and semiconductor capex guidance as the proxy for whether agent adoption is accelerating, and the cross-border e-commerce platforms where agentic volume shows up in margin and FX translation first. USD/CNY and EUR/USD are the translation channels to monitor for listed retail and platform earnings, but neither is a trade on this note alone.
Risks to this view
The thesis is invalidated if agentic commerce stalls at the assistant-and-comparison stage — regulation, payment-security liability or platform resistance could keep the consumer in the loop and preserve store economics. Conversely, faster-than-expected adoption in luxury or experience categories would break ING's core segmentation and pull forward the retail margin hit. A shift in the AI capex cycle would reprice the upstream beneficiaries faster than any retail channel-mix story.
Articles How AI agents could reshape retail as stores move beyond sales Published 13:59 Manufacturing, Construction and Retail AI Share X LinkedIn E-mail Copy link Share X LinkedIn E-mail Copy link Download Agentic commerce, where autonomous AI agents independently select and purchase products on behalf of consumers, is forcing retailers to rethink their role. As AI takes over more of the buying process, physical stores are evolving from a sales channel into a destination for advice, service, experience, and omnichannel support Katinka Jongkind A Galaxy General AI robot retail store, Shanghai. Retail is entering a new phase of digital disruption, where agentic commerce could fundamentally change how consumers shop Agentic commerce fundamentally changes the buying process Retail is entering a new phase of digital disruption.
E-commerce has already transformed where consumers shop; agentic commerce is set to change how they shop. Autonomous artificial intelligence agents can search, compare and purchase products on behalf of consumers, shifting shopping from an active task to an increasingly automated background process. Agentic commerce goes beyond AI assistants and comparison websites Key differences between agentic commerce, AI assistants (such as ChatGPT and Gemini) and comparison websites Source: ING Research "> Source: ING Research Most relevant for routine purchases, less so for luxury goods Agentic commerce is not equally suited to every product category.
Its greatest potential lies in repeat and replacement purchases, as well as functional, standardised products. In these categories, AI agents can efficiently compare prices, specifications, and availability, while using consumption patterns to determine the optimal time to buy. Adoption is likely to be slower for luxury, trend-sensitive and experience-driven products.
In these categories, inspiration, emotion, and personal preference remain central to the purchasing decision, making consumers less inclined to relinquish control. AI agents can assist by pre-selecting products that match predefined preferences, but final purchasing decisions are likely to remain firmly in the hands of the consumer. Consumer electronics and personal care lead the adoption of agentic commerce Within non-food retail, agentic commerce appears particularly well suited to consumer electronics and personal care.
Both sectors are characterised by a high degree of product standardisation and a large share of repeat purchases, making them ideal candidates for AI-driven purchasing. In consumer electronics, purchasing decisions are often based on objective and easily comparable criteria, including price, technical specifications, and availability. This makes products such as televisions, laptops and other electronic devices well suited to automation, allowing AI agents to manage much of the purchasing journey from product selection to the final transaction.
The potential is equally significant in personal care, where many purchases are routine and repetitive. Products such as diapers, shampoo, and toothpaste can be reordered automatically based on usage patterns and preset preferences. By reducing the effort required to order everyday products, AI agents could further accelerate the shift towards online shopping.
Agentic commerce is best suited to consumer electronics and personal care products Product categories most and least suited to agentic commerce Source: ING Research "> Source: ING Research Personal taste limits AI agents' potential in fashion and living For now, fashion and home furnishings appear less suited to fully autonomous agentic commerce. Purchasing decisions in these categories are often shaped by inspiration, aesthetics, personal taste, and broader lifestyle considerations, making consumers less willing to hand over control to an AI agent. While AI can support product discovery and narrow down options based on predefined preferences, the final purchase decision will typically remain with the consumer.
In these segments, AI is therefore more likely to serve as a decision-support tool than as a fully autonomous purchasing agent. Chinese platforms lead adoption Agentic commerce creates the greatest opportunities for large global marketplaces and technology companies. These players combine vast product assortments with the scale, data and infrastructure needed to deploy AI agents effectively.
Their integrated ecosystems, including payment solutions, logistics networks, and data platforms, enable them to integrate AI agents into their business operations faster and more effectively. Chinese platforms such as Alibaba and JD.com currently lead the way, with AI agents already deeply integrated into their respective platform networks. These agents can not only provide product recommendations but also complete purchases autonomously.
Meanwhile, US technology companies including Amazon, Google, Meta, OpenAI and Anthropic are investing heavily in the development and rollout of AI agents, intensifying the global race for leadership in this emerging market. European players are structurally lagging behind their American and Chinese competitors. Stricter regulation and less integrated platform networks limit their ability to bundle scale, data and services.
While Zalando is among Europe’s frontrunners in AI-powered retail, its applications are currently focused on product discovery, comparison, and selection. Fully autonomous purchasing has yet to become a reality. Technology platforms are becoming the gatekeepers of commerce AI agents are rarely neutral actors.
Embedded within commercial ecosystems, they are inherently linked to the interests of platforms, marketplaces and technology companies. Although agentic commerce promises greater convenience and time savings for consumers, it is also likely to strengthen the dominance of large tech firms and marketplaces. These parties are becoming increasingly influential in shaping product choices and controlling consumer access.
Physical stores are not disappearing, but their role is changing How will the role of the physical store evolve in an AI-driven retail landscape? Source: ING Research "> Source: ING Research Redefining the store in the age of AI Agentic commerce is expected to accelerate the shift of sales to online channels, putting further pressure on the existence of physical stores. The extent of this impact will largely depend on consumers’ willingness to delegate purchasing decisions to AI agents.
Physical stores will nevertheless remain relevant for many non-food retailers. Their role is shifting from sales channel to destination for advice, service, inspiration, and support of online sales. Stores that continue to rely mainly on transactional sales risk losing relevance in an increasingly AI-driven retail landscape.
Retailers that successfully integrate their physical and digital channels, define clear roles for their stores, and deploy technology strategically can strengthen their position within the customer journey. In a future shaped by AI-driven commerce, their role will increasingly centre on adding value where automation cannot. Retail Consumption AI Content Disclaimer This publication has been prepared by ING solely for information purposes irrespective of a particular user's means, financial situation or investment objectives.
The information does not constitute investment recommendation, and nor is it investment, legal or tax advice or an offer or solicitation to purchase or sell any financial instrument. Read more Share X LinkedIn E-mail Copy link Share X LinkedIn E-mail Copy link Download Author Katinka Jongkind Senior Economist, Retail, Services and Leisure Katinka Jongkind is a sector economist covering Retail, Services and Leisure. She joined ING in 2012 after a career in financial journalism.
Katinka studied Economics at the University of Amsterdam. In this article Agentic commerce fundamentally changes the buying process Most relevant for routine purchases, less so for luxury goods Consumer electronics and personal care lead the adoption of agentic commerce Personal taste limits AI agents' potential in fashion and living Chinese platforms lead adoption Technology platforms are becoming the gatekeepers of commerce Redefining the store in the age of AI
Sources & References
How we cover this story