What 100+ retail leaders explored about unified commerce, intelligent discovery, and the role of human judgment in AI-enabled retail
The most effective AI is invisible by design. Customers should experience the benefit, not the machinery behind it.
That idea connected many of the conversations at RetailDELUXE NYC: Fall Edition, co-hosted by Allure Commerce and Shopify on October 14, 2026.
More than 100 retail leaders representing over 50 brands and organizations gathered at Shopify’s New York offices. The room included leaders from Macy’s, LVMH, Tory Burch, Fenty Beauty, Dolce & Gabbana, KHAITE, L’Oréal, Shiseido, Longchamp, Maje, Vestiaire Collective, Printemps New York, and many more.
Across two panels, Unified Commerce and AI in Retail, the discussion moved past broad predictions about technology. Instead, it focused on the operating decisions that determine whether an idea becomes useful: who owns the customer experience, what information can be trusted, where AI should assist, when a person should remain in control, and how a brand decides whether an investment is creating real value.
Unified commerce begins with the customer promise
Customers do not experience a brand as a collection of channels. They do not think in terms of ecommerce platforms, store systems, customer service workflows, inventory tools, or organizational ownership.
They experience one brand, and they expect that brand to recognize them and keep its promises wherever they choose to interact.
That expectation can expose the gaps inside a retail organization. A payment method accepted online may not be available in a store. An associate may not be able to change an order without cancelling and recreating it. A return may move through customer service, a warehouse, an ERP, and a payment system before a refund can be completed.
To the business, these are separate processes. To the customer, they are one experience.
The Unified Commerce panel explored what it takes to close that gap. The answer was not simply another platform or integration. Technology matters, but the harder work involves operating models, ownership, incentives, policies, and the information available to the employee serving the customer.
A connected experience requires teams to agree on more than data. They must agree on what happens when inventory is limited, which channel receives credit for a sale, who owns a return, how operational costs are attributed, and who can resolve disagreements when departmental targets compete with the wider customer relationship.
Unified commerce is therefore not only a technology decision. It is an organizational commitment to behave like the single brand the customer already believes they are dealing with.
Store associates should not carry the burden of disconnected systems
When systems and policies do not align, store associates and service teams are often left to explain the inconsistency.
They may have to tell a customer that information is unavailable, a payment cannot be accepted, an order must be recreated, or a service offered through one channel does not apply through another. The employee becomes the human face of an operational problem they did not create and may not have the authority to solve.
This is where better information, and eventually AI, can make a meaningful difference.
Instead of presenting associates with more dashboards, applications, and logins, technology should help them understand the next best action. It could prepare useful context before an appointment, identify relevant preferences, check inventory, reduce repetitive administration, or make important exceptions easier to recognize.
The objective is not to place more technology between the associate and the customer. It is to remove the operational friction that prevents the associate from being fully present.
AI should provide context before it takes action
The second panel examined how AI is moving from personal experimentation into real retail workflows.
The strongest opportunities were not necessarily the most visible ones. They were the applications that helped people prepare, understand, decide, and respond more effectively.
AI can organize fragmented information, identify patterns, prepare an associate for an appointment, support campaign decisions, reduce internal search, classify service requests, or help a team recognize an exception before it becomes a customer problem.
But assistance and autonomy are not the same thing.
When AI influences availability, price, payment, fulfilment, eligibility, customer communication, or the use of personal information, the standard must be higher. The underlying data must be dependable. Ownership must be clear. Decisions should be explainable, recoverable, and subject to appropriate human review.
A useful principle emerged across the discussions: allow AI to prepare and surface information, but automate consequential decisions only when the information is trusted, the process is governed, and the outcome can be corrected.
AI should strengthen human judgment. It should not quietly remove it from moments where discretion, empathy, consent, or accountability still matter.
Product discovery is moving from keywords to questions
Retail discovery is also undergoing a fundamental shift.
Traditional search asks customers to translate what they want into a small number of keywords. AI-enabled discovery allows them to describe a complete need, including the occasion, style, budget, preferences, constraints, and comparisons that matter to them.
This changes how brands earn attention.
Product information must now work for more than a website search bar. Intelligent systems need to understand what a product is, who it may be relevant to, how it differs, when it is available, and which claims can be supported.
That does not mean reducing the brand to structured data. Retailers must make their information clearer and more usable while preserving the creative identity, positioning, and point of view that make the brand distinctive.
The opportunity is not simply to appear in more AI-generated answers. It is to provide information that is accurate and meaningful enough for intelligent systems to represent the brand well.
Better foundations matter more than more AI
The two panels approached the future of retail from different directions, but several shared principles emerged.
Customers experience one brand
A seamless customer journey cannot be produced by disconnected teams optimizing only for their own channels. Shared outcomes, ownership, and accountability are as important as technical integration.
AI is only as dependable as the operation beneath it
If information is unreliable or a workflow is broken, adding AI can make the failure move faster. Brands must understand and improve the underlying process before automating it.
Human judgment remains part of the design
Retail depends on context, relationships, hospitality, creativity, and discretion. The purpose of AI should be to give people better information and more time for those responsibilities.
Consent must shape customer intelligence
The fact that information is available does not automatically mean it should be used in every interaction. Brands need clear boundaries around customer data, relevance, permission, and the situations that require human review.
Investment must be tied to measurable value
AI and unified commerce initiatives should begin with a specific journey, decision, or workflow. Brands need a baseline, an owner, a meaningful outcome, and a clear standard for deciding whether to continue investing.
The speakers behind the conversations
The Unified Commerce panel featured:
-
Ann Marie Ippoliti, digital transformation and AI advisor
-
Andrea Risaliti, former Regional CIO at Bulgari
-
Eduardo Frias, Director of Professional Services at Shopify
-
Renan Goupil, Luxury E-commerce Strategist at L’Oréal Luxe
The AI in Retail panel featured:
-
Kerry O’Donnell, Senior Director of AI at Interparfums
-
Colleen Coulter, Global Ad Solutions Lead at OpenAI
-
Kapil Dabi, Market Lead at Google
-
Lauren Fiorentino, Partner Lead at Google
Both conversations were moderated by Ganesh Patil, CTO at Allure Commerce.
Their combined perspectives brought the discussion closer to the realities of retail: customer expectations, store operations, marketing, organizational adoption, data quality, responsible automation, and measurable business value.
The opportunity is to make complexity disappear
Retail does not need more technology for technology’s sake. It needs systems and intelligence that help brands operate more coherently and serve customers more naturally.
The next era of retail will not be shaped by who deploys the most AI. It will be shaped by who uses it most thoughtfully to remove friction, support better decisions, and strengthen the human relationships that technology cannot replace.
When that happens, customers may never notice the AI at all.
They will simply experience a brand that understands what they need and is prepared to deliver it.