Introduction

Coresight Research is a research partner of Shoptalk Spring 2026, taking place March 24–26 at the Mandalay Bay resort in Las Vegas, Nevada. In this report, we present highlights from the morning sessions of Day 2, covering three themes.

• AI Applications for In-Store Physical Retail—Industry leaders discussed how brands are testing and scaling AI to improve in-store service, support associate workflows, manage inventory and deliver personalized experiences.
• Driving Costs Out of Fulfillment and Delivery—Speakers examined how to reduce hidden exception costs in the delivery process and the physical and technological requirements for scaling drone delivery.
• Technologies Transforming Marketing and Advertising—Panelists explored the impact of agentic platforms on product discovery, the growing importance of human-generated content for AI algorithms and practical frameworks for bringing AI into marketing workstreams.

Shoptalk Spring 2026, Day 2: Coresight Research Insights

1. AI Applications for In-Store Physical Retail

The clearest message from this session was that AI’s most immediate role in physical retail is not replacing human interaction but improving it, giving store associates the information and tools to deliver more seamless, informed and personal experiences. The Vitamin Shoppe and Tecovas both said their primary goal is better hospitality, with AI working behind the scenes rather than as a visible feature of the customer experience. At The Vitamin Shoppe, that takes the form of “Shop Advisor,” which surfaces educational content and product recommendations in-store, helping both customers and associates navigate complex health decisions. To build associate knowledge, the retailer uses a gamification platform called Exonofiy with an “Elevate” tool, which tracks where each associate is in their learning journey and teaches accordingly.

Associates are also encouraged to ask customers for their phone number at the start of the visit to pull up loyalty details, freeing up the rest of the interaction for direct engagement. Andrew Laudato, Chief Operating Officer at The Vitamin Shoppe, said, “We didn’t set out to use AI—we set out to provide a better customer experience.” Laudato added, “AI will take away the worst part of your job.”

On the data side, the company built its infrastructure from the ground up using a clean room and medallion architectures, working from the principle that incomplete data is better than wrong data. It also syndicates its content and product catalogs, including research papers, into large language models to show up in relevant search results. Beyond the store, The Vitamin Shoppe works with Uber Eats and DoorDash for delivery, with the view that the omnichannel customer is the best customer and should be able to shop wherever they prefer.

Tecovas has focused on the operational friction points that come with scale, particularly the challenge of managing 13 sizes in regular and wide footwear backstock. Its “Boot Runner” app lets associates request products from the backroom without leaving the customer, replacing the radio and cutting wait times to an average of 85 seconds. The engineering team built it in two days. Kevin Harwood, Chief Technology Officer at Tecovas, said, “The worst thing that can happen is the associate leaves the customer—we want them present every step of the way.”

To make sure Tecovas shows up in AI-driven discovery, the brand uses the UCP protocol with Google and Shopify and has adopted Algolia and Botify specifically for AEO and GEO content optimization. The company’s data shows that omnichannel customers have a higher lifetime value than single-channel shoppers, which has shaped a push to create experiences across channels. One example: when an online order is returned in-store due to a sizing issue, an AI tool checks the retail statistical area (MSA) to surface store-level inventory and hold the right size for the customer.

Tecovas also runs an internal AI agent that manages upcoming cultural and music event calendars across retail MSAs, scoring events by their likely impact on store performance. Harwood said he has seven to 10 agents running at any given time, and that this has changed how the engineering team works, with code now essentially throwaway. The pace of development has also changed how the company trains: Harwood noted that training takes five times longer than the actual software development, requiring headquarters to continuously retrain teams on how quickly software can now be built.

Despite all this, Tecovas still dedicates 20% of its floor space to experience rather than selling, with the open bar, blow torches, stamp personalization, hat customizations and boot shining all part of the brand. Harwood acknowledged that the company typically only identifies the customer in the final 10% of the journey, usually at the point of sale or during scheduled processes like boot shining or preparing whiskey for buy-online-pickup-in-store orders, because asking for identification earlier is awkward.

The most valuable AI applications in physical retail remove friction and lift human connection, rather than trying to automate it away. Both retailers showed that AI works best when it is embedded into workflows—inventory allocation, associate training, content delivery—where it produces measurable results like higher in-stock rates and better-informed staff. Tecovas’ nearly 10% revenue lift in test categories from AI inventory optimization illustrates the business impact of applying AI to core retail fundamentals. Both speakers pointed to a future where AI makes store experiences more localized and relationship-driven, potentially returning retail to something closer to its local roots.

2. Driving Costs Out of Fulfillment and Delivery

The real cost of delivery failure is hidden, and most retailers aren’t measuring it correctly. Salman Habib, Co-Founder & CEO at Burq, explained that retailers tend to merge the concepts of shipping and delivery, but from the customer’s perspective they are different, and a missed delivery leads to churn. Around 8% of first-time deliveries fail, with each late or failed delivery costing between $17 and $40 once reattempts, support tickets and customer churn are factored in. Habib recommended starting with a delivery promise that varies by fulfillment area, diversifying beyond a single delivery provider to reduce risk, and building AI agents for automated exception recovery such as auto-rerouting. He also made the case that while every company tracks cost per delivery, they should start tracking cost per failed delivery, because the customer who blames the brand and doesn’t return is the hardest cost to measure and the most damaging.

The panel also covered the rapidly growing drone delivery market. Heather Rivera, Chief Business Officer at Wing, said Wing grew five times in the last year, driven by a shift toward small-basket orders wanted quickly, a use case that traditional last-mile delivery, built for large shipments, was not designed for. The drone delivery market is projected to grow 2.5 times by 2030, potentially adding $8.3 billion in incremental sales and $2.4 billion in cost savings. With flight times under five minutes, Wing’s partners are already seeing gains in purchase frequency and customer satisfaction. Rivera was clear that drone delivery is for everyone but not for everything, and it won’t replace the heavy Sunday grocery shop, but it is well suited to a missing ingredient for dinner or an over-the-counter medication for an aging parent.

For retailers thinking about getting started, she pointed to three factors: deciding whether to run a standalone Wing app experience or integrate into the retailer’s own app; adapting the physical real estate footprint for drone charging infrastructure and ensuring the technology can handle dynamic neighborhood conditions like parked cars; and confirming that the delivery partner has enough FAA-approved aircraft to scale if consumer adoption is strong.

Data visibility and the product lifecycle rounded out the fulfillment discussion. Mo Afshar, Co-Founder & CEO at Pipe17, noted that retailers are often locked into logistics providers simply because switching takes six to 12 months to integrate a new one. Disparate inventory pools and disconnected systems produce stale data and canceled orders, costing retailers time, money and lost sales. Afshar put it plainly: retailers cannot optimize for cost if they cannot see. Pipe17 positions itself as the middle tier that connects brands and logistics providers. Chloe Songer, Co-Founder & CEO at SuperCircle, made the case that retailers need to stop thinking of fulfillment as the end of delivery. Returns are part of the first purchase flow, and the millions of units circulating with a brand’s name on them—in supply chains, in stores, everywhere—directly affect that brand.

3. Technologies Transforming Marketing and Advertising

This session focused on agentic platforms and what marketers need to do to stay relevant in them. Jacob Ross, CEO at PebblePost, opened with a straightforward data point: transaction data is the most impactful and predictive basis for understanding who to target and what they will do next, producing two to four times more incremental lift than offline data. Ross drew a direct parallel between AI platforms and search engines: if Google changed its algorithms, companies would feel it, and the same is true for ChatGPT and its peers. Relying on any single platform creates exposure; navigating this space is less a set-and-forget operation and more a capability that has to be continuously maintained. Ross said, “Marketers want to find proof points to build confidence in a world that feels like maximum hype.” He also offered a counterpoint to the agentic commerce discussion: humans still buy things, and there is real value in human-based marketing driven by actual transaction data and genuine acquisition.

James Cadwallader, Co-Founder and CEO at Profound, added that in the near future every company will care about how AI talks about their brand, and every marketer will use AI agents as a kind of “marketing engineer” to get work done faster.

Bringing AI into marketing workstreams without losing brand trust was the focus of Bruce Richards, Head of Industry Strategy and Marketing Retail & Consumer Goods at Adobe. Richards shared that 94% of retail and consumer goods marketing leaders are now expected to directly contribute to revenue, with profitable growth as their top priority. He pointed to 45% of consumers turning to AI tools in their buying journeys, and an Adobe-tracked 700% year-over-year increase in AI visit growth during holiday 2025 retail. The brand’s front door, Richards argued, is shifting from personalization to participation, which means building for both humans and agents.

He outlined three moves for marketers: shift from messaging to invitations, letting customers choose their own experience; move from one-off campaigns to living systems where AI learns from every interaction; and go from capacity to comfort, making interactions feel natural. Richards gave CMOs three questions to start with: what AI initiative could save 50% right away; what would make a measurable impact on customer experience without necessarily saving money; and what governance is needed to keep AI initiatives on-brand. Bruce Richards, Head of Industry Strategy and Marketing Retail & Consumer Goods at Adobe, said, “Marketers get it conceptually (AI), but struggle with the ‘how’—ecosystem education, across the org.”

AI’s rise has also made human-generated content more valuable, not less. Kristen Wiley, Founder & CEO at Statusphere, noted that AI overviews appear in over 60% of search results and 74% of Gen Z uses TikTok as a search engine, creating a new discovery flywheel of social search and AI. Where it once took seven touches to convert a customer, it now takes 15 to 20. Creators add a trust layer that AI cannot replicate, making human-generated content the primary fuel for generative AI, social SEO and traditional SEO. More context, including brand mentions, influencer posts and video reviews, means more visibility across LLMs and social search. Statusphere maps a brand’s discoverability footprint, identifies where human content is missing and brings in creators to fill those gaps. Wiley’s advice was direct: start briefing influencer marketing teams now.