Retailgentic: Agentic Commerce meets Retail and Brands

This week on the retailgentic podcast, we have an experiment for you. NRF concluded Tuesday Jan 13th and Scot's good friend and podcast partner over at the Jason and Scot show, Jason Goldberg (Chief Digital Commerce Officer at Publicis) was invited to ‘tie the bow’ on the conference in the last keynote: Commerce Disrupted: Rise of the AI Native Consumer.This episode is the audio from that presentation on 1/13/26. Jason covers:Agentic commerce, defined clearly: not “AI efficiencies,” but new consumer behaviors (and the full “iceberg” of influence, not just robot-takes-card transactions).Is it hype or historic? Jason frames the debate (skeptics vs believers) and uses live audience polling to set the stakes.A disruption of discovery: the decline of “saw in store,” then “search,” then “social”, and now the robot becomes the front door.The cautionary tale (Tudor Ice): why incumbents often miss the next business model, even when they dominated the last one.Retail growth is concentrated: a few giants capture a huge share of growth, while challengers (Shein/Temu/TikTok Shop) rewrite the playbook.AI shopping ramps fast: from research → to product tiles → to direct purchase inside AI experiences, plus retailer catalogs moving upstream.0-click commerce: the next leap isn’t fewer clicks, it’s no clicks (automation, replenishment, “never let me run out”).Onsite vs offsite agents: retailers’ on-site assistants (ex: Amazon) vs consumers using neutral offsite agents that choose the retailer for them.Winning the robots: why optimization shifts toward what models care about (criteria, sources, “GEO/AEO/ACO”), and why this changes constantly.Change management wins: Doug McMillon’s “How did you use AI to prepare?” as a culture lever—and why org adoption matters as much as tech.If AI changes the way people discover, it changes everything downstream. Brands, retailers, and platforms all have to learn to win the robot’s shelf.Notable timestamps 00:00 — The real disruption: new consumer behavior, not efficiency gains01:43 — Scott sets up the “special episode” + NRF context03:17 — The experiment: “put you in the audience” (keynote replay)06:29 — The big questions: how big is agentic commerce + who wins?08:07 — Skeptics vs believers: “hallucination” vs “overhyped” vs transformational11:05 — Defining agentic commerce + the “influence vs transaction” iceberg15:07 — Tudor Ice story: disruption, incumbents, and missing the next model19:14 — $5.3T retail + who captured the growth (concentration)22:16 — “Disruption of discovery”: saw-in-store → search → social25:04 — Viral demand shock example (whipped coffee)26:53 — Sunscreen example + TikTok → then “ask the robot”29:08 — “How fast this changed” (recent launches + ecosystem momentum)30:19 — From “order” to “never let me run out” (automation framing)31:11 — Clicks: 22 → 4 → “one-click” → 0-click32:40 — Proof points: agent assistants and rising adoption33:49 — Offsite agents: buying without choosing a retailer first34:03 — In-store: pointing ChatGPT at the shelf (real-world agent use)36:29 — Scale math: small % of prompts can still be massive volume38:34 — Robots optimize differently + why brands need new processes/tools39:36 — GEO/AEO vs the bigger strategic layer (Agentic Commerce Optimization)40:33 — Doug McMillon’s culture lever: “How did you use AI to prepare?”42:12 — Final message: change management to avoid being the “ice harvesters”👉 Connect with Jason: https://www.linkedin.com/in/jasongoldberg/👉 Connect with Scot: https://www.linkedin.com/in/thescotwingo/👉 See the slides: https://open.substack.com/pub/scotwingorefibuy/p/jason-goldbergs-full-nrf-keynote🧠 Want to stay ahead in AI commerce? Subscribe and follow along:📰 Subscribe to the free Substack: retailgentic.ai 📺 Watch episodes on YouTube & Subscribe for updates: https://www.youtube.com/@Retailgentic

Today we have a special FLASH episode - no fancy intros, just the news as it's breaking at NRF.

Consumer behavior doesn’t change overnight, even when the technology is obvious. In this episode of Retailgentic, Scot sits down with Hamish Gunasekara to unpack why agentic commerce is inevitable, why it’s moving slower than the hype suggests, and why product data infrastructure, not checkout, is the real unlock.We cover:Why “fully autonomous shopping” is the wrong definition of agentic commerceWhy consumer habits change slower than technologyWhy global mega-catalogs don’t work (and never really have)How vertical AI shopping experiences can thrive alongside ChatGPTThe hidden trust and governance problem no one is talking aboutWhy product catalogs are the hardest unsolved problem in commerceIf you’re a retailer, brand, agent builder, or investor trying to understand what actually matters in the agentic commerce stack, this episode is foundational.Timestamps:02:39 Meet Hamish & the origin of Catalog04:55 From Afterpay to Cash App Commerce11:30 The Naya experiment & lessons learned14:24 How AI is changing discovery16:08 Why catalog data is broken18:30 Fundraising & early traction20:48 Who Catalog is built for23:20 Why global catalogs don’t work28:30 The agentic commerce stack33:40 Defining agentic commerce (correctly)37:30 Why vertical agents will win43:45 Trust, governance, and adoption44:51 Where to learn more👉 Connect with Hamish: https://www.linkedin.com/in/hamish-gunasekara/👉 Learn more about Catalog: https://getcatalog.ai/🧠 Want to stay ahead in AI commerce? Subscribe and follow along:📰 Subscribe to the free Substack: retailgentic.ai 📺 Watch episodes on YouTube & Subscribe for updates 🎧 Listen wherever you get your podcasts: Spotify, Apple, etc: https://retailgentic.transistor.fm

Matt Nichols is back, and a lot has changed since his first appearance on Retailgentic. Scot and Matt revisit the predictions he made six months ago, look at what came true far faster than expected, and dig into the infrastructure challenges that still stand between us and frictionless agentic transactions. From SKU mapping and API-first checkout to the economics of feeds and the future of retailer traffic, this episode is a full-stack view of where Agentic Commerce stands at the end of 2025, and what’s coming in 2026.If you want more of Matt’s backstory and Commerce Ventures’ thesis, listen to Episode 4! It pairs perfectly with this deep dive. In this episode, they cover: Why VC predictions used to take 5–10 years to play out… and now take 5–10 weeksWhat Commerce Ventures is hearing from both startups and their LP base (top retailers + PSPs)Why ACP isn’t “the solution”—and what infrastructure still needs to be builtThe “API vs. fake-it-til-you-make-it” debate in agentic transactionsHow retailers should think about feeds, reviews, and blocking scrapersWhy “AI magic on your own site” is step one for most brandsPredictions for 2026: traffic shifts, wallet battles, Apple’s eventual entry, and moreWhere Matt thinks the industry is overly optimistic, and where the real upside isIf this year taught us anything, it’s that the next six months will change more than the last six years. Stay tuned.Timestamps05:00 — Welcome back, Matt Nichols05:10 — Six months later, the world looks different05:35 — Commerce Ventures in 60 seconds06:30 — What changed in 2025?08:00 — Standards vs. infrastructure10:00 — Retailers’ three priorities right now12:40 — Static feeds → AI-centric feeds15:00 — Reviews, Q&A, and proprietary data17:30 — SKU-level identifiers20:00 — API-first checkout vs. agentic scraping22:30 — Fraud, agent authentication & whitelisting24:00 — Payments are further ahead than commerce infrastructure27:00 — What retailers are actually worried about30:00 — Google’s role in accelerating the shift32:00 — Vertical agents vs. horizontal giants34:00 — Predictions for 2026 & beyond38:00 — Final reflections👉 Connect with Matt: https://www.linkedin.com/in/matthewnichols/👉 Learn more about Commerce Ventures: https://commerce.vc/🧠 Want to stay ahead in AI commerce? Subscribe and follow along:📰 Subscribe to the free Substack: retailgentic.ai 📺 Watch episodes on YouTube & Subscribe for updates 🎧 Listen wherever you get your podcasts: Spotify, Apple, etc: https://retailgentic.transistor.fm

In this episode of Retailgentic, Scot sits down with Luca to unpack a research paper that blends academic rigor with real-world implications:“LLMs Reproduce Human Purchase Intent via Semantic Similarity licitation of Likert Ratings.”The paper, co-authored with Colgate-Palmolive researchers, explores whether AI can accurately simulate human reactions to product concepts, enough to replace or accelerate traditional consumer panels, which are slow, expensive, and hard to scale.This one goes deep, but in ways that any retail or AI leader should care about. Scot and Luca discuss:Colgate’s Challenge: How to test product concepts faster and at scale.Synthetic Consumers: AI models that react to products like human panels.Accuracy Breakthrough: Reaching ~73–74% agreement with real consumers.Fixing LLM Failure Modes: Why naive prompts don’t work, and what does.Bayesian Reasoning: Adding uncertainty so AI stops being confidently wrong.Smarter A/B Testing: Using AI to pre-screen ideas before running live experiments.Digital Clones: Future consumers earning money by sharing preference data safely.Simulated Populations: Matching real audiences for testing and predictions.AI isn’t just helping brands write copy or generate images, it’s beginning to think like their customers. If synthetic consumers continue to evolve at this pace, product development, A/B testing, and personalization may look completely different in just a few years.Timestamps: 03:00 — Luca’s background: Rocket Internet, HelloFresh, Lazada, Stitch Fix08:00 — How PyMC Labs was founded & why Bayesian modeling matters16:00 — Bayesian thinking explained in simple terms19:00 — High-stakes decisions & why probabilistic reasoning matters24:00 — Colgate’s challenge: testing product concepts at scale26:00 — How synthetic consumer panels work30:00 — Accuracy results: humans vs. AI (~73–74%)32:00 — Why naive LLM prompting fails (“mode collapse”)35:00 — How reasoning → scoring solves accuracy issues38:00 — Example: synthetic consumers evaluating PyMC’s own website redesign44:00 — How AI can pre-screen ideas for smarter A/B testing48:00 — Where AI cannot replace causal testing49:00 — Digital clones & monetizable consumer preferences52:00 — Future benchmarks, new LLMs & evaluation methods👉 Connect with Luca: https://www.linkedin.com/in/lfiaschi/👉 Learn more about PyMC Labs: https://www.pymc-labs.com👉 Check out the paper: https://arxiv.org/pdf/2510.08338🧠 Want to stay ahead in AI commerce? Subscribe and follow along:📰 Subscribe to the free Substack: retailgentic.ai 📺 Watch episodes on YouTube & Subscribe for updates 🎧 Listen wherever you get your podcasts: Spotify, Apple, etc: https://retailgentic.transistor.fm#Retailgentic #AgenticCommerce #SyntheticConsumers #AIShopping #ProductTesting #CPGInnovation #GenerativeAI #BayesianModeling #Ecommerce #FutureOfRetail #DataScience #LLMs #AIResearch