The B2B Podcast Index
Index
All categories
MarketingSalesSaaSFinanceHROpsLeadershipCustomer SuccessAI & DataProductStartups & FoundersRevOpsEngineering & DevTools
MethodologySubmit
Best of:MarketingSalesSaaSFinanceHROpsLeadershipCustomer SuccessAI & DataProductStartups & FoundersRevOpsEngineering & DevTools
An independent project byFame
SearchBest episodesGuestsInsightsMethodologySubmit a podcast
Index/Finance/Voice of MPE
Voice of MPE artwork

Commerce is changing. What merchants expect? Podcast w/Shopify.

Voice of MPE · 2026-05-19 · 12 min

0:00--:--

Key moments - from our scoring

Substance score

58 / 100

Five dimensions, 20 points each

Insight Density12 / 20
Originality10 / 20
Guest Caliber14 / 20
Specificity & Evidence13 / 20
Conversational Craft9 / 20

The modern commerce stack, built over two decades around human interaction patterns, breaks when AI agents attempt transactions because it lacks the contextual and deterministic layer needed for reliable agent-merchant negotiation. Fabian Musik, head of solution engineering for Shopify EMEA, explains that existing infrastructure detects and blocks bot traffic - but now merchants need to welcome high-intent, high-net-worth AI agents. Shopify's answer spans three layers: the Universal Commerce Protocol (UCP), an open standard developed with Google and partners that lets agents and merchants communicate mathematically; Agentix storefronts, invisible interfaces optimized for AI interaction; and an open product catalog merchants can publish to once and sell across multiple AI channels. The data already shows 15x growth in AI-originated orders at Shopify merchants between January 2025 and early 2026. For payment executives and merchants, the takeaway is urgent: prepare now by ensuring clean, structured product data, because AI is only as effective as the data feeding it. Shopify's decision to open the catalog beyond its own merchant base reflects a philosophy that when commerce flows freely, the entire ecosystem wins.

Key takeaways

  • →Agentic commerce is growing 15x month-over-month at Shopify merchants (Jan-Feb 2025) and requires fundamentally different infrastructure than human-facing e-commerce.
  • →The Universal Commerce Protocol enables AI agents and merchants to negotiate terms in milliseconds by sharing intent and rules deterministically, replacing guesswork with structured data exchange.
  • →Merchants do not need to be Shopify customers to participate in agentic commerce through Shopify's catalog, which distributes products once to Google, ChatGPT, Microsoft Copilot, and other AI channels.
  • →Current commerce infrastructure was built to block bots; agentic commerce requires a 'green lane' that contextualizes transactions, applying loyalty programs, age verification, and personalization rules before payment processing.
  • →Clean, structured product data with correct meta fields and inventory levels is the foundation of successful AI-driven commerce, since agents perform as well as their training data.

Guests

Fabian Musik

Topics in this episode

ChatGPTGoogle GeminiMicrosoft CopilotAgentic commerceUniversal Commerce Protocol (UCP)Agentix storefrontsShopify catalogpoint of saleShopify Commerce operating systemstructured product data

Questions this episode answers

What is the Universal Commerce Protocol (UCP) and what problem does it solve?

UCP is an open standard developed by Shopify, Google, and partners that creates a standardized language for AI agents and merchants to communicate deterministically. It solves the problem of agents scraping human-facing websites and failing - instead, agents declare their intent and wallet details, merchants declare their rules (loyalty programs, age verification, shipping constraints), and transactions settle in milliseconds with the right context.

Do merchants need to use Shopify to sell through AI agents?

No. Shopify is opening its product catalog to any brand, not just Shopify merchants, allowing them to publish products once and sell across Google Gemini, Microsoft Copilot, ChatGPT, and other AI channels without switching platforms.

How much is agentic commerce actually growing in real merchant data?

Shopify sees a 15x increase in AI-originated orders at its merchant base between January 2025 and February 2026, indicating this is not a future scenario but an active, rapidly growing channel.

What is an Agentix storefront and why do merchants need one?

An Agentix storefront is an invisible, AI-optimized interface where merchants define product data and rules (shipping, inventory, compliance) in a standardized format that AI agents can reliably read and interact with, unlike human-facing storefronts.

What should payment executives prioritize to prepare for agentic commerce?

Ensure structured product data with correct meta fields and real-time inventory levels, since AI agents are only as effective as the data they access; also begin integrating with agentic commerce infrastructure now rather than waiting.

What our scoring noted

Our reviewer’s read on each dimension, with quotes from the episode.

Insight Density

12 / 20

The episode covers agentic commerce with some substantive technical points (UCP as translation layer, cross-border logic, inventory/metadata requirements), but relies heavily on Shopify's marketing narrative and lacks depth on implementation challenges, failure modes, or counterpoints. The content is informative but surface-level for operators who need to act on it.

The infrastructure we have in most cases is built to detect bot traffic and block it. And for the first time we also have good ones. We have bots that are high net worth customers, high intent, that want to spend money with you.
AI is only as good as the data it works with. So that's a key part, making sure that you have structured product data with the right meta fields, with right inventory levels, so AI is able to facilitate the purchase.

Originality

10 / 20

The discussion of agentic commerce and UCP represents genuine emerging territory, but the framing is entirely vendor-centric and promotional. There is no contrarian perspective, no discussion of risks or failures, and the 'good bots vs bad bots' framing is Shopify's standard positioning rather than independent analysis.

The problem I just described, where bots are scraping an Internet that isn't built for them, but that is built for humans. There's something missing there. What's missing is a translation layer.
commerce flows freely, everyone wins and so everyone can participate.

Guest Caliber

14 / 20

Fabian Musik is Head of Solution Engineering for EMEA at Shopify, a senior practitioner role with real operational responsibility and customer-facing expertise. However, he is speaking in a role that requires him to represent Shopify's interests, which limits the candor and independence typical of ideal guests.

Happy to be here.
we power millions of merchants and we do that across something we call the Commerce operating system.

Specificity & Evidence

13 / 20

The episode includes one concrete data point (15x increase in AI-originated orders Jan 2025-26) and named merchant examples (Gymshark, Glossier, Mattel). However, most technical claims lack supporting evidence, timelines are vague ('a couple of years'), and the UCP details are explained conceptually rather than with implementation specifics or adoption metrics.

we run merchants, local merchants from Europe like Huell, Gymshark, West Wing, Jagermeister, uh, global brands like Mattel Supreme Glossier.
from January 2025 to 26 of 15 times, 15 times more order that originated in AI.

Conversational Craft

9 / 20

The host asks reasonable setup questions and shows interest, but rarely pushes back, probe for problems, or challenge the narrative. Questions are softball-adjacent - mostly invitations for Musik to expand on Shopify's roadmap. There is no genuine tension, disagreement, or skeptical follow-up about adoption friction, merchant hesitation, or competitive risks.

What, um, actually happens when an AI agent, uh, tries to complete a purchase on behalf of a customer? Like, is there anything that breaks the day?
Why should I ask more?

Conversation analysis

Computed from the transcript - who did the talking, and the words that came up most.

Share of words spoken

  • Speaker B61%
  • Speaker A39%

Most-used words

shopify16merchants16commerce14agent11catalog8agentic7infrastructure7agents7across6customer6built6layer6build5experience5payment5data5

Episode notes

In this episode, we dive into the rise of AI-powered commerce and explore what happens when agents start shopping on behalf of consumers. From payment orchestration to merchant readiness, we unpack the infrastructure being built right now to support this new era. We also explore the UCP protocol launched with Google, designed to translate the “language of payments” between all parties involved, and how Shopify is building the orchestration layer that could connect countless AI agents across commerce ecosystems. These are the conversations shaping the future of payments and commerce. See you at MPE 2027 → #merchantpayments #AIcommerce #AgenticCommerce #payments #fintech Thanks for listening!

Full transcript

12 min

Transcribed and scored by The B2B Podcast Index.

Speaker A: Hi everybody. Uh, I'm Matteo Gamba. Uh, we are here at mp and uh, today we're going to be recording one more, uh, session of the voice of mp and uh, with us we have Fabian Musik from Shopify, uh, the head of uh, solution engineering, uh, for uh, the EMEA market.

Speaker B: That's right. Happy to be here. Hi Matteo.

Speaker A: Welcome. Uh, so yesterday we had you on stage, um, both at the panel and at the keynote. That was great to have you. And uh, I'm sure today you're going to tell us a little bit more about something that everybody is talking about. Uh, Agent E Commerce. Um, Shopify is really leading, uh, the race. There, uh, is a leading player. And uh, we know that Shopify powers millions, um, of merchants out there, like in 175 plus countries. So I'm curious to know a little bit more about your keynote yesterday and also why, uh, Shopify is present at the payments conference.

Speaker B: Right. So as you said, we power millions of merchants and we do that across something we call the Commerce operating system. So we build that operating system over the past two decades. And it's not just online and online checkouts. It spans across brick and mortar with point of sale. It spans across B2B. Um, it is social channels, it's pop up stores, and now it's also agentic commerce. And we run merchants, local merchants from Europe like Huell, Gymshark, West Wing, Jagermeister, uh, global brands like Mattel Supreme Glossier. And they use this operating system to bridge the gaps between all different channels. And that puts us in a unique position, that gives us a unique radar. We are at the intersection of merchant, buyer and bank. And that means we see the signal before the transaction even happens. And we don't just see payments volume, but we see the entire journey and we see actual behavior change.

Speaker A: So not just small merchants, also bigger ones are using Shopify.

Speaker B: All sizes, from entrepreneurs to global brands.

Speaker A: Nice. Um, so everyone at mp, but not only at mp, uh, lately is talking about this agentic economy. Agentic commerce. What, um, actually happens when an AI agent, uh, tries to complete a purchase on behalf of a customer? Like, is there anything that breaks the day? Is that a smooth experience? How can you tell us a little bit about this journey?

Speaker B: The infrastructure we built, the industry built over the past, uh, where the Internet, uh, was starting with, uh, the search bar, ah, that was the front door to commerce. And if an agent uses that type of infrastructure, tries to make a purchase, it will fail. There's so many traps there where we cannot solve commerce logic at the payments layer. So imagine we have things like cross border logic, uh, taxes, difficult discounts. All these things need to be solved before we hit the payment. Because the payment layer doesn't. The last layer, it treats uh, a loyal five year VIP customer the same as it does uh, for a new customer. And the cure here is, is to provide context. So what we need to do for agentic commerce is rethink the way we build the infrastructure. Rethink the approach. The infrastructure we have in most cases is built to detect bot traffic and block it.

Speaker A: The bad ones.

Speaker B: The bad ones. And for the first time we also have good ones. We have bots that are high net worth customers, high intent, that want to spend money with you. So our mission at Shopify is to build a green lane and to build a way, um, for our merchants to participate from that opportunity.

Speaker A: Nice. So a hyper contextualized and aware experience, uh, where the customer understands uh, more. Also thanks to the technology that brings that in front of them.

Speaker B: Exactly.

Speaker A: And uh, in connection to this, uh, Shopify introduced something new called ucp, like the Universal Commerce Protocol, together with Google and other key partners at the end of last year. Um, for someone that is hearing this for the first time, what problem does this new framework solve and how?

Speaker B: The problem I just described, where bots are scraping an Internet that isn't built for them, but that is built for humans. There's something missing there. What's missing is a translation layer. A standardized language that lets you. AI agents.

Speaker A: Mhm.

Speaker B: Talk to merchants in a, uh, deterministically, mathematically correct way, understand what these merchants offer. And that's the ucp. The UCP is as you said, an open standard, A language that we released that allows merchants and AI agents to communicate in a standardized format and solve all these problems and therefore make sure that we have high authorization rates, that we have successful transactions on this new channel.

Speaker A: Yeah. And that agent end up buying the right stuff.

Speaker B: Exactly.

Speaker A: That's always a concern. It's like what if my agent buys something else? But if you have the right context, if you have the right information, the right pricing, obviously that's much clearer for an agent. What to buy.

Speaker B: Exactly. In the past, agents were guessing, they were scraping the site, they were sometimes maybe making things up. And that guessing is replaced with negotiation. Um, in the UCP there's this negotiation that happens in milliseconds where agents declare their intent, merchants declare their rules. Could be something like uh, an agent saying I have a wallet, I have a shipping address. Trying to find a store where they can successfully complete that order and the merchant declares their rules. I have this loyalty program. I need age verification. And when that negotiation is successful, the entire transaction can be settled in milliseconds.

Speaker A: And you have context also about the customer itself. Uh, like if you know, if they're a loyal customer, uh, when they purchased the last time, their transaction history, you can bring some of that context in front of the agent, making the experience more live and uh, real time as well.

Speaker B: That's, that's so important because many protocols out there, they focus purely on the payment token. And we know, because we see the entire picture. We know commerce is complex and we need to abstract that complexity and make it easy for our merchants. And that's why UCP spends not just across payments, but across checkout experience and the shopping experience. Right.

Speaker A: Makes a lot of sense. So besides ucp, uh, also Shopify, you built Agentix, uh, storefronts and uh, opened up your catalog infrastructure. Uh, that's really an interesting move. Uh, does a merchant need to be on Shopify to participate in agent E commerce and what do they keep control of?

Speaker B: That's the big shift that we are opening up our catalog to merchants who are not on Shopify. But maybe to take a step back to explain what this is concept actually is, ucp, which we talked about, is sort of the infrastructure layer, the protocol behind it. Agentix storefronts are storefronts that are invisible to humans. They are optimized for these AI agents to interact with. So merchants have a standardized way where they can define their product data, certain rules like shipping rules and so on, and that is accessible for AI agents in a standardized way. And the catalog which you just mentioned, that's the catalog where merchants publish their products so they can be discovered in any AI channel. The beauty is they do it once and it can be discovered and sold on Google, gemini, Microsoft, Copilot, ChatGPT. So that's the catalog. And what we are doing now, which you just mentioned, is we are opening up that catalog to any brand, not just Shopify merchants. So if you are not on Shopify today with the Agentic plan, you will be able to publish your products into the Shopify catalog and sell to all these AI channels. And we do that because we think if commerce flows freely, everyone wins and so everyone can participate.

Speaker A: So you build infrastructure layer that allows to orchestrate across all the multitude potentially of uh, agents that are going to be out there. So I don't have to worry about that. That's great. Uh, you reduce the effort and give Me more sales channels? Why should I ask more? Um, but if we fast forward a couple of years, um, maybe 1 5. Um, AI moves very quickly so that ah, time frame could be, uh, different. Uh, but a human takes minutes to browse and fil and checkout. An AI agent does that in a few milliseconds. What should the payment executives and the people that are listening to the show, um, understand and think about to prepare for this change?

Speaker B: It's important to understand that this change is happening right now. It's not a future vision. It's something that is impacting businesses right now. We see if we look at our Shopify stores globally, millions of stores, and we look at the data, we see an uh, uptick of orders that originated in AI, uh, from January 2025 to 26 of 15 times, 15 times more order that originated in AI. So it's not something that is far away. It's happening already and it's just the beginning. No one has the crystal ball to know exactly what it's going to look like. But agentic commerce will be shaking the ground. And so the recommendation for payment executives, merchants out there is to act now and to start building the foundation for agentic commerce. And the key piece we just spoke about, the catalog, is making sure the product data is right. Because AI is only as good as the data it works with. So that's a key part, making sure that you have structured product data with the right meta fields, with right inventory levels, so AI is able to facilitate the purchase.

Speaker A: Perfect. So thanks for being here. Uh, I think we got an understanding that agent E commerce might not be visible one, uh, hundred percent today, but it's not something that we can pass on. Uh, we need to keep an eye on that and it might come faster than we expect. Uh, and, well, it's good to see that Shopify is building solutions that are not just for Shopify Merchant, but also for anybody else, uh, to really give space to the industry to embrace this, uh, new trend. So thanks for being here and good luck with the rest of the conference.

Speaker B: Thank you. Matteo.

Related episodes across the Index

Other episodes covering the same guests and topics, from across The B2B Podcast Index.

  • Is Your Business Invisible to AI Search? (And How to Fix It) ft. Ray YoungRevenue Science · on ChatGPT85 / 100
  • Episode 018: Season 2, the $75 Consult and the Frankenstein StackAI Tools for Practicing Lawyers · on ChatGPT84 / 100
  • SVP at Adobe Express: Why AI Content Doesn't Get Views | Govind BalakrishnanSilicon Valley Girl · on Microsoft Copilot82 / 100
  • PwC's Chief People Officer on Training 80,000 People for the AI Era With Human Skills at the CenterFuture Ready Leadership With Jacob Morgan · on ChatGPT81 / 100
  • The emerging world of machine customers - Interview with Katja ForbesPunk CX · on Agentic commerce81 / 100
  • S02E14 It's a Yes & conversation .. guiding your teams to use AI | Emilie Schmitz | Wired for WonderWired for Wonder · on ChatGPT80 / 100

More from Voice of MPE

All episodes →
  • Payments infrastructure 2.0 w/ payabl.55 / 100
  • Fraud 3.0: From risk mitigation to revenue enablement66 / 100
  • What happens to commerce and payments in a programmable world?71 / 100
  • Orchestration layer: Controlling complexity w/Payrails
  • How AI is shaping merchant payments in 2026 w/Mastercard
Explore the best B2B Finance podcasts →
All Voice of MPE episodes →