
For fifteen years, the biggest brands won product discovery by default.
They had the ad budgets, the shelf space, and the enterprise software to sit on page one of Google and Amazon while everyone else fought for scraps on page seven. If you are a Shopify brand doing $500K to $10M, you have felt this ceiling. Here’s the counterintuitive part: the rise of AI shopping and agentic commerce might be the best thing that ever happened to you, because an AI agent does not care about page seven. It just wants the right product.
Morten Poulsen has been watching this shift longer than almost anyone. He spent nearly four years inside Google working on programmatic buying, then founded Plytix, the product information management (PIM) platform serving brands across 29 markets and putting enterprise-grade product content tools within reach of small and mid-market merchants. His argument is simple and a little provocative: AI agents have no loyalty to big brands. They reward whoever has the most accurate, structured, and complete product content, and that finally levels a field that has been tilted for a very long time.
In this conversation, Morten breaks down content-led growth, his umbrella framework for winning discovery whether it happens on Google, on Amazon, or inside a ChatGPT or Claude conversation. You’ll get the practical playbook for building a single source of truth for your product data, why localization is far more than translation, and how to get agentic-ready today without ripping up everything you already know about SEO. Whether you’re running a single Shopify store or selling across a dozen marketplaces and markets, this is your roadmap.
Let’s dive in. 👇
✅ Why AI agents may favor the little guy. How product discovery through ChatGPT and Claude can break the page-one monopoly enterprise brands have enjoyed for over a decade—and what that shift means for brands doing $500K to $10M.
✅ What “content-led growth” actually means. Morten explains his term in plain language and shows why it sits over SEO, product listing optimization, and conversion, instead of replacing them with yet another shiny acronym.
✅ The hidden cost of scattered product data. What really happens to your margins and your sanity when product content lives in spreadsheets, browser tabs, and Dropbox folders—and why a single source of truth stops being optional as you add channels and markets.
✅ Why localization is not just translation. The units, currencies, and channel rules that quietly break when you expand into new markets, plus the “content tree” problem that turns one product update into a maintenance nightmare across every branch.
✅ How to get agentic-ready without abandoning SEO. A crawl–walk–run approach to readying your data for AI discovery, and what emerging Shopify data on agentic conversions signals about where buying is actually headed.
✅ Why specialized AI can beat raw ChatGPT on product content. It’s not about having a bigger model. Morten reveals the “harness” that lets a focused system consistently out-produce a frontier chatbot on your catalog specifically.
MAI Fulfillment is a 3PL partner for brands that have outgrown basic fulfillment but do not want to get lost inside a massive, inflexible network. We specialize in ecommerce fulfillment, B2B distribution, retail-ready shipping, inventory management, returns, and value-added services that help brands operate with more control and confidence. Our value is in the combination of technology, execution, and real partnership.
We give you real-time visibility, reliable workflows, and operational support that helps prevent small fulfillment issues from becoming expensive customer problems. Whether you are managing online orders, wholesale growth, seasonal spikes, or multi-channel complexity, MAI provides the structure to scale without sacrificing accuracy or service quality. We are practical, responsive, and built around your growth. That is what makes us different from 3PLs that treat every brand the same.
For most of ecommerce history, the game was rigged. Morten Poulsen saw it from the inside at Google, where the running joke was that the best place to hide a dead body is page two of search results, because nobody ever goes there. With roughly 90% of purchases happening on page one of Google and Amazon, the brands that owned that real estate owned the sale, while everyone else stayed invisible. Enterprise retailers grew six times faster than small brands not just because of scale, but because they could afford software the little guys could never touch.
Morten explains why AI is quietly dismantling that advantage. When you ask Claude or ChatGPT to find a tent for your next camping trip, the agent does not stop at the same four brands you always see. It reads product specs and use cases and surfaces the genuinely best match, even if that brand never cracked page one. That is the heart of what Morten calls content-led growth: structuring your product content so it wins discovery everywhere, whether that is a marketplace filter, an organic search result, or a conversation inside an AI app.
Most merchants get stuck long before that. Product content is scattered across spreadsheets, the Shopify admin, an Amazon seller account, and someone’s desktop. Ask which of four prices is correct and nobody can answer. Multiply that by every market, language, currency, and channel you sell in, and you get a tree of content branches that all have to stay accurate every time you change a single attribute. This is the problem Plytix, Morten’s PIM platform, was built to solve: one clean source of truth that feeds every channel and stays in sync.
You’ll hear the real story on getting agentic-ready without torching your SEO, why Shopify is already logging agentic-referenced conversions, and how a specialized system can out-produce raw ChatGPT on your catalog. Not because it uses a bigger model, but because it carries the full context of your products, your brand guidelines, and a team constantly training on channel best practices. Morten’s advice is refreshingly un-hyped: build the foundation first, adopt with a crawl-walk-run mindset, and make discoverability a top-three priority now
This isn’t a sales pitch. It’s a blueprint for making your products the ones AI chooses to recommend.
👉 AI discovery rewards accuracy, not ad budgets. Agentic systems do not carry brand loyalty; they surface whichever product has the most complete, structured, and accurate information for the query. For smaller brands, that is the first genuinely level playing field in product discovery in over a decade.
👉 A tactic is not a strategy. The temptation is to abandon SEO the moment a new three-letter acronym appears, but that’s like closing your physical stores the day ecommerce arrived. Content-led growth treats SEO, marketplace optimization, and AI discoverability as tactics serving one strategy, not competing religions.
👉 You cannot optimize what you have not centralized. Without a single source of truth, every channel becomes its own silo with its own version of the product, and nobody can say which price or description is correct. The foundation has to come first, because the AI and automations that enrich content only work when they sit on top of clean, structured data.
👉 Localization is a maintenance problem, not a translation task. Expanding into new markets means adapting currencies, units of measure, and channel rules, then keeping all of it accurate when your core catalog changes. Automation rules can get you part of the way, but they break at scale without an architecture designed to absorb constant change.
👉 Discoverability and transaction are separate problems. Where the purchase actually happens—whether in-app or through a link out—is still unsettled across Google, Apple, and the marketplaces. What is already happening is buyers researching and shortlisting inside AI systems, so getting found during that research phase is the urgent move.
👉 Being early to AI visibility is the opportunity; being late is just catching up. Ranking number one on Google as a new brand is nearly impossible today, but agentic discovery is still the wild west, much like the early days of Google AdWords. The brands that structure their data now can win visibility before the rest of the market floods in.
Morten Poulsen
Founder & CEO, Plytix
Morten Poulsen is the founder and CEO of Plytix, the product information management platform serving brands across 29 markets and one of the most widely used PIMs among small and mid-market merchants worldwide. Before Plytix, he spent nearly four years at Google, starting as an account manager and moving into the New Products division, where he ran programmatic buying for the Nordics. He then co-founded a marketing agency specializing in programmatic and Google Ads API work before launching Plytix in 2015 alongside co-founder Alejandro López.
What makes Morten’s perspective valuable is that he sits at the intersection of two worlds most operators only see one side of: the technical guts of how product data flows across channels, and the strategic question of how AI is reshaping product discovery. He has built his company on a single conviction: powerful technology should not be locked behind enterprise pricing. It shows in how he talks about leveling the field for smaller brands. So when he predicts where agentic commerce is headed, it’s worth paying attention, because his team spends all day figuring out what actually works.
Featured in This Episode:
Over 9 seasons, I’ve been incredibly fortunate to chat with some of the brightest founders building amazing Shopify brands, as well as the partners shaping the app and marketing ecosystem. Every conversation has taught me something new, and I’m grateful for the chance to learn alongside you.
What matters most is that this podcast helps you solve real challenges and discover new ways to grow. Your support, feedback, and stories have made this journey truly special. Thanks for tuning in, sharing your wins and losses, and being part of the eCommerce Fastlane community.
Stay Connected: Leave an Honest Rating/Review on Apple Podcasts or Spotify. Follow & Subscribe on YouTube for new episodes.
Steve Hutt:
Welcome back to eCommerce Fastlane. I’m your host, Steve. Today’s conversation is all about content-led growth and why your product content, in my opinion, is deciding whether AI systems recommend your product or pass you by. If you’re selling in more than one place—which most of us are—your product content is probably scattered across a lot of different areas right now. Think about spreadsheets, browser tabs, Dropbox folders, or wherever else it lives.My guest has actually solved that problem. His name is Morten, and he’s the founder of a company called Plytix—that’s P L Y T I X.
Steve Hutt:
He spent years inside Google and learned a lot about what it means to build product content software, long before most people thought about where commerce was headed. Plytix now serves brands in, I think, 29 markets. They’re connected everywhere.By the end of today’s episode, I’m hoping you’ll understand what content-led growth means in plain language, what Morten believes is happening right now and in the very near future with AI agents, and how they handle smaller, edge brands that don’t have large ad budgets. How do those brands get discovered? How do agents buy from other agents?
The whole idea is getting your product ready, and I think that’s one of the advantages of Plytix. So let’s welcome Morten to the show.
Morten Poulsen:
Thank you, Steve. You did the entire episode for me there.
Steve Hutt:
I know, right? Tiny bit of research before recording.
We had one of your partners on previously and talked a lot about Plytix as a platform. I’ll link that in the show notes. We went deep into the platform itself and its history. It was firmly entrenched in the enterprise market, but now you’re firmly in the mid-market and really helping Shopify businesses from a PIM perspective—product information management for those who don’t know—and also the DAM side, digital asset management, as a source of truth for content.
We’re pivoting now to what’s happening right now: this content-led growth concept you’ve coined and what’s happening with agentic commerce, because it’s an incredibly hot topic. Can you take me back a bit? You spent years inside Google, you built an agency, and now you’re making a bet on where the future is headed. Talk a little about that and what led us to chatting today.
Morten Poulsen:
First of all, when I started Plytix back in the day, I had no idea we’d end up where we are. This is not our first pivot, so to speak. To be honest, I didn’t even know what a PIM was when we started Plytix.
Fun story, Steve: the name Plytix actually comes from “product analytics.” The first concept of Plytix was to facilitate the exchange between brands and retailers when it came to product images and product information. We wanted to build a bridge between them so brands could see how end users interact with the products.
Obviously, we didn’t get a lot of traction on that, and that’s where we discovered the big need in the market for democratizing PIM, as I like to call it—basically making PIM available for the little guys.
You’re absolutely right. I’ve been in ecommerce for a long time. First Google, then running an agency. But it would be a lie to say I started Plytix with the idea of what it is today. That’s the fun of building software—you don’t always know where it will take you.
Steve Hutt:
Yeah, exactly. I love it.
You’ve made a case that enterprise retailers are growing a lot faster than smaller brands. From what I read on LinkedIn, you believe there’s a technology gap between what enterprise retailers have available to them versus what mid-market and early-stage brands have.
Can you talk about that gap? Is it closing or widening right now—enterprise versus mid-market and early-stage brands?
Morten Poulsen:
It depends on how you look at it. I was once quoted on some insights and stats showing big retail brands growing six times faster than the little guys. They occupy all the prime real estate on marketplaces and search engines, and it’s very hard for smaller brands to break through that ceiling.
There are many reasons why big brands have leverage. One is economy of scale in purchasing power. Another is software. They had access to very powerful enterprise software that gave them an unfair competitive advantage.
We felt that economy of scale is fair enough, but powerful technology shouldn’t only be accessible to big enterprise brands. Our whole notion since we started Plytix was to democratize technology. There’s no difference between a small brand and a big brand in terms of needing access to technology—only in how much they can pay.
Technology costs the same to build whether you build it for small businesses or big ones. Making it super expensive just because software companies can maximize profit was something we never liked.
As a byproduct of that philosophy, we’ve become known as the PIM for SMBs—the PIM for the little guys. That’s not entirely true; we have many enterprise brands. What we’ve always said is we want to democratize technology so everybody, big or small, can access it without discrimination.
Steve Hutt:
What’s interesting is I get a lot of DMs from founders on social. Many of them are worried that AI shopping favors larger brands right now because they have economies of scale or access to very expensive, enterprise-level technology that only large players can afford.
From what I understand, you argue something quite different. You feel these AI agents actually don’t have a preference for who they recommend. It comes down to your software leveling the playing field for mid-market and early-stage brands. Enterprise had that luxury, and Plytix helps close that gap.
Talk about the agentic landscape and how you’re readying the source of truth and the data so it can be discovered in agentic systems—for purchasing, research, citations, and that sort of thing.
Morten Poulsen:
That’s a very good topic and very close to my heart. It’s a hot topic. I did say that I believe the best thing that could happen for small and medium-sized brands was the rise of agentic commerce.
That doesn’t mean the overall performance gap is closing in every area, but it does in one particular area: product discovery.
For a long time, we had this joke at Google that the best place to hide a dead body is page two of Google, because nobody goes there. If 90% of products are bought from page one on Amazon and page one on Google, whoever occupies page one gets the sale.
As a consumer, I don’t want to go to page seven either, but I’m also annoyed that every time I want new shoes, I’m exposed to the same four brands.
That’s what AI is really democratizing. AI doesn’t care about going to page seven. It cares about finding the best product.
So when I start a conversation with Claude or ChatGPT about needing a new tent for camping, I can have a conversation about everything I need. It will, to a greater extent, find the right product for me than I can by myself. That might be the perfect product out there. Back in the day, I would never get exposed to it because I didn’t know how to put together the right keywords.
The brand couldn’t bid on all the generic keywords, and it all came down to one thing: occupy Google, occupy the major marketplaces, and you’re set. That made it very hard for the little guys to get eyeballs.
I believe that’s being democratized with AI because it’s a different way of discovering products.
Steve Hutt:
Yeah. For years, a lot of listeners have spent time on SEO—search engine optimization—doing whatever it takes to be on the first page of Google. Over the years, that’s been the highest ROI.
The energy goes into content production, link building, and improving domain authority. That’s incredibly important. No wonder Reddit shows up so much. There’s a relationship between OpenAI and Reddit, and a lot of those results are displayed accordingly. There’s a baked-in trust because Reddit has high domain authority and strict spam policies, and people publish honest conversations about products.
No wonder those Reddit threads and listicles are being ingested by AI systems and displayed during research.
So, what’s your thought about people who’ve put all this energy into SEO? Does it bleed into these agentic systems? How is Plytix thinking about content-led growth in relation to that? Is it similar or different? Is there still benefit to having high SEO value in these agentic systems?
I’d love your mindset, because you’re definitely on the front lines.
Morten Poulsen:
Absolutely. This is a good time to introduce the concept of content-led growth.
We made up the term, so I can define it however we like. If you think about it, we all first got familiar with SEO—optimizing for search engines. Then a lot of product searches moved onto marketplaces, and another topic came up: PLO, product listing optimization.
Then you also had to optimize for conversion, which brought in CRO. So you have all these different acronyms.
For us, content-led growth is just the umbrella term for all of those things.
Do you need to completely abandon SEO because now there’s agentic commerce? Absolutely not. That would be like a physical retailer shutting down their stores the day ecommerce arrived.
If you go back to when ecommerce first came about, everybody thought physical shopping would die completely. It didn’t. The total retail market grew.
What we’re preaching is: do not abandon everything and go all in on the next three-letter acronym.
Think about your strategy. Each strategy is made up of different tactics. A tactic is not a strategy.
You can see some brands that are only on Amazon. A couple of years ago, 50% of new retail brands were born directly on marketplaces without even having a website.
So, whatever your go-to-market strategy is and whatever three-letter acronym tactics support that strategy, you shouldn’t abandon them. But you should think hard about how product discovery is changing.
We’re now having conversations. Whether you shop on marketplaces, through ChatGPT, or directly on Google, you want maximum exposure for your products.
I wouldn’t say “forget everything you know about SEO because it doesn’t work anymore.” It does work. But it’s also true that a lot of product discovery has now moved into LLMs.
Steve Hutt:
Right.
Another acronym I’ll throw in is “AI visibility.” I see many agencies pivoting some of their SEO marketing toward AI visibility. They’re running quick visibility audits with generic queries to see where brands show up.
Instead of just keywords, they’re testing long-tail prompts like, “What are the best shoes under $200 for someone with a high arch?” That might be your niche, and you want to show up there. That’s knowing your market.
We should talk about that, because one advantage of what you’re doing with content-led growth is thinking about the customer—how these systems, marketplaces, and the Shopify store work together organically.
Let’s talk about that. Many brands are truly multichannel retailers. They’ve got retail, their online store, an Amazon shop, maybe Etsy. There needs to be a source of truth. That’s where Plytix comes in.
Knowing that’s the source of truth, what’s the thought process around getting that data organized so it’s agentic-friendly and marketplace-friendly?
That’s the challenge for many people listening right now. They’re doing decently on their Shopify store and Amazon with built-in audiences, but they don’t feel like they’re building a competitive moat. They have too many separate systems and data pools that don’t talk to each other. They don’t know what’s working where.
If they modify product listings on Amazon, how does that bleed into SEO or how they’re being cited in large language models?
To summarize: there are many multichannel retailers with separate systems. What do you recommend they do today so they’re agentic-friendly and continue to grow their organic reach and marketplace presence?
Morten Poulsen:
That is a big question—the multimillion-dollar question, I guess.
First, let’s look at how you get ready for agentic commerce. And let’s talk about what agentic commerce actually is. Many people believe it’s an AI chat interface that does the shopping for you. That might happen, but I also see a future where it’s less about the transaction and more about discoverability.
If we’re heading toward a world where a “Shopping GPT” does all the shopping for you, I don’t think certain big players are going to be happy about that. There will be pushback.
What we’re seeing is Alibaba already doing a great job with their own agentic commerce on top of their platform. I think Amazon will do the same.
For me, agentic shopping isn’t just about whether there’s a winner-takes-all chat interface handling every transaction. It’s about having that agentic behavior integrated into all your platforms. That’s my prediction.
But even if we do end up with a single shopping chat interface, the fundamentals of how you optimize product content are the same. You need to think about how people interact with these chats to discover products.
If you want sneakers under $200, you could filter for that on marketplaces or listicles—or you could ask directly in a chat. The same way you structure data so it appears in filters is the logic you use to make it digestible for AI.
The problem isn’t only knowing what works on each channel and finding the time to optimize. It’s realizing that if you start with base content, then add Amazon and start tweaking, you create a new branch. For every channel and market, you create multiple branches.
Maintaining the accuracy of all that information is hard. You might update a product—maybe the price or colors change—and you have to remember all variations of that core content across all those branches.
It’s not just about knowing what works on Amazon and making that branch once. And by the way, that’s forever changing. It’s also about keeping all those branches accurate whenever your core catalog changes.
That becomes very difficult. Many people in ecommerce wish they could go back to the old days when it was just about SEO. I don’t think we’re going back. On top of everything we have, we now have to think about enriching content with FAQs, use cases, and more attributes so it’s easier for AI, whether a popular model or one integrated directly into Amazon.
It all has to be digestible for these systems. That’s a different ballgame. But you don’t need to throw the baby out with the bathwater and abandon SEO tactics.
Steve Hutt:
Yeah.
When we think about a brand wanting to expand into a new channel, that’s where Plytix fits. You’re the source of truth. Often the primary connection is through Shopify, because that’s the starting point for many brands. They build a Shopify store, then expand into other marketplaces.
There’s another layer when we think about content-led growth: what happens when we expand into other countries? That’s where language and localized currencies come in. Historically, cross-border operations like that have been mostly workable for enterprise companies, because they had access to complex cross-border systems.
Shopify has helped a lot with Shopify Markets and their connection to Global-e, which is useful for American brands connecting to Canada and globally from a duty and tax perspective.
But there’s still the question of how you show your site in other markets and languages. What’s the source of truth for that product data? You still want to show up in agentic systems and in organic search in, say, Italian Google.
What’s your thought about all of that? Internationalization is complex. How does Plytix look at it? Do you think there’s an agentic or content-led system around internationalization that brands are working on right now?
Morten Poulsen:
Absolutely.
If you go back to the analogy: you have your core content, then you have market adaptations. For us, you have your core content, then what we call market adaptation—different languages and local versions. Then for each market, you look at the channels you’re on. For each market, you have further proliferation.
If you think about that tree, it grows into something very complex and difficult to manage. Shopify has made it easier with Markets to expand and sell, take payments, and so on.
But you still have to maintain and optimize product content. Localization is not just translation. It’s currencies, units—centimeters, ounces, liters—and more. Everything has to change.
Technology has been there to set rules and automations, but they can only take you so far. You still see automation breaking and needing very complicated rules in order to get somewhat automated output. And it doesn’t automatically change if something in the core changes.
I think the mix of automations and AI will make it much easier to manage all of that without losing your head in the process.
You mentioned we’re doing a pivot. I think we’re more evolving. Nobody needs a PIM that’s only a static repository of product content. You want that repository to also enrich product content, make it complete, and optimize it for different markets and channels.
Now, with AI, we can layer that on top so you can maximize product discoverability on every market and every channel. That’s what we’re excited about. We’re so focused on that part that it might look like we’re pivoting into becoming “AI for product content,” but it’s still built on the core premise that we’re a PIM.
I don’t want this to turn into a sales pitch, but I’m genuinely excited about what AI has done. Combined with automation, it means you don’t have to prioritize which channels or markets to be on. You don’t have to compromise and say, “I know best practices for Google Shopping, but who has the time to set up use cases and FAQs?”
Now you can do that in bulk if you have the right system and AI on top. And when I say the right system, I don’t mean asking ChatGPT to fill a spreadsheet and watching it get tired after 50 products. I’m talking about deploying powerful models on top of a very structured database of product content.
Steve Hutt:
You’re right. Claude and ChatGPT are getting better at persistent memory and context, and that’s important. But that’s one advantage of Plytix—it has all of that baked in. Plytix is the source of truth for your product data. That’s true; there’s no doubt. Even G2 shows you have a high rating. The market clearly loves Plytix.
Again, this isn’t a pitch. Today is more about education around where we’re headed. That’s why this concept of content-led growth and agentic systems matters. Are you readying yourself for where things are headed?
There are massive investments worldwide in server farms. There’s one here in Alberta—Meta just spent about $9 billion on a huge AI server farm in a neighboring province. Memory shortages, memory pricing—all of that is happening. Knowledge workers like me can clearly see we get more work done, better, with these tools.
That’s where Plytix fits: the future.
You’re suggesting that some listeners may be thinking, “I have a lot of disparate systems: spreadsheets, processes. Maybe my life could be better; maybe there’s a better way.”
It comes down to your concept: content-led growth, having a source of truth, and this crawl–walk–run approach. You don’t have to adopt the full system overnight. Start where your shortcomings are.
Maybe you’re not showing up when someone researches your product on Claude. Great—let’s talk about that. Your system lets you set up product data so it has FAQs and proper schema, and so product detail pages and collection pages are correct.
There’s a lot going on, and Plytix fits into that mix. It’s not just “Are you ready for a DAM? Are you ready for a PIM? Are you connecting third-party listing tools?” Yes, Plytix can do all of that. But the key is: where are you today in your business, and where are your shortcomings? There are ways to turn on parts of Plytix to make all this magic happen.
So I’m curious about your feedback: where do you start from where you are, and what’s the hand-holding or process to use the platform so you become agentic-friendly? The future is coming. There’s no doubt. I’m not sure about app-to-app buying, but discoverability is clearly happening now.
Where the link routes to for purchase is a separate conversation. Google, Apple, and others are all figuring it out. We don’t know where the payment gateway lands. But you still want to be discoverable outside of Google organic.
What’s your feedback on all of this?
Morten Poulsen:
I think the interesting part is we don’t know where it’s going to land or where transactions will actually happen. But does it matter?
What you want is for people to find your products. At that point, it doesn’t matter where they pay. That’s not the part of the agenda we’re solving. We want to make sure that if you sell products, you get in front of your customers.
Whether discoverability happens in a chatbot or directly in a marketplace doesn’t really matter, because the fundamentals of optimizing and structuring product content are the same.
You said something interesting: brands need to optimize their product detail pages and collection pages.
It used to be that we had a very siloed way of working. You had an SEO specialist with their own spreadsheet, an ecommerce manager who only worked in Shopify, and someone on Amazon only doing that. You had multiple tools, all with product information.
If you ask, “Can I get all the information in one place?” nobody knows where to find it. There might be information on the desktop of the Amazon person that nobody else has access to.
If you say, “There are four different prices here. Which is the right one?” they come up short.
Back in the day, it was already getting difficult around 2015 and beyond, when ecommerce proliferated and it was more about optimizing for marketplaces than just your web store. Things became hard when everyone worked in their own system.
That complexity is now accelerating with AI and the speed of change. At least back then you could patch things with a feed management tool and some rules. The output wasn’t always great, but it was automatic and somewhat optimized.
Now, with AI and the tools available, you can deploy automations and AI to create really strong content optimized for each channel and keep everything synchronized and accurate.
But you can only do that if you start with the foundation. You need core data. You need a database, whether that’s Plytix or something else. Before you set up rules and agents to optimize for SEO, Amazon, and localization to Spanish, it all starts with a strong foundation.
My recommendation is obviously to buy Plytix, but if not Plytix, really think about building that foundation first. Then layer all the tactics on top.
The cool thing is, with AI and automations, you don’t have to do the manual typing. You can set it all up programmatically. It’s a big upfront investment, but one I think everybody should consider. Once you have that architecture, the world can throw anything at you—“Now it’s FAQs,” “Now Amazon changed its algorithm; long titles aren’t important, now it’s FAQs”—and you can adapt very quickly.
You cannot do that without an architecture built on a strong foundation.
Steve Hutt:
Yeah, it’s amazing. Back to my comment about context-aware systems: having a system that understands you. Persistent memory is incredibly important.
When you spin up an instance of Plytix for your brand, you don’t have to use the whole platform. Crawl–walk–run is really the way brands should adopt any new technology where they see value.
Your AI content agent inside Plytix actually understands your business individually. It’s not generic. It’s ingesting your product data from Shopify through the associated app as a starting point. That’s a two-way sync. From there, it continues into other systems.
Localization, like we talked about, is a big part of it.
The nice thing about Plytix—again, not trying to pitch, just being honest—is that it’s incredibly affordable. If you add up the mindshare required to run all these separate systems to track what’s going on at Amazon, plus point solutions managing isolated product data silos, I’m suggesting there’s a better way.
With persistent memory and a source of truth, we come back to the PIM concept. The idea is: things are happening now. There’s a move toward being agentic-ready and discoverable in these systems because that side is growing.
Shopify even released research recently showing how brands are being discovered and what agentic-referenced conversions look like. They’re logging them all now. Roughly 3–5% of conversions are happening because of a touchpoint from an agentic system—ChatGPT, Claude, or others.
That’s interesting. Plytix’s flag in the sand is right there. There are roughly 5 million Shopify brands, maybe 20,000 on the Plus side—mid-market to enterprise. Many mid-market brands know they want to be discovered and they’re doing SEO, but they have too many systems and need a source of truth.
You’re thinking, “Where can we be so brands are agentic-ready today? How can we make them discoverable, with a source of truth inside our platform?”
That’s exciting for where we are now and where we’re headed.
Morten Poulsen:
Absolutely. And I’m not going to lie: we have a whole team dedicated to figuring out what’s happening in the market—specialized in knowing what works on Google Shopping and what works now for agentic recommendations.
Our own internal resources show that in certain verticals, that 5% is very low. For products you don’t buy from brands you know every month—like T-shirts or washing powder—but things you buy occasionally, we actually see the agentic portion much higher.
That was a sidebar.
The interesting part, coming back on track, is: if you think about agentic commerce and how to get ready architecturally, consider this…
Steve Hutt:
Yeah, no worries.
Morten Poulsen:
Right, I remember where I was heading.
You asked: what is the big unfair competitive advantage we offer at Plytix, and why is it so important to have everything in a central source of truth?
Steve Hutt:
Right.
Morten Poulsen:
Think about it this way. I often get the question when I boldly say Plytix AI will be better than ChatGPT or even Claude at producing product content. People ask, “How can you say that? These are massive companies.”
It’s true: when it comes to blueberry muffins, we’re not great at recipes. But we’re using their models—we use Claude and OpenAI’s models, and we toggle between which is best for any given purpose.
What they don’t have is our context. We have all the context about the products. We know how you typically write descriptions and think about your catalog. On top of that, we constantly train our engine specifically on best practices for different ecommerce channels.
Combined with using those frontier models, of course we’ll be better at enriching product content. Of course Plytix will be better at optimizing content for different platforms. That’s not even a bold statement.
We have the context of your entire catalog. If you upload your branding and content guidelines, plus the work of a dedicated ecommerce team updating and refining best practices, that’s what we offer: not only a PIM to store information, but a tool already trained on ecommerce.
All it needs is your guidelines and catalog. Combined, that should create the best AI out there for producing and enriching product content.
Steve Hutt:
I agree. Your company isn’t alone in using frontier models—OpenAI, Anthropic, Google Gemini, and fantastic open-source models.
Your competitive advantage, from my perspective, is the harness. You have access to data and a source of truth, and your harness on top of that is ongoing ecommerce best-practice research.
You also have major enterprise players as customers. You can anonymize data and see what “great” looks like for them. Maybe you see what Reebok or others are doing. That lets you hypothesize what success looks like.
That’s why this is iterative. Nobody’s fully figured it out. We talked about SEO—we don’t want to give it up. Google is scraping the top 10 and putting those into AI Overviews.
We know that’s part of their learning algorithm for Gemini. If you search in Gemini, you’re kind of getting that top 10-ish Google perspective.
But Claude and ChatGPT aren’t crawling Google. They’re doing their own thing. That’s the benefit of having Plytix and this agentic system with a source of truth: these systems are more likely to find you because you’re readying your data correctly for ingestion—beyond Google’s top 10 scraped results.
Morten Poulsen:
Absolutely. And you said something important: it becomes a beast that compounds on itself.
The more brands we onboard, the more insights we get. That’s used to self-learn.
But it doesn’t mean the throughput—the product content output—looks identical for everyone. That’s where you upload your own guidelines and brand rules.
We want to get there. But you can accomplish something similar without Plytix. You start with a database structure, then you need a team constantly updating and optimizing best practices, and you set up agents that can do all the tasks.
I’m not saying you absolutely need Plytix. But this is where our commitment to democratizing software comes in. If you don’t have the capacity to hire people for that entire engine, we’ve made sure you can get those benefits at a fair price.
Steve Hutt:
Yeah, 100%.
So, what do you believe the next steps are? This show has founders, marketers, and agency partners listening. Some agencies are doing done-for-you services, which is another conversation for another episode.
I think some agency partners are wondering, “Could I have a done-for-you service that becomes the source of truth for my ongoing retainer clients, where I’m doing marketing and helping them with AI discoverability and data readiness?”
That’s a separate conversation, but I’ll throw it out there: it’s probably easy to go to plytix.com, reach out to the sales team, and say, “Hey, I’m an agency with 50 accounts. I’m thinking about using Plytix as a source of truth for all my clients, as a value-add. I don’t even have to tell clients I’m using Plytix. The results are what matter.”
For those listening in the mid-market—which is where the Shopify app fits—what do you think their takeaway should be, and potential next steps?
People know they need to be discoverable during the research phase. That’s where we’re headed: research inside agentic systems. Where the transaction happens—whether in-app or via a link—is evolving. But discoverability in research is incredibly important. It’s not just scraping Google’s top 10; it’s other systems and processes, where your source of truth matters.
So I’ll pass the baton to you. What do you believe the next steps are for those who want to learn more about Plytix and this system and process you’ve built?
Morten Poulsen:
I try not to force my opinion or be too “salesy.” Every business is different and has different priorities.
But I do believe that whether agentic commerce ends up being the thing we imagine—full transaction capability—or it’s primarily about agentic discoverability, brands should make this at least a top-three priority.
There are agencies betting heavily on this, helping existing and new clients transition into another tactic—call it agentic commerce or agentic discoverability.
Without making too many predictions I might regret later, I do believe you should rethink discoverability. That’s not just the future—it’s happening now.
We’ve written white papers and best practices on optimizing product content for discoverability. We’re not the only ones. It’s still the wild west, and the dust has to settle before clear guidelines emerge. But we’re testing this with hundreds of thousands of products and different best practices. Many agencies we collaborate with do the same.
I’d say those content resources and best practices are a good place to start.
It’s still an opportunity. Good luck trying to rank number one on Google if you’re a new brand. But agentic discovery is like the early days of Google AdWords and search. If you invest now, there may be a period before the rest of the market catches up.
Soon it will be something you have to do just to keep up. Right now, it might be an opportunity to look ahead.
Steve Hutt:
What’s great is you have a lot of research. There’s great content on your website. I’ll send people to plytix.com. There are ebooks, guides, playbooks, and very pointed resources: “If you have this challenge, learn about it here.”
If you’re in the very early days of your business, start by building your business acumen. You need to think about AI in relation to discoverability and the future.
I agree—it’s happening now. People are definitely researching through AI systems before making product decisions. I used AI for my vacation planning and spent maybe 20 hours setting up a project and going through options.
Maybe I’m more on the leading edge of AI adoption than the average consumer, but people are using ChatGPT and Claude. Claude’s the number one app on the Apple App Store. People clearly use it and love it. It’s no longer just a coding tool; it’s a consumer product for research.
If you’re not showing up, that’s why new companies are popping up to track queries in AI systems. They’re checking whether you show up in discoverability. They’re not giving you the answer on how to fix it, but they show you where you stand.
“Best shoes under $200 with a high arch”—do you show up? You can put that query into these modified SEO tools. They don’t tell you how to fix it; they just tell you you don’t show up, or that you need to write another content piece. That’s not necessarily the answer.
Morten Poulsen:
Yeah, it’s like: you’re not showing up, but your competitors are.
Steve Hutt:
Exactly. So what’s the answer?
The answer is a source of truth through a system like Plytix.
The education is here. I’ll put all of this in the show notes. This has honestly been really educational.
Thank you for sharing. I’ve got pages of notes. I feel blessed to have the opportunity to chat with you because anybody watching or listening is on a learning journey right now. We don’t know the exact outcome, but it’s about the journey—both as consumers researching and making buying decisions, and as brands beyond Google organic listings.
Knowing all this research is happening and how you buy products, Plytix fits into the mix by readying your data and being a source of truth across marketplaces and your Shopify store.
This has been very educational for me, and I hope everyone watching or listening feels the same.
Once again, thank you, Morten, for coming on the show today.
Morten Poulsen:
Thank you very much, Steve. I really enjoyed this. Have a wonderful trip to Europe.
Steve Hutt:
Thank you so much. All right, have yourself a great day.
Morten Poulsen:
You too.
Steve Hutt:
Okay, bye-bye.
Well, that’s it for today’s episode. I’d like to thank you personally for being a loyal listener of eCommerce Fastlane. It’s my hope this podcast offers you a ton of value through growth strategies, tactics, and exclusive insider tips on the best Shopify apps and marketing platforms—always with my goal to help you build, manage, grow, and scale a successful, thriving company powered by Shopify.
Thanks for investing some time today and listening to the show. I’m proud and excited that you have a growth mindset and are a constant learner. I truly appreciate you and your entrepreneurial journey.
Enjoy the rest of the week and keep thriving with Shopify.