Cover: SPECIAL EDITION - Launch of Order Network eXchange

SPECIAL EDITION - Launch of Order Network eXchange

18. November 2025 31:41 Min. Zu Gast: Kelly Goetsch
0:00 31:41

Worum es geht

In dieser Sonderausgabe geht es um den Launch des Order Network eXchange (ONX) Standards der Commerce Operations Foundation. Vorgestellt werden der offene Standard, das zugehörige GitHub-Repository und der Kontext rund um E-Commerce-Plattformen wie Shopware. Die Folge gibt einen ersten Einblick in Zielsetzung und Bausteine des neuen ONX-Standards.

Als Artikel lesen
Autor: Commerce Famous Podcast

Zu Gast

Chairman der MACH Alliance, CSO von commercetools, Co-Host des COMMERCE TOMORROW Podcasts

Kelly Goetsch ist Mitgründer und aktueller Chairman der MACH Alliance sowie CSO von commercetools. Seit 2018 ist er Co-Host des COMMERCE TOMORROW Podcasts, hat vier Bücher geschrieben und ist für seine pointierten Beiträge auf LinkedIn bekannt.

Folge teilen

Transkript

Welcome to Commerce Famous, a podcast that celebrates the celebrities of our industry. Join us for a conversation with the taste makers in ecommerce. In their daily lives, these are just normal everyday people like you and me. But to us, they are commerce famous. All right.

Welcome to a special episode of Commerce Famous. We have a return guest joining us today for a special episode. Welcome, Kelly Goetzch. Thanks for having me. Great to be here.

Well, I don't remember how long it's been, but you shocked the world, when you moved on and and left Commerce Tools and joined a company called Pipe seventeen, which is a pretty big news in our industry. I'd love to hear from you how it's going, and then we'll use that as a segment to talk about your next big project. It's going amazing. I I just got promoted to president, last week, which I'm happy about and been setting myself to four for quite a while. And generally, I, I see a lot of the action moving to order management.

You know, if I, if I look around at the order management space, there's a, a lot of legacy entrenched solutions, and I think those are perfectly fine and they work well, but I think there's a a need in the market for something like a Shopify or something like, you know, Shopware where it's a much easier type of solution. It's not a, you know, bunch of little, you know, APIs that you have to put together. I think there's a lot of folks out there who just need somewhere to dump their orders to, and we fill that need really, really, really well. It's the simple button for order management. And how we're doing that is we've got a network of 300 selling channels back office systems of record and fulfillment channels.

And that network is really the foundation of our product. So you just kinda take your API keys in, you plug them into the network. And then on top of that, we have AI based order management, and we can route around, hurricanes and labor strikes, and we can do least cost routing, you know, all the traditional and and advanced order management features. And I'd worked for the CEO and founder of Pipe seventeen for many years. Love what he and team were doing in a lot of, ex Oracle employees.

I'd been an advisor to the company for the better part of a year before I joined. And it just felt like a very natural fit to, to make the move into order management. So I've always done order capture. Yep. It's a it's a great move.

I I tend to agree. I've obviously been in e commerce for twenty five years and order management tends to be, a little bit more sleepy in the way that it thinks about innovation and evolving the industry. And so I was thrilled to see you landed there. Obviously, we're partners. But talk a little bit about what you're up to next, the order network exchange, and and how it's essentially gonna set a new standard as it relates to order management?

Sure. So, you know, looking around when I I landed at Pipe seventeen, I thought, how ridiculous is it that there isn't a standard yet between selling channels and selling channels include commerce platforms like you guys, AI heads, you know, whether it's a chat GPT or Claude or somebody like that doing shopping, we're seeing a lot of social commerce. Like TikTok is a 100,000,000,000 GMV marketplaces. You know, there's a big wide world out there of places that can capture an order. And there's no bridge between that and that big wide world of fulfillment.

And with every single commerce replatform, you have to do a big heavy lift to integrate with a traditional order management system. You have to bolt those two together, weld those two together. It's often the longest, you know, pull in any, any implementation. And we look across the industry there's standardization and industry cooperation everywhere you look. So we have barcodes, right?

We have GTINs we have PCI, right? Payments vendors got sick of, each enforcing their own standard and got together and released PCI. We see AP two and ACP now, you know, with the agentic heads starting to standardize. But, you know, we don't have this yet. And, you know, I'm sure, you know, going back in time when, you know, swift and ACH were built, you know, I'm I'm sure some guys in suits in Manhattan and smoke filled offices said, hey, this is getting pretty ridiculous.

Like, you know, Earl, why are we doing these point to point connections to every single bank in the world? Wouldn't it be better if we just built a standard and then we could just pass money through an ACH or swift standard? Right? We don't have that in our space. We just don't, you know, the closest thing approaching it is EDI.

And I think we can all agree that EDI is a gigantic mess that everybody hates. So that, and then MCP model context protocol have really risen quickly. Still less than a year old. It's coming up on its one year anniversary on November 25, but model context protocol is the protocol for how you get data from AI heads and other heads to SaaS applications. It's basically a layer over the top of your APIs.

So if you're in ChatGPT and you wanna access and write to a calendar, you connect it up through MCP. If you wanna connect up, you know, ChatGPT to Shopware. Right. You would do that through MCP. Right.

It's all MCP based at this point. And MCP makes integrations between these desperate systems dramatically easier. So we see a lot more selling channels. We see MCP finally rising. We see agent commerce really start to take off, and I see an entire lack of any specification or standards.

And I checked around with some friends who would know these things and nobody was working on it. So I said, you know what? Let's just do this. Let's give it a shot and see how it works. Obviously I have some experience doing this with the mock Alliance and this being much more standard focused, much more technical.

So look, we're at 62 members right now. And we hope to launch with, at least 10 more so we could realistically crack 70 members who are backing the standard. Commerce famous is proudly presented by Shopware, the leading open source e commerce platform for mid market and lower enterprise merchants. More than 50,000 clients already process over $25,000,000,000 in annual GMV through Shopware. Find out more about Shopware and the best value in e commerce at shopware.com.

Well, let's talk a little bit about ONX. I don't know. Is that the is that the working title, ONX, or is it do do I say the whole thing, one network exchange? It's order network exchange, and we're, calling it Onyx. Onyx.

Standard. Well, Shopware is thrilled to be a part of it. We're one of those 62. But talk a little bit about the conversations with the tech partners and and other firms who are part of this. What was the conversation like?

Did they embrace it? Was there pushback? Kind of maybe maybe that gives you a sense or the audience a sense for what kind of an idea we're talking about here. Enthusiastic embrace. And my pitch to them was, look.

You're doing MCP anyway. Right? And everybody said, yeah. We're either doing it now or we plan to do it imminently or whatever. But every single vendor is doing MCP in some way.

And my pitch to them was, Hey, you can pull this spec off the shelf and we'll save you a ton of work for having to build this yourself because you're having to build one anyway. And if you pull this off the shelf, you not only get to collaborate with your peers from across the industry, but you get interoperability and interoperability makes a huge difference because with interoperability, now you can start, capturing orders from any selling channel from anywhere. And even within the OMS space, you know, the biggest challenge with order management has been integrations. And we at Pipe seventeen have solved this very neatly through our network. But for the traditional OMSs who I pitched this to, this is great because if a three PL adopts the standard, then all they have to do is call the three PL's MCP server.

So it's MCP up to the selling channel from the OMS. It's also MCP from the OMS back down to the three PL. So it makes that network basically for them. And once the world is adopting the standard, we're in really good shape here because everyone will be speaking the same language and we'll finally have the, you know, swift equivalent, but for, for commerce orders. Now, what are the so compare and contrast a little bit so people understand the benefit of of of this approach versus, say, a classic EDI approach and a, you know, a snowflake integration type approach.

Like, what where does it get felt? Does it get felt in the in the merchant? Does it get felt at the agency? Does it get I mean, it's kinda put more context around it. So EDI is barely a standard anymore.

So everybody just implements their own there's a document that you have to produce along with your EDI implementation. Right. And there are lots of companies out there whose sole job it is to, you know, build a network out of EDI and EDI is for B2B orders primarily. So if you're general motors and you're buying steel, right, you're probably buying that steel from a factory over EDI. Right.

EDI has never really touched B2C orders. And a challenge historically has been inventory visibility, especially as you have channel fragmentation and you have fulfillment channel fragmentation. And that to me surfaces in a way like, okay, you're on a selling channel and you see inventory available. Great. But that could be cashed across three, four different systems.

Right. It could be cashed, you know, in, in many. So you could imagine, and this is the standard out there. You've got a three PL, which has a view of inventory because there's, you know, a guy punching in the numbers in a green screen right there in the warehouse. And it goes from there all the way up to an OMS of some sort.

And then it goes from the OMS up to the commerce platform. And then it goes from a commerce platform up to, say, TikTok. Opportunities for this all to get, out of date very quickly. And again, with lean inventories these days, if, if you've got, you know, 10 pairs of size thirty two jeans that you're trying to sell, it becomes very challenging to manage inventory if you're selling those 10 jeans across 10 different selling channels, which we see quite frequently. So the customer benefit here is the selling channel like you or AI or, a social head like TikTok, they can directly query the three PLs or the WMS's MCP endpoint to retrieve the inventory at ground source source.

They can get, delivery promise dates. They can offer up shipment options and you bypass a lot of those intermediary layers very, very quickly. So you're you're getting, you know, you're avoiding stock outs, for example, you're avoiding issues where somebody thinks that they bought something, but there isn't actually inventory to fulfill that thing. You're getting much more accurate delivery times. And for the brands, you're getting much better customer experience because, again, you're just collapsing a lot of these back office layers of inefficiency.

It's dramatically easier to do. Yep. It's I mean and so I have to imagine, Kelly, that the merchant benefits because, obviously, their customer has a better experience, which it's which is the gold standard. But, secondarily, doesn't this lead us to a road where it is less expensive to set up up to to update, to run, etcetera? Yes.

Because the network that OMSs have to build out becomes easier because it's all MCP based, and all of the back office systems are communicating with each other through the same standard. So that's one beneficial area. The other beneficial area is, OMS integrations. You know, you don't have to do them anymore because now you just point it at an MCP endpoint and you're good to go. So all of the work of implementing a new OMS is it's gone, basically.

Much, much, much easier to do. Yeah. So this in a way levels the playing field for a merchant who's trying to add more sales channels because ultimately everybody plays with the same standard. You have your data, you have it in the way everybody wants. You should be able to point it at any end point and begin using it and interacting with it.

Is that Correct. Is that it? Correct. And I think we're entering an era where selling channels are becoming more and more disposable. So you can imagine someone taking a commerce platform and using AI to build a pop up storefront for one day.

You know, maybe it's Black Friday storefront, right? Or a special product launch storefront. And in that world, you wanna be able to very quickly connect up to an OMS and you wanna be able to dump those captured orders somewhere. And this provides a fantastic way to do that. Much easier to do this versus point to point bespoke integrations every single time.

Yep. Well, that's an interesting point because, we you know, if I think a little bit about today's conversations with merchants, everybody's talking about GEO and how to be optimized for these LLMs so that you're showing up in their searches. And we actually have a merchant in our in our pipeline who had more traffic from ChatGPT than they got from Google through organic search for the first time a couple of months ago. And so that was a great surprise me. Yep.

Yeah. So maybe talk a little bit about kinda how the ONX Onyx and and the ability to kinda reach all these different endpoints connects. Yeah. So at the moment, we're not doing product listings Yeah. At all.

So that's outside of it. It's something that we've talked about doing. You know, we have the order network exchange. You know, maybe we extend this in the future to do a product network exchange, for example. But we we really wanna nail this standard, you know, or order protocol first.

But I guess the question was in more of a way of saying, like, if, if if a merchant wants to have all these new different selling channels and endpoints, right now one of the reasons you don't do it is because of the return on aggravation of setting them all up and getting all of the data to flow back and forth. In the world where you've got Onyx set up, you've taken a decent portion of that challenge away from them. So there could be a thousand more channels that you could now support made up number because of the standard. That's kinda where I was trying to go. Selling channels, also fulfillment channels as well.

Yeah. And back office systems and everything. Right? Dramatically easier to do. Yeah.

Yeah. One of the things you and I have talked about over beers is, you know, in the world of AI, there's a lot of conversations around the role of brand and what will happen to brand.com web. Do you have a point of view around that you wanna have a conversation around? I don't know. I I I have theories.

I can also talk myself out of those theories as well. You know, one side argues that the web is going to die as we know it, you know, historically. And after the web dies, at that point, we're gonna have AI be the head, basically, and AI will step in. And, you know, basically you're gonna be browsing what we would historically call the web through chat GPT. Right?

I could see that. And that's much more personalized. It's much more dynamically assembled. It's much more intent based. You know, it's an easier easier way to do things.

So that's one approach. You could also argue that the Chargeepts of the world are gonna be great at product discovery, and they're gonna be good at helping folks select, but they're still gonna go to thebrand.com to actually transact. You know, so I could see that happening as well. There's another school of thought that says the web is gonna continue because people like their brand experiences. And, you know, maybe there's some discovery that happens, but, you know, largely it's going to be business as usual for a long time.

Yeah. I only ask you these questions because you're such a thought leader on almost everything. I don't know where to think about. Advantage of the moment. I I could very easily argue any one of those points definitively.

I think it's going to be a non I think it's gonna be a material portion of transactions will happen through AI over the coming years. Yeah. Well, let's let me ask you what that looks like long term, though. Well, so so obviously well, I'll tell a little story first, and then I'll get into the question. So friend of mine is a small business, and he's really, really good financially.

And he starts studying his product margin and gross sales over time. And what he realized over the last probably five to seven years is that his cost of Google had gone up. His cost to make his product has gone up. His revenue had gone up slightly. And what he came to the realization was he was really just in business to fuel Google.

I mean, that's who was making all of the money. And so we're at this really interesting paradigm where ChatGPT and and other AI engines are starting to steal or take share. Mhmm. And what obviously, that has a different business model than your traditional kind of PPC model. And so as you think about kind of what the end state is for how they're gonna monetize it, is it, well, you tell me I've kind of led the witness enough.

You know, the options. Where's this going? Alright. So when Google so so when Yahoo first came on the scene, right, you had a directory listing of websites. Right?

Human curated by and large. When Google hit the scene in, like, February, it was dramatically better. Right? Like an order of magnitude better customer experience. And at the time, it didn't have ads, and it was a pure uncorrupted experience.

The web itself didn't have all those stupid GDPR consent pop up banners, and it didn't have the join our mailing list for 15% off banners. And it didn't have a 100 cookies per page visit. You know, the web was dramatically better then. And Google was a huge step forward because it gave you the page of what you were looking for. I see AI as being in that same category of being dramatically better because it's not just giving you the page anymore.

It's actually giving you exactly what you want. Right? Like, I don't wanna I don't wanna go to weather.com and see the weather today. I want to ask what the weather is. Right?

Now the problem is is AI knows everything about you. You know, a lot of people use their AI as a personal therapist. AI is connected up to people's emails and calendars, and they work with it all day long. And they have pictures from vacations and like AI knows stuff about you. And I, I was, I had a, one of those, you know, black to me, you know, laptop bags.

I had one for twenty years and it finally got to the point where I had to replace it. And I was debating options with chat GPT. And then later I was looking for a laptop sleeve to go in the bag that I bought. And it said, oh, that would match the one you just decided to buy. And I thought that was really weird because it was a couple weeks after, and it remembered which one I had said I was gonna buy.

And then it told me that it would match it. And I thought, wow. That's a little unnerving. But look, AI is just, I mean, categorically, dramatically better. And it you know, we talked about personalization a long time over the decades.

This is personalization. This is actual personalization. Now it's a good thing for customers. It's also a bad thing for customers because the open AIs of the world are absolutely going to monetize this with ads. Right?

They've brought in a bunch of Facebook people. This stuff is very expensive. Advertisers would pay dearly, you know, if somebody is constantly on their chat GPT talking about how depressed they are all day, and then you've got a pharma company who's selling a new antidepressant. Right? What pharma company wouldn't pay a $100 to, serve up a personalized ad personalized ad to the person who's, you know, been talking to their chattyPT all day about how depressed they are.

Right? That's ridiculously valuable. It's ridiculous margins, you know, 90% plus gross margins on that stuff. It's a cash printing machine. And being that OpenAI and others are for profit, nakedly for profit, inevitably, they're gonna corrupt the experience in the same way that Google has corrupted the experience.

And whenever Google was, you know, down in the earnings, they would just add another ad to the page. And now all of a sudden, the Google experience is nothing but ads and over SEO garbage. Yep. It's totally true. But it's so funny you you you go back in the way back machine because I think that's the best way to get perspective.

Who remembers taking an Uber ride for $6 when it when a taxi costed 40? Take an Uber now, costs 40. Or or, or Netflix used to cost $6.99. Now I don't even know what it cost, four or five times that. Right.

Disney just doubled their price again for the third time. So once you give tech companies leverage, the one thing that I think is true is they will use it. Well, my favorite word for 2025 is if you've heard that term before. I haven't. What does mean?

Tech companies inevitably destroy good products. And what happens over time is advertising corrupts them, legal frameworks often that those companies themselves lobby for destroy products and experiences. The raw focus on profits and engagement destroys experiences. You know, I'll give you an example. Yesterday, my wife and I were looking for something to do with the kids.

So we went to Facebook events because in the Midwest, that's what you do. Yep. And I saw that they removed the option to select what day you wanted to see events for. And they did that presumably because they want people to scroll through a list of a 100 events rather than selecting the actual day that you want to see the events. Right.

That's an perfect example of intuitification was otherwise a fine product and they purposely degraded it so that they could juice their engagement numbers. Right. You see a lot of that type of behavior. You see feature bloat, you know, I think you look at any modern vehicle. Right.

That's probably not a Tesla and you see that type of, of button, you see a lot of lag in the, in the UIs. Right. The, the new infotainment systems are borderline unusable because they're just so bloated and have so many useless features. Like I get in my car and I use Spotify. Right.

I'm a Spotify guy. I've only ever used Spotify. It constantly pops up whenever I start it, that I should, listen to the preview of, Sirius XM. And I'm sure GM is paid a fortune to have Sirius XM on there, but like, that's an example of where, you know, a correct experience would be to realize that I've only ever in my entire, you know, life having that car selected Spotify, and then it's constantly trying to upsell me whenever I get in. Great example of incentivication.

So we're in the very early stage of AI. I have no doubt AI will get worse, and I'm sure at some point, it'll turn into, you know, something very evil and not good. You know, I'm sure, you know, a couple CEOs after Sam Altman, they're gonna be, you know, auctioning off your data and selling it to the highest bidder. Yeah. That's my pessimistic view of the world.

I think you're probably right. What was that thing that, you could send in your your swab and they could tell you your ancestries? What was that one? 20 three and Me. 23 and Me.

Yeah. And they did exactly that. They're not selling all your data. Yeah. Let's sell it to insurance companies so that insurance companies can deny you of coverage.

Yes. Yeah. It's now a preexisting condition. Being alive is a preexisting condition. Very good.

Well, let's let's transition back, in our in our last ten to fifteen minutes and talk a little bit more about Onyx because I think it's really cool. When will when will you publish the standard so people can start to see what it is, how to engage with it, and how to start playing ball? It's available. It's available today. So you go to GitHub and search for Commerce Operations Foundation.

Also, if you go to commerceopsfoundation.org, you will see the link to the spec right there. Okay. We'll put them in the chat. And, you know, you said you had 62 getting to 70. You wanna name some of the companies that are in the first round of support?

I think that'd be a good call out for them. Yeah. I I think I have every single OMS vendor in the entire category. I've got Manhattan, Sterling, Kibo, Fluent, One Stock, Fulfillment Tools, Deck. I've got a whole bunch of those.

I've got three PLS, including Radial and Rider. I've got Geotis. You know, these are all pretty big sizable companies. CEO of Ryder was kind enough to give us a nice quote for our launch, which was which was very nice of him. And then we've got WMSs.

We've got three p ls, and then we've got commerce platforms, you among them. So we've got pretty broad support. And then on top of that, we've got some brands as well. I gotta say, I think it's really cool that well, actually, I'll I'll tell a little story. So it's really interesting in where we're at in the world, and Shopware has an agentic commerce alliance.

And when people ask why are we doing it, I say, hey. I it I'm old, and it feels a little bit like that moment where you've got a lot of really big companies trying to control the airwaves. And if you think back to AOL versus the open Internet, it was a similar type of a moment. Right? And so what I always say in that conversation is that open should win.

And and and so, that's kinda why we're doing what we're doing, and it feels a lot like what you're trying to do. Do you agree with that? Yeah. I always default to open. I believe in having a big tent.

And look, this is not rocket science, what we're doing here. This is not it at all. We all gain more as an industry collectively if we all adopt it, you know, and again, going back to standards, you know, if some random company wants to invent their own barcode standard, like they do so at their own peril these days and barcodes are nice. Everybody works with them, right? The whole ecosystem is built out around barcodes working as they should.

It's a standard. And Onyx is that standard now for flowing orders between selling channels and fulfillment and back office channels. Yeah. Phenomenal. So if folks wanna get involved, how should they reach out?

Should they go to the website? Should they contact you directly? What's the most appropriate way for people to get involved in Onyx? Commerceopsfoundation.org, and we have a sign up button. So that would be helpful.

If you wanna sign up and actually back the standard as a brand or, a vendor, you are absolutely encouraged to do that. We'd love to grow the tent. So contact me directly. You can find me on LinkedIn or, through the Commerce Ops Foundation website. And then are also looking for folks to join the various committees and, and boards within the organization.

So we've still got a spot or two left on the executive board. You know, I have the CEO of Kibo on the board with me. I've got the CTO of, Manhattan on the board with me, you know, a lot of good solid industry representation, you know, SVP from, IBM Sterling who leads that whole program for them is on the board. So would love, you know, if there are any folks out there who wanna join and participate, you know, would would love to have you, join us. Fabulous.

And the the mantra of build a bigger table, bigger, bigger tent, I love it. Is there anything required for these brands? By by the way, the other question I was gonna ask is, can merchants be a part of it, or does it should it always be tech partners? Can agencies belong? Like, who who's who's who's able to be a part of it?

Yeah. So brands absolutely can. And if the brands are running any Onyx compliant order management system, they're by default then compliant with it. So if you're running, you know, Kibo, for example, then as a brand, you are therefore running the Onyx standard or soon will be, and would love to have your participation directly as well. So you can join a brand, you know, as a brand, you can join the, the organization and advocate and advance the standard.

So that's one way to, participate. And then we're looking for folks to actually drive the tech standard forward. You know, that's been an area of focus. We have an industry advisory board, a vendor advisory board, and a couple others. So nice opportunities to get involved network with peers who are all going through the same transition at the same time as you.

And also, I mean, just, you know, speaking selfishly, it gives a lot of folks out there an agentic story that they can tell their board or tell the public markets, because I know everyone's on the lookout right now for, you know, for something to be able to talk about. Absolutely. Well, I will say I admire the great work you did on Mach, getting that thing established. And I think some of the principles you've brought over, into into Onyx are really admirable. Like, the fact that you've created all these these committees and the the ability to be involved and essentially creating a community around something that rises the tide for the entire industry is really impressive.

So Appreciate that. I I'm doing what I did with the mock alliance, which is I pointed myself as benevolent dictator for a year. And then and then I stepped aside because first, I don't need two jobs forever. You know, logistically, it gets hard. But second, you know, these are these are meant to be community initiatives.

And now look, the mock alliance has a full time staff of 12, and we've had four different presidents now, which is great. And, you know, if I got hit by a bus, the mock alliance would be just fine without me. Right. And that's how you should evolve these nonprofits to get to where, you know, you're it's not just one person driving it. It's a community where all of us together are driving it, not just, you know, one person or one company.

So, you know, we at Pipe seventeen have donated our standard as the starting point. It's a very nice neutral, open standard, open source. And, we're finalizing the standard right now and we'll have that published by January 1 and then we'll all get started, adopting it or finish adopting it in many cases. Amazing. Well, I'll give you that as the last word.

Kelly, thank you for joining us on, Commerce Famous today. There's very few people more Commerce Famous than you. So appreciate you being here and and creating this standard. Alright. Thanks for having me and all the great work you guys are doing.

Mehr von Commerce Famous Podcast Alle Folgen als zitierfähige Episodenseiten

Veröffentlicht mit CiteCast