WEBVTT

00:00:00.000 --> 00:00:17.495
<v Speaker 1>You know, there's there's there was a choice that always had to be made. We wanna we wanna do something really fast, you know, if a merchant or brand wants to do something really quickly and get to market. Then they might be, you know, making some sacrifices in terms of the power extensibility of the platform. Right? We wanna maintain that level of of enterprise, enterprise and scalability.

00:00:17.495 --> 00:00:24.855
At the same time, there's a question of, hey, can we do something a little bit faster? Right? Can we do something a little bit more easier? Can we lower the TCO for day to day operations? And the answer is yes.

00:00:24.855 --> 00:00:27.095
Right? It's twenty twenty six. Let's combine the two together.

00:00:27.095 --> 00:01:00.720
<v Speaker 2>Welcome to the Retailgentic podcast where we explore the intersection of retail, ecommerce, and AI agents, which we call Retailgentic. Our mission is simple. We are shining a bright light on the new agentic AI shopping megatrend, which will disrupt the trillion dollar digital retail market. The stakes are high. Google, OpenAI, Microsoft, Perplexity, Amazon, Apple, Meta, and Anthropic are all fighting to be the consumer AI shopping agent winner.

00:01:00.880 --> 00:01:17.675
There are gonna be some tough decisions for online brands and retailers to make. What happens to checkout? How should payments work? Could you continue to invest in retail media networks? What does the world look like when over half of your traffic is not from humans but from AI agents?

00:01:18.315 --> 00:01:41.570
These existential questions are the tip of the iceberg. On this podcast, we're talking to industry experts, analysts, and influencers to get answers to these questions, identify trends, and best practices. This is Retailgentic, your podcast partner in navigating our AI shopping agent future. And now here's your host, Wingo.

00:01:43.385 --> 00:02:03.560
<v Speaker 3>Welcome to this episode of Retailgentic. Today, we have a rare doubleheader. Due to some scheduling issues with the two folks I wanna interview, we didn't do a three way interview. We did two singles. So I interviewed these leaders sequentially, and they are two of the leaders from Salesforce.

00:02:03.640 --> 00:02:31.540
And the topic today is their big June release, which they just announced last week. That includes the launch of their newest platform that they're very excited about called Storefront Next. So up first to explain to us what's going on and why he's so excited, we have the now executive vice president and general manager of Salesforce Commerce, Nitin Mangtani. Nitin's been on the show before. He nagged it way back.

00:02:31.540 --> 00:03:00.010
He ran Google Shopping for a long time. So he he knows a lot about ecommerce, and he is a great leader that Mr. Put in charge of this at Salesforce. And you can tell he's chopping it up over there and pulling a lot of things together that they really needed to do to get kind of serious about get back in the game as it were. Nitin shares with us the big picture strategy of where this release came from, what was the inspiration for it, the big themes.

00:03:00.330 --> 00:03:17.625
We haven't talked since they did their acquisition of Simulate. Congrats to our friends at Simulate. And he explains how that fits in. Also how this hits, there's the commerce side, the point of sale side, and then the warehouse side. So he kind of ties all that together for us.

00:03:18.185 --> 00:03:51.480
And yeah, so he positions this as their biggest release in five years. So, and there's a lot of really interesting things, in this storefront next. I don't know if this is due to competition with Salesforce and BigCommerce and all that kind of thing, or does seem like they're, like, really modernizing the platform, which could be the ultimate impetus of this. After Nitin, we have the VP of product management for Agent Forest Commerce, Lennart Stevens. He we go through his background briefly because he's new to the pod.

00:03:51.480 --> 00:04:25.160
Then he takes us deeper into all the cool new Agintic commerce features, which is our whole focus here at Retailgentic. And he tells us all about those aspects of storefront next, how it fits in with the the other storefronts that Salesforce has, and much more. I hope you enjoy this doubleheader conversation with Nitin and Lennart. Welcome to this episode of Retailgentic. It's my pleasure to be joined for the second time by my good friend, Nitin Mangtani.

00:04:25.675 --> 00:04:42.140
He is now the EVP and GM of Salesforce. When, Nitin, when you were here last time, you didn't have this many cool things in your title. So I know it's only a data point of one, but I I think the conclusion I would draw is your, an appearance on Retailgentic is good for a promotion. Congratulations. Glad we can help.

00:04:42.140 --> 00:04:51.180
<v Speaker 4>Yeah. Well, Scot, you know, I've known you for more than a decade now, and, it's always it's always a pleasure, being being on your podcast.

00:04:51.420 --> 00:05:14.450
<v Speaker 3>Yeah. So you're here to tell us some of the big news out of Salesforce. Earlier on this episode, we we kind of went into it with your your colleague, Lennart. Tell me what's going on from the the EVP GM side of the fence? Like, what's what's big about Commerce Next and what what you guys are up to?

00:05:14.450 --> 00:05:35.745
<v Speaker 4>Yeah. It's been a it's been a good trend, and it's almost been two years since I joined Salesforce. So I sold my company, August, September 2024. So there's a lot of good things we have done in the last eighteen to twenty four months, including the Simulate acquisition. I'm trying to recall.

00:05:35.745 --> 00:05:38.950
I don't think I had a... We had acquired Simulate last time I spoke to you.

00:05:38.950 --> 00:05:40.630
<v Speaker 3>That's correct. Yeah. You did that after.

00:05:40.630 --> 00:05:47.270
<v Speaker 4>Yeah. Yeah. So that's been great. So as you know, search is having a having a moment. Right?

00:05:47.270 --> 00:06:02.095
Like, it's rethinking the entire search in the LLMs and the agentic world. Right? Obviously, common search has been around for many, many years. And so we found these these founders, Vivek and John. They're based in Boston.

00:06:02.575 --> 00:06:17.600
Amazing founders. They had several folks, PhDs from MIT, Stanford, CMU. Top notch talent. And I was just blown away when I looked at the product. But I'm like, who am I to calibrate a company this young?

00:06:17.600 --> 00:06:30.435
You know, sometimes, you know, you just don't know. And so I started calling people. And in fact, their referral came from a good friend, John at. He told me, he's like, John Cosa. He's like, you you...

00:06:30.435 --> 00:06:46.015
Nitin, you should look at them. And, fast forward, we closed that acquisition a few months ago. And I'm so happy, Scot. Like, this is our biggest innovation when it comes to rethinking the entire discovery experience. And then...

00:06:46.015 --> 00:07:02.570
And that's huge. And then combined to... There are three other things we have done in this release. We are calling it the June release or the kind of summer release, but early summer release, not not late summer because summer definition is a little longer. So it's June.

00:07:02.570 --> 00:07:14.730
You don't need to wait till July or August. And as part of it... So Simulate is now part of the Salesforce family. It's available to Commerce Cloud customers. In fact, not only Commerce Cloud, but even non Commerce Cloud customers.

00:07:14.730 --> 00:07:30.150
So So we made a decision that whatever reason, if you just want a better search and a better shopper agent, we'll offer it to you. And we'll keep it... Because, you know, the feedback was, can you make your platform more modular? And we are like, not a problem. It's API first.

00:07:30.150 --> 00:07:42.105
Licensing one, so we'll do it. Second is the storefront next. You know, front end experiences, as you have... As you've seen, they have evolved. Like, you know, early days, we were trying to build desktop pre iPhone, pre 2007.

00:07:42.105 --> 00:07:52.265
Like, everybody was like, how do I pixel perfect my desktop? And then iPhone came, and everybody's like mobile. And then mobile became mobile first. Right? So it kind of went through this interesting journey.

00:07:52.505 --> 00:08:23.590
And then native apps came in. And then now everybody's like, can I just wipe code the entire front end? Which is mind boggling because what was considered this very herculean task and a very complex project, all of a sudden, kind of you reduce the barrier, and that's where a lot of focus of Storefront Next is is really bringing white coating and kind of the modern way of thinking and how do you build front end experiences. So that's a big thing. The third part of this release is a new merchandising experience.

00:08:23.590 --> 00:09:05.445
Completely redone merchandising, where administrators can visually and prompt based, say, boost these products, you know, new arrivals or demote this product because I'm running low on stock. And now you can just do it both prompt based as well as you can see a visual workbench, instead of, like, you know, a lot of complex manual clicking going on. So that's the third big, innovation that we are bringing to kind of this release. We're making a lot of changes also to Business Manager, which is kind of our admin console for Commerce Cloud because it's an important part of it. And then on the POS, we're kind of really trying to make the product global.

00:09:05.605 --> 00:09:20.620
We are already in 20 countries, and now we're gonna make this available in 50 plus countries. So it's a huge jump from... Because point of sale has a very complex fiscal requirements. To run it in France, you need to go through a multiyear process. It's not as easy.

00:09:20.620 --> 00:09:36.665
So now we're gonna go from 20 to 50 countries. So that's a huge thing. And then we are, later this year, we're making our point of sale available on Android and Windows as well. So not just iOS, which is an important partnership for us. But now along with it, we are offering it in...

00:09:36.745 --> 00:10:01.120
In fact, we we just launched our beta program on Android last week. So that's gonna be a huge kind of, you know, enabler. And then, on the OMS side, really rethinking the routing logic. So all this is coming together in this June release. I would say this is our biggest release when I bring all of these pieces together in the last five years, and I'll tell you in a minute why this is our biggest release in the last five years.

00:10:01.120 --> 00:10:25.000
<v Speaker 3>You mentioned this on the last podcast that you're... There's kinda, like three legs to the stool, it's, so the June release has seen major upgrades to all three, which is interesting. For for folks that aren't as familiar with the the the platforms you have, so the Storefront Nest is, everyone kinda come along for the ride? Or if they're on a previous version, they're gonna have to kinda upgrade to this new version? Or...

00:10:25.000 --> 00:10:33.735
And and if they if they do need to move to it, is it like a replatforming? Or do you guys give them tools to go from version a to b to c to d kind of thing?

00:10:33.735 --> 00:10:41.015
<v Speaker 4>Yeah. Yeah. So it's available for anybody. We we don't want to limit. And we kind of have a spectrum, Scot, when it comes to customers.

00:10:41.095 --> 00:10:52.590
There's a set of customer which is like few clicks and get me live. I don't even want to look at the quote. Right? I don't have in house IT, and I don't want to really manage and deploy code. Right?

00:10:52.590 --> 00:10:59.390
It's all... Everything is done by a business user. And it's... You know? So that's set one.

00:10:59.630 --> 00:11:19.570
Set two is I need a high degree of customization, but I'm still not willing to, like, really write ton of code. I might, like, redo my PDP because PDP is very important to me, but rest of it, I just want to use your page designer. And how can you make it component wise and modular? Right? And then the other side of it...

00:11:19.570 --> 00:11:32.345
There's kind of in between also I want to use an external CMS. I have using x y z CMS, and my business users love it. Can I plug it into you? And the fourth set is give me headless. Let me just build this.

00:11:32.345 --> 00:11:43.790
Give me your API. I'm just gonna build my own front end. I might still use a CMS, but I'm not using any of your UI components. And that's fine. It's a huge spectrum, and our current customer base is in spectrum.

00:11:43.790 --> 00:11:51.390
So it's for all four audiences. It's not limiting any audiences. It's for all four audiences. As far as your question on, is this a replatforming? No.

00:11:51.390 --> 00:12:05.055
It's much more simpler. I wouldn't say it's like one click. So if you had returned some complex front end code, you know, we have built a bunch of migration utilities. Especially with Cloud Code, you're able to do a lot of interesting things right now. You can...

00:12:05.055 --> 00:12:12.180
You know? So in the traditional days, Scot, as you remember, redoing a front end was like a massive project.

00:12:12.180 --> 00:12:13.940
<v Speaker 3>Capital t project. Yeah. Mhmm.

00:12:13.940 --> 00:12:19.620
<v Speaker 4>Yeah. Yeah. This just, like, shrinks. Right? It just fundamentally shrinks the effort you need.

00:12:19.620 --> 00:12:29.975
And then on top of it, we have all this jump start migration packages that we are giving. So if you're using SFRA, you know, here's the migration utility to storefront next.

00:12:29.975 --> 00:12:42.220
<v Speaker 3>And imagine... So clogged code, you must be surfacing all this through MCP. So so you're kinda saying, here's all the components you can you can interface your UI with. Here's the here's the OMS. Here's the content piece.

00:12:42.220 --> 00:12:46.700
Here's the catalog. Here's... You know? Is that is that kind of how how that would work?

00:12:46.700 --> 00:12:52.315
<v Speaker 4>Yeah. Multiple... I mean, MCP, REST is still... Nothing wrong with REST. You still have some very typed endpoints.

00:12:52.315 --> 00:13:08.080
And if you, you know, you just want to make calls. But, also, we have a lot of out of the box baked UI components. Right? So if you don't want to build UI, right, we still have the page designer, which is our declarative drag and drop tool to design your home page. Right?

00:13:08.080 --> 00:13:35.675
So if you don't want to go directly to our APIs, which you can, you're welcome to, you can stop there. But then when you look at your home page and you're like, wait a second. I actually want this very special butter bar on the top, and I want to make changes. And that's where you can take our code because the code is open source, that part of it. And you can just prompt engineer it and say, let me change this butter bar to be this very unique styling and something very special that I have creativity and I want to do it.

00:13:36.000 --> 00:13:45.120
So it's not... I mean, you could start from scratch, but that's not the intent. You're starting from a 90, and then the last 10%, you can white coat it.

00:13:45.120 --> 00:14:00.195
<v Speaker 3>The big trend in AI developers is ephemeral UIs where they just kinda create them on the fly. Do you guys have any elements of that where you can kind of, you know, maybe like create a landing page on the fly or or something? Or is it all kind of statically determined before the visit?

00:14:00.195 --> 00:14:12.030
<v Speaker 4>You could do a lot of dynamic UIs and that... That's gonna continue. The other place you would see dynamic UIs is in the is in the chat interface. Right? So if you think about, right, our our...

00:14:12.030 --> 00:14:33.720
What we call it shopper agent. Our shopper agent is an answer to powering agentic commerce on owned and operated properties. So one education, and I think we talked about last time, but I kind of emphasize. When we think about agentic commerce, we clearly think about agentic commerce on owned and operated channels. This is your website.

00:14:33.720 --> 00:14:43.320
If you are Ralph Lauren, if you're Michael Kors, if you're Coach, if you're Louis Vuitton, you're... It's your website. It's your mobile app. It's your physical store. Right?

00:14:43.400 --> 00:15:08.330
This is the channel that you own and operate. And you do need agentic experiences on your owned and operated channels. But you also want to participate through UCP on agentic commerce that will happen on Gemini and other avenues, OpenAI, eventually TikTok and others. Right? And in some ways, I think some of the articles out there don't clearly explain the difference.

00:15:08.330 --> 00:15:22.705
You and I get it. We have lived this life. Like, back in the days when I was at Google, you know, we partnered together with your company. You're the founder, CEO, and chairman. I think I try to emphasize this clarity so that people understand that those are two distinct things.

00:15:23.105 --> 00:15:29.350
And one is not mutually exclusive to other or one is not better than other, but you need both. You cannot skip one.

00:15:29.350 --> 00:15:32.310
<v Speaker 3>You get... Gotta have your own Rufus and you also gotta play off-site.

00:15:32.310 --> 00:15:37.270
<v Speaker 4>Exactly. Or or now, I guess, they they... I think instead of Rufus, I think we have

00:15:37.430 --> 00:15:38.790
<v Speaker 3>Alexa for shopping. Yeah.

00:15:38.790 --> 00:15:41.785
<v Speaker 4>Mhmm. Alexa or what? Yeah. Yeah. But what I think...

00:15:41.785 --> 00:15:54.745
And a lot of the noise around was like, oh, Rufus is no longer... I think that's not the point. The point here is... I mean, you know, I'm just holding, you know, the phone here, right? This is kind of Alexa.

00:15:54.870 --> 00:16:11.285
The point is not like whether Rufus or Alexa. The point is Amazon absolutely believes in this. That's why one of the only four tabs they have in the bottom iOS app... In the entire app, there are only four tabs at the bottom. And one entire tab is dedicated to this, what I call it, shopper agent.

00:16:11.285 --> 00:16:16.725
<v Speaker 3>Yeah. It's clearly... Jassy talks about it. It's almost the lead of the... Every conference call now.

00:16:16.725 --> 00:16:27.680
Yep. And it's it's moving the needle like a non... A shopper agent can... You know? And I know you guys have put some stats out there, which is always very helpful, but it's definitely moving the needle for retailers to implement it.

00:16:27.680 --> 00:16:36.160
<v Speaker 4>Yeah. And I think it's still early. I mean, I'm surprised. Like, many of our biggest competitors still don't have their act together. They don't have a shopper agent.

00:16:36.195 --> 00:16:56.910
They talk a lot about syndicated commerce. They talk about... And which is important to us, and we have very close partnership with Google and OpenAI and TikTok and Instagram and everybody in town. But to me, this is this is foundational getting. It's like in the old days, saying, I participate on Google Shopping, but I don't have a search on my website.

00:16:56.990 --> 00:16:57.950
<v Speaker 3>Yeah. It's kinda crazy.

00:16:57.950 --> 00:17:04.350
<v Speaker 4>It'll be like a massive mess, you know, if if the website had no search, but you're like, look. But on Google Shopping, I have all my results.

00:17:04.685 --> 00:17:20.320
<v Speaker 3>If the acquisition only just closed, does Simulate kind of integrated in here, or is it still... It's... It hasn't kind of been fully integrated in this June release. Is is this like the underlying technology, Simulate? Obviously, you can use it with any of your platforms, but, has that been, like, baked into the cake at this point?

00:17:20.320 --> 00:17:26.400
<v Speaker 4>It is. It is. Yeah. So we have done the integration so that way you don't need to submit the catalog twice to us. Right?

00:17:26.400 --> 00:17:32.000
So you're not submitting catalog to Commerce Cloud and then simulate. It's just one catalog. We get it. We have... This is your catalog.

00:17:32.435 --> 00:17:54.490
And then, you know, if you turn it on, it's integrated in the search box, and Simulate is indexing it. Right now, we are selling Simulate as a kind of a stand alone SKU. So it can be avail... Because it's premium, it has all the, you know, features. But you could literally turn on similar as a Commerce Cloud customer relatively easily.

00:17:54.890 --> 00:18:15.895
And the way the architecture is... And this is going to evolve again. Right now, Scott, most websites have a search box and an icon to start the agentic thing. Eventually, me, the search and agentic will collapse. Right now, there are still two entry points because the consumer behavior is still like, do I search or do I do a conversation?

00:18:17.470 --> 00:18:24.270
It's just a very like a transitional time. Google is, AI mode will become the mode. It's just a matter of time.

00:18:24.270 --> 00:18:26.990
<v Speaker 3>Yeah, Alexa first shopping will be the Amazon search.

00:18:26.990 --> 00:18:45.135
<v Speaker 4>Exactly, exactly. So whether you type in two keywords or a full sentence, it doesn't matter. Point here is, like, there's only one. So today, there is still two interfaces. But behind the scenes, the way we have done is that Shopper Agent, which is fully conversational, understands these LLM prompts and everything, still needs to do a search because you still need vectorized search.

00:18:45.135 --> 00:19:04.375
You still need to understand the semantics. So I might have a complex prompt, but that complex prompt has to be converted into a search. So this is where Simulet comes in. We have plugged in Simulet, both at a standalone search engine, but also it's powering every single prompt on Shopper Agent to make sure that the search results are most relevant.

00:19:04.375 --> 00:19:24.910
<v Speaker 3>When you guys announced Storefront Next, there... There's an interesting framing in the press release that talks about before you had to pick Enterprise or Easy. And and talk about... Is that is that... When you came into Salesforce, is that one of the first things you noticed that that there was a weird choice there that that maybe you didn't have to exist anymore?

00:19:24.910 --> 00:19:40.755
<v Speaker 4>Yeah. So I told you this is our biggest innovation in the last five years. Right? All combined. Like, you know, when you combine this package between acquiring similar, the whole thing around shopper agent, Storefront Next, the POS advancements, the OMS, and then b to b.

00:19:40.755 --> 00:19:57.460
You know, we we need to do a separate podcast and just talk about b to b commerce because I'm I'm very bullish on where b to b is going, especially in agentic world. Because traditionally, b to b has been like a static website with two images and a contact us form or 1800 number. Right? That was the start of In

00:19:57.460 --> 00:20:00.005
<v Speaker 3>a PDF. Yeah. Download a PDF.

00:20:00.005 --> 00:20:02.485
<v Speaker 4>Yeah. Or a PDF. Yeah. Yeah. Yeah.

00:20:02.485 --> 00:20:19.140
So so b two b has evolved tremendously. And all this is back... The reason I say it's our biggest release in the five years, multiple multiple kind of tenants went in. One is focus. We had a very clear focus coming into the last eighteen months, what we are doing, right?

00:20:19.140 --> 00:20:31.355
A few initiatives we are doing, and I kind of consolidated. I said, look, future is around agentic and this new front end experience. And the last thing is unified commerce, POS and OMS. Right? So we put a lot of...

00:20:31.355 --> 00:20:54.430
And focus is magic, as you know, Running a company, there's nothing like focus. Once you get the... So we got the focus and strategy right. And then we really iterated with our customers because I believe that the top down design doesn't work. If you can reduce the iteration cycle to weeks instead of months or quarters, you get a very different feedback loop.

00:20:54.605 --> 00:21:21.630
Because I don't think I'm like this big visionary. I can sit here and tell you two years, I'm going to write a PRD for two years and I'm going to be done. I can barely think ahead maybe like, well, I have a clear vision, but actual iterations, I like to do it in units of literally two weeks, traditional kind of two week iteration cycle. So we got really good at iteration as a team. And the final thing was the AI native codes, Cloud Code also helped us a ton in accelerating.

00:21:21.630 --> 00:21:42.200
So when you do fewer things, automatically you have more resources. Because all of a sudden, agent, you know, has more resources than I had before because I was not doing 20 things. I was doing four things or maybe five things. We bought in an acquisition, so there was a capital investment which accelerated the work which would have taken. So that was like a little bit of a cheat code.

00:21:42.280 --> 00:22:11.320
Instead of building the search in house and taking five years, like, okay, let's acquire a company. And so so it's it's a rotation of that. But, yeah, one of the themes was this friction between can you be enterprise ready and meet this crazy scalability demand and be this infinitely scalable and customizable? And can you give me headless API? And I need a component, and I need this very complex checkout with like eight options, which, by the way, I don't recommend anybody.

00:22:11.320 --> 00:22:29.655
Just put Apple Pay on your PDP and try to streamline. What is good maybe on a good day, but not more. But anyways, look, enterprise is complex. And there's this friction where enterprise means three year deployments, multimillion dollar projects. And we're like, well, that distinction is actually false.

00:22:30.295 --> 00:23:09.250
You could actually simplify things. You could have a discipline where you say, let me simplify for simple use cases and medium use cases. If the customer wants out of the box storefront, like, you know, storefront next, it should be clicks, not code. But for whatever reason you want to write code, in the modern era, this code should be written by Claude through prompting and not by you. And can we create the structure in our code and the guardrails and the thought process so that it's easy to take the default code we are shipping, but then somebody can iterate on it and do a lot more changes, but not in, like, again, a twelve month complex IT project.

00:23:09.250 --> 00:23:21.010
<v Speaker 3>We only have you for about three more minutes. What else would you tell all the merchants out there? Let's talk... You know, let's punch up on Agent to Commerce. What do you think about everything going on there?

00:23:21.170 --> 00:23:30.025
And were you surprised ChatGPT got out of the checkout game? And what's your view of where Agent to Commerce is going, and how does Storefront Next tie into that?

00:23:30.025 --> 00:23:45.490
<v Speaker 4>No, I was not surprised. First of all, like, I'm a pro technologist just like you. Joke with some of the people. Was like, look, I I drive the FSD, Tesla FSD, and I'm, you know, I'm comfortable with that risk because the the innovation is so beautiful. Right?

00:23:45.810 --> 00:24:07.080
And so we are definitely pro technologists, so we're not going to bet against what the art of possible. But you and I both know how hard it is to do syndicated commerce, whether it's OpenAI or Gemini. I mean, we tried on Google, remember, back in 2008 or 'nine. So in my mind, I think I was very clear that the referral traffic coming from agentic is already happening. That's like a no brainer.

00:24:07.080 --> 00:24:26.755
That's what you are doing, Scot, and your company is doing is phenomenal. I said commerce is possible, but it's going to be very hard and it's going to take some time and patience, right? And so ChatGPT, OpenAI made that decision to focus less on commerce instead of focusing more on ads and discovery. It's totally fine. And by the way, I would have done the same thing.

00:24:26.755 --> 00:24:55.125
Google wants to actually do the full commerce with UCP, so we are still working with both of them. And I was not surprised to answer your question. I thought it's just I knew this movie, how it's going to play out. And this is where I keep reminding this, that agentic commerce on your own and people, by the way, totally there was so much noise and people totally forgot about the agent e commerce on their owned and operated property. And I was like, I don't go in every CIO meeting and CEO meeting.

00:24:55.125 --> 00:25:00.805
Was like, guys, I get it. You want to be on OpenAI. Trust me. I will get you there. That's like, I'm not stopping you.

00:25:01.110 --> 00:25:17.965
But please don't forget that you still have your own website and mobile app. And once you get the customer on your website, you still need to deliver a delightful experience. And that doesn't go away. Commerce is not going to be 100% on syndicated channels. So I think it's a fair game here.

00:25:17.965 --> 00:25:26.525
But I think it will happen. It will take some time. It will be iterations. And there is a lot of complexity. And like home furnishing, you have to pick a date and a delivery time.

00:25:26.525 --> 00:25:36.600
All those things will take some time to get syndicated commerce right. But we are we are being very neutral as a platform company and, you know, providing the tools on both sides of the house.

00:25:36.680 --> 00:25:48.715
<v Speaker 3>Sounds like... Sounds smart. Gonna gonna neutral in the middle of these things is what I've learned for sure. Nitin, you taking time out of your busy schedule to join us. Congrats on the June release.

00:25:48.715 --> 00:25:55.710
Sounds like you're you're getting some steam there. So congrats on that and congrats on the promotion and the acquisition. You've been busy.

00:25:55.710 --> 00:26:07.710
<v Speaker 4>It's been a good busy start. And great to see all the progress on your company. I feel you add tremendous value to this retail ecosystem, and I'm glad leaders like you continue to play this role in the agent e commerce. So thank you.

00:26:07.875 --> 00:26:10.035
<v Speaker 3>Thanks a lot. Thanks. I'll see you at the trade shows.

00:26:10.035 --> 00:26:11.715
<v Speaker 4>Yes. See you. Take care. Bye.

00:26:12.115 --> 00:26:41.175
<v Speaker 3>Afternoon, we have the VP of product management for Agent Force Commerce, Lennart Stevens. He... We go through his background briefly cause he's new to the pod. Then he takes us deeper into all the cool new Gendic commerce features, which is our whole focus here at Retailgentic. And he tells us all about those aspects of storefront next, how it fits in with the the other storefronts that Salesforce has, and much more.

00:26:41.175 --> 00:27:01.210
Welcome to this episode of Retailgentic. I'm your host, Wingo, and I am joined by a special guest, his first time appearing on Retailgentic. We have Lennart Stevens. He is the vice president of product management for Agent Force Commerce at Salesforce. Lennart, welcome to the show.

00:27:01.210 --> 00:27:08.145
<v Speaker 1>Thank you, and thanks for having me. Excited to be on your on your show. I've, you know, been a follower for a while, so it's great to be on the show.

00:27:08.145 --> 00:27:19.025
<v Speaker 3>Awesome. Excited to finally have you on. Since you're new to the show, let's start with a little bit of background. We know we know you're at Salesforce, but let's talk about the before Salesforce times. Like, what did you do before Salesforce?

00:27:19.170 --> 00:27:40.315
<v Speaker 1>So I... I've I've been in the commerce space for for quite some time, about eighteen, nineteen something years at this point. I've kinda stopped counting, mostly on the partner side. So I started working with, you know, Demandware back in the day as well as, some of the other platforms, that that that kind of came and went and have had their journey. So I really see the world from the perspective of the customer.

00:27:40.315 --> 00:28:03.950
Right? I've always I've always represented what the customer is trying to achieve with their brand, and the platform has been just kind of a means to to getting to that to that achievement. So, you know, at some point, Salesforce said, hey. Why don't you come on board and, take some of that perspective with us to to to help, us move the product forward, especially from a storefront perspective? So that's what I'm here doing at Salesforce, bringing that, bring that customer vision to, to the front.

00:28:04.005 --> 00:28:09.765
<v Speaker 3>Cool. So were you at Demandware when Salesforce acquired Demandware and and then kind of rebranded it to Salesforce Commerce Cloud?

00:28:09.765 --> 00:28:20.010
<v Speaker 1>I wasn't at Demandware. I was a partner. So I was one of the early partners, implementing, Demandware at the time, but we also did many of the other platforms. Yeah. So I kind of followed that journey along the way.

00:28:20.010 --> 00:28:30.570
<v Speaker 3>Well, that knowledge was super helpful, I'm sure, as you've you've, you know, kind of moved into this role. The... And then at Salesforce, what all is under your your purview?

00:28:30.905 --> 00:29:01.305
<v Speaker 1>Yeah. So my purview, the main focus is gonna be Storefront Next, which is a collection of a few things. Of course, focused on the storefront itself, the experience that we're trying to build out, and, kind of bringing, bringing that to, you know, a new audience in the marketplace. Within that, there falls a a few other pieces, including, like, the transactional aspect, you know, considering with checkout, as well as, what we're doing for our commerce applications framework, right, making it a lot easier to install, plugins and extensions provided by ISVs and partners. So really excited to kind of bring this all together.

00:29:01.545 --> 00:29:12.025
It's really under this, this umbrella of of making making commerce quicker, providing more value upfront, and and allowing our customers really focus on the areas that they that they... That drives your business.

00:29:12.240 --> 00:29:19.920
<v Speaker 3>Yeah. Let's talk about that. So you guys just announced this yesterday as we're recording this. So congratulations. It seems like, I'm sure you've been working on a long time.

00:29:19.920 --> 00:29:30.755
It's good to kinda, like, get get... You know, birth the product, as we say. The... Help us understand where this fits. Is this like the next...

00:29:30.755 --> 00:29:44.750
Is this like the the artist previously known as Salesforce Commerce Club is now called this... Or not club, but cloud. And this is the next generation, or is there... Are there two things now? Help us understand the the...

00:29:45.150 --> 00:29:52.350
If I'm brand new to Salesforce, what are my choices in store platforms? And is this a new thing or separate, or, like, how does it all fit together?

00:29:52.395 --> 00:29:59.515
<v Speaker 1>It's a great question. I mean, it's it's... If you're brand new to Salesforce, it's it's the future. Right? It's everything that you're gonna see from here on forward.

00:29:59.915 --> 00:30:08.870
So it's not really a choice between the two. Obviously, we know we have a lot of customers with a lot of legacy that have been with us for for a very long time. So they have potentially other storefront versions. Right? We've...

00:30:09.110 --> 00:30:21.925
For those kind of familiar in the in the in the industry or in this platform, you know, your site Genesis or SFR or some of the other previous platforms that have existed. And those were just the storefront. Think of it as kind of the storefront templates. Right? So this is the...

00:30:22.005 --> 00:30:46.460
This is that new and latest offering on that, which is also then tied into all the work that we're doing for, agentic merchandising and some of the other things that are driving down total cost of ownership on a day to day, as well as allowing you to stand up your site much quicker. So, again, if you're new, this is this is really, this is the future of of where we're going. This is our investment into the product, to to take it into, the direction that we want it to be.

00:30:46.825 --> 00:30:53.705
<v Speaker 3>Cool. And it sounds like legacy customers were gonna have to kinda do go through a migration piece to get to this platform.

00:30:53.705 --> 00:31:12.350
<v Speaker 1>They don't have to migrate. You know, that's a that's a that's a good question that comes up a lot. We support our customers with where they're at. What we really want is our customers to be excited enough to say, hey, you know, maybe there's something we can unlock by taking this next step. Maybe we wanna, you know, launch a microsite or we've acquired a new brand and we wanna stand that up more quickly.

00:31:12.350 --> 00:31:31.200
And this is a great starting point for that. But it doesn't change where customers are today. They can continue to be where they are today, and we continue to support them and continue to build on on those road maps as well. But this is really the new directional change if you're looking at doing something net new. The idea is, you know, to be really excited about that and and and and weigh that as an option.

00:31:31.200 --> 00:31:48.035
<v Speaker 3>When you guys announced this, you had a pretty interesting framing where you you basically said, look. It used to be you had to choose. You had to choose Easy or you or or you had to choose, like, enterprise level features, and and you feel like this is kind of split the baby where you get both enterprise and easy. Talk us through.

00:31:48.355 --> 00:32:06.350
<v Speaker 1>Yeah, that's right. And I and I love that that messaging that we we kind of came up for it, which is, you know, there's there's there was a choice that always had to be made. We want to we want to do something really fast. Know, if a merchant or brand wants to do something really quickly and get to market, then they might be, you know, making some sacrifices in terms of the power extensibility of the platform. Right?

00:32:06.350 --> 00:32:23.475
We're obviously very proud of the architecture that we have underlying this, and that architecture continues to remain, you know, achieving our 100% up times in this last, holiday season, etcetera. Right? So we wanna maintain that level of of enterprise, enterprise and scalability. At the same time, there's a question of, hey, can we do something a little bit faster? Right?

00:32:23.475 --> 00:32:29.040
Can we do something a little bit more easier? Can we lower the TCO for day to day operations? The answer is yes. Right? It's twenty twenty six.

00:32:29.040 --> 00:32:45.925
Let's combine the two together. So we're taking, you know, this easy direct, way to to get, to get selling and then also taking the enterprise side of the house and and bringing them two together, to say, you can still have the both. Right? This is this is a challenge that that can be solved for, and we've done that.

00:32:45.925 --> 00:33:04.780
<v Speaker 3>What makes this possible? Is it a total reimagining from soup to nuts of how the platform works? Or, in the in the release, you also talk about it being there, you know, it's a headless kind of a a situation, so it's more composed kind of a a situation. What is it... What's the unlock that that helped you find that middle ground?

00:33:04.780 --> 00:33:14.465
<v Speaker 1>Yeah. The unlock is a is a lot of investment areas across the the the... We kind of broke it up into personas. Right? You've got the admin persona setting up the site initially.

00:33:14.785 --> 00:33:32.330
You've got the the the folks that are running the site on a day to day basis. Right? So that's merchandising, managing content, managing promotions, and things like that. And then we also have, you know, the focus on on on the infrastructure side. So what we've what we've done is we've leveraged what's...

00:33:32.330 --> 00:34:03.840
What we have that's a really great starting point. Right? We have this, you know, administration strength. We have, the back end enterprise, architecture behind it, right, surfacing up these APIs. And we said, let's put a new let's put a new layer on that that utilizes those APIs, right, and uses those APIs in kind of a headless way, which required us to revisit the APIs, make sure that they're in a place where they need to be, make sure that they're extensible and composable in the areas that they need to be, and then looked for unlocks to, kind of bring that together in a more simplified manner.

00:34:03.875 --> 00:34:13.475
Right? We had all the mechanisms there that can support the enterprise, enterprise structure. But what was needed was a way to kind of smooth that out. Right? Think of take...

00:34:13.475 --> 00:34:36.215
Turning it from a from a a code and configuration into a clicks type, type scenario. So that's what we did. So we reinvested into the into the strengths that we have and then brought them together into a way that you can now set up a new storefront simply by going... Logging in, creating a new storefront name, connecting it to your code repo if you want, clicking go, and it automatically sets it up. Sets up your code repo, sets up the storefront.

00:34:36.375 --> 00:34:57.920
And under thirty minutes later, you're clicking on preview, and you're on a storefront. And then within that storefront, we did a lot of unlocks too by moving away from this idea of that a a reference architecture is is just a reference. Right? It's just a template that looks a lot like a wireframe. And we turned it into something that's aspirational, something that truly converts, out of the box.

00:34:57.920 --> 00:35:30.125
It's made this high performance. And what that does is allows our customers to have a better starting point, but also allows them to focus on the areas that they want to build, way that they want to customize on that's gonna move their business rather than building out the basics for it. So, that's what we kind of did. We really focused on those areas of strength and then brought them together in a much cleaner and cleaner way. And then, of course, the storefront architecture itself has been updated to, to use the latest in in in modern technology, you know, React and Tailwind and SHATCN.

00:35:30.525 --> 00:36:02.525
And the reasons for that is those are really well, you know, agentically enabled, if you will. And and the the ecosystem of developers and those working on kind of web technologies understand it really well. So we've really opened up the network to who can be implementing on this platform and makes it much easier for, for organizations that might be more used to it, you know, a traditional kind of React application at this point or have an in house team that's working on it to be more familiar with the technologies.

00:36:02.525 --> 00:36:41.745
<v Speaker 3>And then when you say it's agent ready, this obviously is a podcast about e commerce, I know that's not what you mean. We can get to that in a minute, but at our company, because we were born in 2025, everything is an agent underneath the hood, we put a SaaS interface on it because our customers would be freaked out if they worked with the agents like we do, which are basically these little digital entities that we talk to. And is that... Is like... Are you basically taking the guts of the storefront and there's little agents inside of there kind of doing the the workflows, kind of a situation at this point, or you're you're you're getting towards that?

00:36:42.120 --> 00:37:28.205
<v Speaker 1>It's it's multiple areas. So, obviously, it fits into our, you know, our agent force strategy, right, whether that's shopper agent, but also, boosting the the merchandise productivity through AgenTic experiences, whether it's helping you set up promotions or provide insights into products or let you know if if something isn't fully complete yet. But it's also unlocking AgenTic, capabilities within the storefront when it comes to actually implementing setup or doing customizations. So that's been unlocked through providing an MCP server, a lot of investments into our CLI. So that's the kind of the command line interface that allows, you know, Claude or, you know, Copilot or someone else to have a a a place where they can access certain tooling, and it's discoverable through MCP.

00:37:28.605 --> 00:37:57.215
And then, of course, lots of, lots of different skills have been created for it as well. So, you know, 40 and counting. So when you're working with, you know, LLM assisted or guided development, it's now knows exactly where to go and can do that really efficiently and can do some really impressive stuff. In the, in in the initial setups that we were doing, you know, was handing this off to nondevelopers, and they were updating... You know, changing all sorts of code on the storefront and changing the PDP and PLP just through prompts.

00:37:57.215 --> 00:38:04.095
Right? And that's made available by surfacing surfacing this this code base through those those those Egentic abstractions.

00:38:05.240 --> 00:38:17.000
<v Speaker 3>In your... I imagine you had some design partners here and and and whatnot. You you haven't done this in a vacuum. And did you find that that was... People were hesitant to engage with that that new form?

00:38:17.000 --> 00:38:34.660
Looks like so different than how we've worked for thirty years, right? Did you... But I'm kind of addicted to it personally because that's how we interact with everything now. But I do think there's a curve there. Like, we have a similar functionality and we found some customers take to it like a ducksy water and others are what do I ask this thing?

00:38:34.660 --> 00:38:38.980
Like, I I don't get it. And so it's it's interesting to watch. There's a bifurcation kinda there.

00:38:39.060 --> 00:38:44.500
<v Speaker 1>It really is. Yeah. And we wanna meet everybody where they're at. Right? You know, this is not something you have to do, but this is an option.

00:38:45.140 --> 00:39:02.690
In the same way that, you know, from a... From an agentic future, we're also trying to meet the... You know, our merchants' customers where they're at. Right? Whether that's unlocking, through UCP, right, so through Google UCP or OpenAI or, you know, surfacing JSON LD on the storefront so that LLMs can crawl it.

00:39:02.690 --> 00:39:16.690
You know, we don't know where those customers are necessarily always gonna be or where they want to be, so we wanna make sure we have the places for those customers. And the same goes for those implementing the... Implementing it. We've taken kind of a BYO LLM approach, right, where you don't have to use Clod. You could use Copilot.

00:39:16.690 --> 00:39:43.940
You could use Cursor. You could use whatever you need. Or, you know, you can work with it, in in whatever traditional way you like. Some other areas that we've unlocked that are, you know, kind of adjacent to that is the commerce application framework. And this is a new way to do one click installs of things like tax and payments and, all of those, kind of standard ISVs that you normally would have to support your your merchant experience, to be able to click and install those.

00:39:43.940 --> 00:40:05.480
And that's that's done two things. A, it's made a lot easier to stand up a site quickly, but it's also made it easier for agentic tooling to understand, like, what constructs do I need to work with if I'm trying to build an extension. And now there's that framework there. So you point it at the spec, and you can point it at the application framework and have it build those those out much more quickly. So, it's many, many surfaces in which we've...

00:40:05.480 --> 00:40:49.075
We're addressing, where the future is going, but we also know we wanna meet our our customers and developers and partners, where they may be at and where their comfort levels are. So we're, you know, we're keeping that open. And then to address that one thing you said too, like, we've worked with design. So we've been working with, you know, a set of pilot customers and pre early release, early access customers, so from such as, you know, Colony Brands and Iceland Foods and some really big, some really big, organizations that have been, great to work with along with their partners during this process. So, you know, we're we're trying our own, our own paths and directions, and we've used that to help guide where we've, where we've ended up here with Storefront Next.

00:40:49.075 --> 00:40:51.795
And I think we're in a really great spot that we're we're all super proud of.

00:40:52.390 --> 00:41:17.795
<v Speaker 3>In the press release, they... Another interesting kind of framing for why now is it talks about how you surveyed ecommerce leaders, and the number one thing that came up, in their concern list was AI readiness. What what's that mean to you, and what are some of the AI readiness things that the platform delivers on? I mean, is that UCP and stuff like that? Or, like, what what else is in that list, I guess, is what I'm curious about?

00:41:17.795 --> 00:41:28.010
<v Speaker 1>I think it's a lot, and I think it's the persona you're talking to. And we try to talk to all the personas. You know, from from... Often from the merchant perspective, it's it's AI readiness when it comes to dealing with their consumers. Right?

00:41:28.010 --> 00:42:09.785
And that's that's when we're talking about the syndicated feeds, making sure your products are available in those LLM channels so they can purchase directly through it. Another area is, you know, agent for shopper, some of the guided selling experiences that we're doing, our investments in search, of course. So those are all, I think, areas that are really, really top of mind is how do we unlock these agentic experiences for our customers. And then I think second to that is the agentic experiences required in the day to day operation. So whether that's through merchandising tasks, you know, setting up new products, setting up content, doing all of those day to day things and really lowering the amount of effort and the amount of, kind of overhead spent on doing these tasks that happen, you know, day to day.

00:42:10.105 --> 00:42:25.160
And then, of course, the pieces we were talking about as well, which is the NGENTIC unlocks from a from a development and customization and business differentiator perspective. So we wanted to address all of those surfaces, and, it's part and parcel to, of course, the, entire Salesforce story.

00:42:25.445 --> 00:42:42.860
<v Speaker 3>Now let's talk about Agintic commerce. Wouldn't be, wouldn't be our podcast if we didn't do that. What, so you mentioned UCP, so I imagine this one's kind of got that in as a, a forethought and not a kind of add on. What are some of the other cool Agintic e commerce y things? How about at the at the at the web...

00:42:42.860 --> 00:42:58.595
At the at the website layer? I imagine it has an incredible agent, you know, powered by the agent force piece. Is that, is that, like, out of the box ready to go? And then any... Anything else on the web layer and then also, like, outside the the website layer?

00:42:58.675 --> 00:43:08.115
<v Speaker 1>Yeah. On on the UCPs piece, let's talk about that for a second too. And if we... Absolutely, it wasn't a forethought. And and just to be clear too, this is available for any customer on any version of the storefront.

00:43:08.515 --> 00:43:32.295
Whether they're going back ten, fifteen years, they have this available. This is a built in syndicated feed option. You you click a button and you and you export your you export your products if you're choosing and get started on that. And we've been in close, in in close cooperation with, of course, Google on not only getting that set up, but also helping them define the ECP feed. So it's been a it's been a great partnership, bringing that to the surface.

00:43:32.855 --> 00:44:03.045
From from the storefront perspective, of course, we do have, like we're talking about, our investments into search. We also have our AgenTic shopper, you know, helping helping customers, you know, communicate and talk to the the product buying experience in the way that they would, prefer to communicate. You know, not needing to search by, you know, keywords, but rather having a conversation saying, look, I'm planning on going a hiking trip. It's, know, Pacific Northwest. It might be a little bit rainy.

00:44:03.045 --> 00:44:09.620
It might be a little cloudy. What what do I need to have? What do need to get prepared for that? Right? And bringing that experience to to the customers.

00:44:09.620 --> 00:44:40.220
I heard something interesting, talking to one of our brands, and they said, you know, they were selling kind of, you know, like engagement and rings and that sort of thing, jewelry. They said, some of our some of our customers are more, you almost embarrassed to ask certain questions like, how big should the ring be? How much should it cost? And they're more inclined to ask an agent about that than they would be to necessarily always ask a friend. And that was kinda interesting to hear that, you know, there's some unlocks there to say we can be, you know, your adviser and your confidant through that through that commerce experience.

00:44:40.220 --> 00:44:43.100
<v Speaker 3>Any other cool, adjunctive key things we should know about?

00:44:43.100 --> 00:45:04.145
<v Speaker 1>I mean, there's a lot. We're doing a lot also in the merchandising side. You know, I was talking about helping me set up commonly used tasks. Some areas that I think are really neat is, you know, helping score whether something is fully set up, whether it's a product that's fully set up. I can't tell you how many times in my, you know, almost two decades of working with customers, there is a product that was launched and they weren't ready for it.

00:45:04.145 --> 00:45:30.735
You know, they didn't have the right description or they're missing something. And so just these little, these little areas really make a big difference, in in helping our customer. In addition to, of course, you know, doing things like surfacing real time insights and, in other areas that we're looking at. We're looking at AI in more ways than just, you know, kind of a chat interface, but using its intelligence to help us drive drive those experiences and and lower that TCO.

00:45:31.055 --> 00:46:02.435
<v Speaker 3>How about anything else interesting in the Merchant Center side where, you know, so you can kind of chat with your store basically and kind of set up some things through your merchant facing chat. Can you touch the whole back end that way? Could you check on, can you see all your inventory? Can you, you know, I'm at you kind of use the use case if you can edit pages and anything else in the Merchant Center side, like the merchant facing user interface side?

00:46:02.730 --> 00:46:19.695
<v Speaker 1>We're always looking to surface more and more. I can say that. Right? So, you know, we're starting out with those those, you know, if you wanna call it the top 10 tasks that everyone's working on and then looking how do we unlock and how do we surface more. I think an interesting area that's kinda tied into this is our investments into Page Designer.

00:46:19.695 --> 00:46:41.270
So this is our content administration piece. You know, previously, you had a lot of power, but you had to install it with the storefront and do some of those setup tasks. And now it's just there, right? It's just ready in there for whenever you need it. And some of those areas that we've looked at too are, you know, how do we have a collaboration between, you know, a Figma design or the Figma style guide and how do we pipe that into content?

00:46:41.430 --> 00:47:13.660
Right? So we're doing a lot of these integrations between our commonly used, tool sets and partners such as Figma to bring those in and to kind of stream that, stream that content creation process, back into into into Page Designer and making that more easy. So I think that's a huge area of unlock too, is just, knowing how much time is spent in those areas and how much time is spent in that design developer administration kind of collaboration cycle and really compressing that down into a singular default experience that's that that's already unlocked.

00:47:13.740 --> 00:47:23.265
<v Speaker 3>Great. Well, I know you have a tight window suit here, so we really appreciate you hopping on and telling us what's new. Where can people... Should... Do...

00:47:23.265 --> 00:47:32.940
Are you big on one of the social media platforms? Should people follow you on TikTok? Or... I'm just kidding. I'm imagining LinkedIn.

00:47:32.940 --> 00:47:34.140
Do you... Are you...

00:47:34.460 --> 00:47:43.820
<v Speaker 1>I've got a LinkedIn. Yeah. Absolutely. Reach out to me on LinkedIn. You know, reach out to your to your, you know, account team if you have some questions and they know how to get a hold of me.

00:47:43.820 --> 00:47:58.435
But you can always can always message me direct to LinkedIn. I'm always happy to talk, just shop and where things are going. And, you know, like I said, I spent a lot of time working with brands. I love to hear what they're doing. I love to see how we can how we can learn from that as well.

00:47:58.435 --> 00:47:59.715
<v Speaker 3>Thanks for coming on. Have a great day.

00:48:00.050 --> 00:48:00.930
<v Speaker 1>Thanks, Scot.

00:48:00.930 --> 00:48:03.810
<v Speaker 2>Follow, rate, and share to stay up to

00:48:03.810 --> 00:48:03.970
<v Speaker 3>date

00:48:03.970 --> 00:48:09.890
<v Speaker 2>with Retailgentic, your guide to the future of AI shopping agents.
