Shelley E. Kohan (00:03)
Hi everybody and thanks for joining our weekly podcast. I’m Shelley Kohan and I am so thrilled and excited to welcome back Janice Wang, CEO of Alvanon Welcome Janice.

Janice Wang (00:15)
So nice to be here, Shelley. ⁓ As we were talking earlier, this is my first podcast.

Shelley E. Kohan (00:18)
I love… Yeah, go ahead.

That’s very exciting that one that you picked retail on raft to be your first podcast but also you’ve been we’ve done a lot of webinars together and I love the brand and the company which we’ll get to in a second Alvanon on so today we’re going to talk about a couple different things but one is AI across the supply chain and I know you believe that brands have to invest in this foundational technologies today because that’s how you can really leverage

AI for trend prediction, product development, and consumers trust for tomorrow. So your whole philosophy is don’t wait for AI to arrive, which I think is great. And I think it needs to be a wake up call for our industry. The other thing that we’re going to talk about is TechFest 2025. I’m very excited. I love the fact that it’s free access for all. So hopefully listeners here will get advanced tips on how they can make

the biggest advantages from the Tech Fest. So let’s start first. I’d love for you to talk a little bit about Alvanon on the company and kind of where you came from.

Janice Wang (01:32)
Well, in industry, right? Alvinon got its name from Thomas Alva Edison. And when my father founded this company, he actually founded it as a tech company. However, we’re really known for these lovely forms that we have behind us. We’re known as a mannequin company. Now, the reason is, is that we’re very pragmatic people. We were born in 2001.

right after the tech bust, per se, the first ever one, your dot com boom and bust. And there’s a lot of pragmatism involved. My father’s original idea really was to be able to allow people to shop online and buy the things that they really wanted to buy and have them fit when you bought them. And you can imagine this was in his first…

thought process for this was like in late 1990s. So obviously we really didn’t have any of the tools per se. ⁓ And so his whole very long term philosophy was that this had to happen at some point, but there were so many different things that needed to change in order to get there that one had to figure out what was the easiest way in and actually that became figuring out what the sizes were.

and how demographics are different in different locations. Be able to set up a standard inside a supply chain before the brand itself and then hopefully at one point the technology would catch up.

and the use of technology would catch up in the industry. Because obviously about 25 years ago, and I can’t believe next year we’re 25, I really never thought that we would be trying to this at this point. But one of the things that happened was you can only move as fast as the…

Shelley E. Kohan (03:19)
Yay!

Janice Wang (03:33)
as the people who are using these things go along with, ⁓ this is very different today now with AI. Because we’ve never ever seen a situation where things are changing as fast as they are. And I think, I mean, in previous iterations, in the past 25 years, we’ve seen a lot, a lot of brands, right? ⁓ In previous iterations, you still had time.

you still were able to say, next season I’ll do this, and I have enough time to fix this, and go forward. But I think that is a little bit over now, because everybody’s moving at warp speed. So.

Shelley E. Kohan (04:16)
It’s so

true. The whole industry is moving very fast. And what’s interesting is when you think about just something as simple as sizing and you look at the U.S. our multi-general consumers here pose a lot of different challenges. And then when you look at globally, all the different parameters have to go into that. There’s really no such thing as like common sizing. Standardized sizing doesn’t really exist.

Janice Wang (04:34)
Yes.

No. No.

Shelley E. Kohan (04:45)
What I find really interesting about Alvinon is you really kind of address two things. You address a industry issue problem, but you’re also addressing a huge consumer problem. So tell us a little bit about that crossover.

Janice Wang (04:59)
So, you know, when we actually first thought about this, I mean, when my father’s original idea was you had to be able to get the image from the consumer and match them to one of the sizes that actually the brand actually had standardized, right? So the first thing we did, and we realized that there’s preference involved in everything. So if you strip it all away, the most important thing is to be able to match the consumer’s shape onto the size that you actually have standard. And then you can see where the differences are.

And then as a consumer, can figure out, well, OK, maybe I’ll buy that a little bit looser, or maybe I won’t. Or maybe I like them tight and I just want to have like a bandeau dress. So fine, you expect these kinds of things. But actually, at that point, you have to question, how am I getting these images of the person? How do I identify who my consumer is? And then how do I match that from that consumer part to the industrial part?

In doing so, there’s a lot of different factors, a lot of different technologies. And that’s why, actually, in the past 10 to 15 years, we’ve seen a lot of size finders come out. And to a certain extent, those have become commoditized, right? People hate using them. mean, have you met anybody, any woman who actually likes this? I don’t know what my busts

Shelley E. Kohan (06:16)
Yeah.

Janice Wang (06:21)
Measurement is I mean and I work in this I still don’t know what it is, right? I don’t know what my waist and where am I measuring my waist from? You know, it’s it there was a lot of factors that why some of these size finders posed challenges to the consumer themselves but so what we’ve actually figured out is with this advent of AI that there are going to be many more tools that you can actually recognize the shape of a person very easily and Be able to map that now, however

Over.

biggest issue is there is that as a brand or as a site that sells multiple brands, you have to have all of those standards of each of those sizes of each of those brands so that you can match it against that. And so that’s kind of the part that, you know, us having been in for 25 years in the industry being everybody’s standards and actually to a certain extent, you know, helping the technical and production people become the disciplinarian for that. It required somebody like us to

actually be able to help empower that process. So now at certain brands what we have is very strong, ⁓ probably ⁓ efficiency and technical expertise and a kind of disciplined workflow. But there’s a multitude of brands who don’t have that as well, by the way.

Shelley E. Kohan (07:46)
Yeah, I’m sure

there’s a lot out there that don’t have that.

Janice Wang (07:48)
Yes, I mean, and that actually is just a function of kind of industry movement and kind of companies in the way that they change. But then what now we also have is new technology, right, from the AI side that actually allows us to make those matching possible.

And if we think about that for an efficiency position, it actually changes a lot of inventory hold problems. ⁓ So the biggest issue is I would tell you that in most companies, the production people versus the design people versus the…

merchants who buy the merchandise, there’s a store manager, all of these things, they’re so siloed, right? So when an executive sits on top of them, they don’t probably see all of the tiny, tiny, granular problems that happen on a day-to-day basis.

And there’s a lot of unintentional consequences, right? So the intention is that we want, know, everybody wants to make the perfect product for the perfect consumer. No such thing as perfect consumer, no such thing as perfect product, right? Especially not at the $29.99 range. We’re not talking about Couture here, we’re talking about, you know,

Shelley E. Kohan (08:52)
Of Of course.

Janice Wang (09:07)
⁓ lots of inventory out there that doesn’t get used because there’s the multitude of issues that are surrounding this. But I think with what we’ve seen in the past two years, now we can use a lot of the tools that are out there in AI to be able to analyze various parts of the whole demand and supply chain. And this will allow us to kind of become much more efficient, right, as an industry.

Shelley E. Kohan (09:36)
Absolutely, and when you talk about so there’s the excess supply that’s at the end of a season and that’s not good for our environments, not good for the companies. And then you also have a lot of the online businesses that are really about selling online fashion. And we know that online fashion has a tremendously high rate of return. And the number one reason is sizing. So if you are able to fix those, just those two things.

Janice Wang (09:41)
Yes.

Yes.

Yes. Yes.

Shelley E. Kohan (10:02)
⁓ you could make a really dynamic impact on the industry as a whole and sustainability.

Janice Wang (10:08)
Yes, and well that’s the hope, right? I mean that was the original premise. The fact that we had to wait, we’re waiting 25 years to do this, right, is, well, thank goodness and we thank all of our clients for actually supporting the view, long-term view that we’ve actually had, but it is actually good business sense. This is about top line and bottom line. And actually for the longest time, brands have chased only top line.

but you can’t anymore.

Shelley E. Kohan (10:40)
It’s impossible. So let me ask you a question. So we talked a little bit about, you know, the brands and they’re having this, you know, proper sizing and using AI and all that. A lot of companies are trying to like forecast trends for next season, two years from now. So how does the, how does your technology help with trend forecasting?

Janice Wang (10:53)
Yes.

So trend forecasting is a, look, 80 % of most of what you make is actually core product. 20 % is your fashion product. Trend forecasting is going to work on that 20%.

Shelley E. Kohan (11:09)
True.

Janice Wang (11:14)
One of the key things is what we actually work on is actually figuring out, this is consumer demographic that you actually say is your loyal consumer. How do you take all of that information and actually be able to use different systems and parse that information that actually will allow you to buy the right amount for the next season? And actually, what does your sizing hold? If anybody’s still using that kind of one, two, two, one, whatever matrix that they are using.

Shelley E. Kohan (11:39)
You

Janice Wang (11:43)
Let’s first figure out what is that 2 to begin with in your medium, right? Because your medium, if it was at one point a size 8 or a 10, but you’re actually selling in places where the medium is more like a 14, then what is this matrix look like? And I think sometimes…

People assume that this is being looked at and analyzed by all the people in the company. But I think when you get down to that granular level and you start to actually look at what is the matrix that they have bought, is it a historical matrix or how does it change? And actually, I think a lot of people don’t recognize that if you change that matrix a little bit, there is a vast cascade of problems that happens at the production level as well.

So if, for instance, I’m about to actually make more so that my skew actually looks like one, two, three being in an extra large, the costing on your garment is going to change. So there are all these cascading problems that I think that… ⁓ I think…

Shelley E. Kohan (12:50)
Right.

Janice Wang (12:56)
The people working at the working level and using it at that level can raise flags going, hey, things are going to change. But when it comes up to when you’re actually at this level, when you’re actually looking at consolidated information and financial information, you don’t really see those nuances anymore. But actually the granularity matters. Now, if you layer, if you were to layer, and there are many people working on this problem, and it’s not just us, we require all the collaborators.

in each one of these sectors to actually make for a better efficient kind of demand chain, supply chain, right? So those are the things that you want to kind of look at as a company to say, okay, trend forecasting part, let’s see what the trends may well be for that 20%. Do we change the size matrix there? Are these trends more suited towards what kinds of bodies, right, and how do we buy there?

then

how do we actually adopt that to kind of product development? Do these things change? They do not change. This is your standard, right? It just stays the same. But now you can actually baseline off of that and say, in this sector, I’m going to have more of this. In this sector, I’m going to have less of this. And so it allows for a more dynamic shift of ⁓ kind of the inventory hold that you’re holding for that. And that translates directly into bottom line, right?

Shelley E. Kohan (14:26)
amazing. ⁓ the other thing that you’re a big proponent of is being able to use what you call these digital foundations. So understanding you have size sets and product libraries and data standards. Can you talk a little bit about what that means today and maybe how retailers can really use that to unleash the power of AI?

Janice Wang (14:46)
Right, so size sets in any executors view, it’s the domain of the technical design department or the production department. It doesn’t look, it haven’t looked at it from a inventory hold standpoint and actually in a financial standpoint of actually how that works, right? So that’s one thing that can be immediate. It’s actually an immediate ⁓ act to the bottom line.

So that’s one thing. And this is circular to a certain extent. You start doing this and then you start standardizing. So you start standardizing then a pattern library on top of these. So then you can actually say, when this new trend comes in, it says, I have to have a bigger drop shoulder, right? I’m not changing this base pattern of the armhole. I’m just dropping the shoulder.

So, and you can use a lot of kind of iterative generation AI to actually look at that, to say, okay, there’s this new trend and it’s hot pink, let me visualize what that looks like, let me put that back onto my foundational kind of pattern library, foundational bodies. You’re actually speeding up the product design, product development.

and with some accuracies, without some accuracies, but at least people can picture what they’re about to make, right? So then, therefore, that product development cycle will kind of scrunch. You scrunch that time as money, right? If…

Shelley E. Kohan (16:20)
Absolutely.

Janice Wang (16:21)
if for instance they were holding fabric on the other side for you and just dying immediately. Then you actually have even more of a tight crunch. So you’re actually making much closer to the market than the 18 months out that we’re holding today. So ⁓ that was a very long-winded answer for actually what the question that…

Shelley E. Kohan (16:43)
No, it was great. let me ask. when so best sellers, let’s talk about best sellers for a second. So every company knows its best sellers. They go through, they can give you here are the 50 best sellers in the past 10 years, whatever. But do we, can we now really look at why from a design, you know, technical perspective, why these are best sellers? Can you find commonalities about the consumer base?

Janice Wang (16:51)
Yes. Yes.

So a lot of this has to do with how well they’ve got that information that they’ve got. How well they mine kind of the consumer loyalty side. Consumer loyalty for the longest time has actually just been about what you bought before and what price point you bought for and your email address. That’s gold mine of stuff in there, right, that you could use to actually be able to give the right things to the right people. And actually you see there’s a few different kind of products out there.

days and that actually can help that streamline a lot of that kind of ⁓ consumer insight.

Shelley E. Kohan (17:45)
That’s great.

Janice Wang (17:47)
And so if you layer on kind of having, if you layer on a very disciplined baseline, you can actually see the difference between them. And then you analyze the delta, right? So I mean, part of this is it sounds like, it sounds really common sense. Actually, I would say that a lot of people would say, wait, are we not doing that already?

And what you find is that because the data has been so unstructured for so long, the problem is that two years ago you couldn’t actually, you couldn’t structure that data. But today,

Shelley E. Kohan (18:16)
That’s right.

Janice Wang (18:26)
We can use tools to restructure that data and I’m sure that the business analysts inside these big companies are actually looking at that and saying, how do we use AI to restructure some of this data? We put them in buckets. And I’ll give you the most dumb kind of example of anything. Is that 25 years of our photo library database. In…

Shelley E. Kohan (18:52)
Excellent, yes.

Janice Wang (18:55)
20,000 places which we don’t even know where this stuff is. ⁓ And a meaning of our photos of kind of, know, people who used to work with us, know, factory photos, you know, what lindu we used to use, all this kind of unstructured data. So we’re like, but next year we’re turning 25. Really, maybe actually we should.

We really have to actually go and thank all the wonderful people who have been with us for so long. So we’re like, ⁓ but don’t know where all the photos are. Got to go find all the photos. Immediately, right, you can use a very quick way of actually saying, let’s shove them all into one place, run an AI on top of them. could see me and Jason and John from 2003. And ⁓ yes, of course, we look exactly the same.

Shelley E. Kohan (19:45)
Of course.

Janice Wang (19:46)
Of course.

But yet it can actually identify us in a picture. And so then you’ve actually got, and it literally can actually pick up all the people who used to work. Once we typed it, who those people were, immediately, boom, you have a bunch of structured data that you can use for whatever use it is. But this is about cultural mindset.

everybody in the company itself and I think that you we have been significantly lucky that everybody who works with us and inside our company has this culture of let’s look at what tools are available, recognize they are tools, break things, break things that we used to think were kind of the

know, workflow processes that we used to actually have and say, how can we make things faster? How can we give ourselves two hours back of our day, right, to do important things like go hang out with our kids? Who doesn’t want that? And I think that’s the part that, that, that.

Shelley E. Kohan (20:46)
you

Janice Wang (20:56)
gets overwhelming sometimes if you’re somebody who works in the workflow process is that actually all I want is my two hours back, but if you tell me that I’ve got to learn something new or do something new, then it changes the things that I used to do. A little bit of scary place. But now with all the tools that AI provides, mindset has to change to say that actually don’t have to be scared of this stuff anymore. It’s going to make me…

happier because I can go do the things that I want to do. So let’s use…

Shelley E. Kohan (21:28)
I love that Janice.

I love that philosophy. That’s great. And you know, since Alvinon was born, you’ve always been tech and data driven from day one. And so that kind of drives through. So I think it’s great. One big key takeaway that I take from everything that you just said is retailers and brands have been collecting all this data for the past 15, 20 years. A lot of them don’t know what to do with all the data, but now…

Janice Wang (21:39)
Yes.

Yes.

Shelley E. Kohan (21:56)
you can unleash the power of that data in a different way. You can go back to the archives or the searchable libraries and really make a tremendous impact on the future.

Janice Wang (22:06)
Yes. But I think all of this is about being the bridge, right? The bridge mentality that has to be there is that, yes, we can get this done.

Shelley E. Kohan (22:10)
Right.

Janice Wang (22:17)
And, you know, that’s kind of how we’ve always thought about it is because there’s always been so, look silos are not new things. They exist so that we can do our work effectively. But now we’ve got to bring all the silos together and actually be able to bridge all of these gaps. And that requires for everybody to have a different type of mindset going forward. ⁓ And I think that AI actually helps us do that as long as we are not

Shelley E. Kohan (22:25)
But. ⁓

Janice Wang (22:47)
scared. And what I find ⁓ is that a lot of companies haven’t really looked into the processes to see how can AI actually be applied in a very pragmatic matter, right, that gives you that two hours back.

⁓ And from our point of view, we know kind of where the, you know, which brands are actually going to be able to cross this divide and which ones aren’t because they also have kind of discipline on the bottom. So it’s really a mix of kind of discipline plus tools, discipline plus mindset. ⁓ And I think that’s kind of where ⁓ our whole ethos have always stemmed from. So.

Shelley E. Kohan (23:36)
That’s great, I love it. So before we leave though, we have to talk about your fantastic 3D Tech Fest, because I think you’ve been running this for five years and it’s coming up very soon, October 9th. And so let’s talk a little bit about that. I love the fact that you have opened it up for all and you say that in order to improve our industry, we have to be open about what we’re doing, right?

Janice Wang (23:43)
Yes. Yes.

Yes, so first things is that we took away the 3D part of Tech Fest. So this year is going to be Tech Fest because actually I think people got fixated on this idea that 3D is a system that we have to use and we have to focus on and actually there’s pattern making on top of that and that’s actually a whole new subset of…

Shelley E. Kohan (24:02)
Okay.

Janice Wang (24:18)
what are we to learn. It never actually was that way, but it was just kind of the feeling that actually came along with it. And so, this year, what we’re going to focus on is all the new technologies that are available that actually help you make, become much more efficient. So we’re gonna start with the design process. And one of my great friends is gonna, from Karl Lagerfeld, gonna actually, yeah, Hun is gonna kind of actually draw a couple of things and see how he uses that to actually, how he uses AI to

to create kind of new ideas and also be able to show them to merchandising, right? And actually, so that the concept can actually become reality very quickly and they can say, oh, this is really what I’d like. Then we have, you know, people who are working in, with the technologies in the factory. How do you actually use AI inside a factory session to actually…

take all of the information that’s coming in, fill out all those kind of tech packs that work properly, and then actually ⁓ move that onto a factory floor, and how those kind of joins will actually happen.

We’ve got some other people who are going to talk about, you know, how do we actually culturally change people who are a little bit scared of all of this. But mostly, also demand forecasting. So demand forecasting is a huge thing that you can use AI for. And so there’s technologies that you can use for there. And actually, by the way, we are technology agnostic. So it doesn’t matter what you use. This is about, you know, whatever works for you, great. But there are a lot of systems out there and there’s a lot of help out there. So

We’re just going to introduce those concepts into part of that. And also, how do those unintended consequences cascade into other people’s workflows? And we always wanted this to be open and free. listen, I think I’ve said this multiple years back on TechFest, is that we always send people of certain levels in the company to go to conferences.

And that’s awesome, but you YouTube exists, right? So you can learn anything, but it needs to be a little bit curated and we all kind of should use this as a talking point inside companies, say, okay, so we’ve had this kind of learning event that actually happens for three and a half hours or four hours. What do we take from this and how do we apply it? Because how does an application help the company afterwards?

Shelley E. Kohan (26:44)
Yeah, and the other thing Janice, I’m sure you recognize you’re doing this, but by giving access to people that may not be at the top, but are in the middle levels or entry levels, you are like giving these young executives appear into how they can lead their future in their companies.

Janice Wang (26:52)
Yes.

Yes, it’s also, young executives and also people who are even trainees or even we’ve had people, know, lot of students join and I love it when students join because I think they need to see the reality sometimes of industry. ⁓ And for all the educators who are out there, we are very, very, very happy that everybody uses this inside classes. And so, you know, part of our mission has been to improve the

industry in whichever way that we can empower the people who are inside this industry. ⁓

And we just hope that this is why we actually, we do so, we teach a lot and we, and we, you know, make sure that students actually know a lot of things because we have to provide some hope for this next generation that comes in here says, no, AI is going to do everything. No, AI is not going to do everything. It’s going to do what you tell it to do. Right. And so let’s please tell them things that actually are useful and helpful. Yeah.

Shelley E. Kohan (28:05)
I love that. And

Janice, I have to say for you and Jason, a big heartfelt thank you for all of your support with students and university studies. I know at Syracuse University, they obviously use Alvinon at Fashion Institute of Technology. It’s a staple in many, many of our courses. And it’s so funny, I had a student come up to me, I think it was last semester, and said, professor, professor, I want to do my project on inclusive sizing and there’s this great company. And I’m like, yeah.

So yeah, really thank you for your contributions to the future generations.

Janice Wang (28:42)
Well, I really thank them for entering the actually, because we really need more makers, and we really need engineers and people who are thoughtful about what’s to come.

Shelley E. Kohan (28:58)
Absolutely.

So is there any closing thoughts you’d like to share with our listeners?

Janice Wang (29:05)
I’d like to say that not to be scared of new technology and things that are coming down because it’s there to make us better. And please, if you actually really think about it, have some discipline in the baseline.

make sure that actually, you know, your fundamentals are in place as brands and as companies because once you have those in place, building stuff on top of it is much easier. wherever you can do that, and this may be just in like list writing, know, wherever you are in this supply chain, in this demand chain, really examining those things and moving forward is, you know, the way it’s built.

Shelley E. Kohan (29:49)
Love it. Thank you, Janice. And I’m sure when we’re in our podcast and the write-up, we’ll put a link to how that people can register for Tech Fest 2025, because I would love for all of our listeners to participate in that this year.

Janice Wang (29:57)
Kiss.

Yes, it would be wonderful. Thank you so much for today, Shelly. See you soon.

Shelley E. Kohan (30:07)
Thank you, Janice.

And my favorite thing that you do is you bridge supply, demand gaps in retail media, advertising a very, very hot topic for retailers and brands today. Yeah. Yeah. Thanks so much for having me, Shelley. This is, uh, this is great and very exciting and a big fan of the podcast. So thanks for being to letting me be on.

Oh, absolutely. It’s a pleasure having you here. So retail media networks, I know we’re gonna talk about that in a minute, but you know, Macy’s Media Network, which is their retail media network, just did a partnership with Amazon Retail Ad Services, and that’s gonna help. Fourth quarter, and of course Walmart has [00:01:00] generated, I think it’s over a hundred million in revenue from retail media networks, from their Walmart Connect, which is responsible for get this, Dave, 12% of the company’s profits.

So when I look at these numbers and I understand the direction retailers are going. You probably have a pretty good point of view about the missed opportunities that may be out there for them. But before we jump into all that, tell us about omics and what is mood. Sure. Yeah. Uh, so, so. Uh, mood Media is a fascinating company.

Um, it’s, I think 60 years old. Uh, the original Elevator music, the Mu, that’s where the company was created and its origins are in, uh, the, the notion of experiential retail experiences. Uh, and, and so what Mood Media does is, is this, it’s the largest provider of in-store or retail based experiential solutions.

We do everything from. Big complex audio and visual build outs and retails, uh, retailers. You’ll see, you know, 20 foot [00:02:00] tall LED screens that are sort of a very rich experience. Mood will do that all the way down to like designing the speaker and the headset that goes into a drive-through. Um, so audio, visual scent.

Uh, if, if, if you’re of my age, you shopped at Abercrombie with blasting music and that really strong smell, that’s actually a designed experience, uh, that, that Abercrombie picked. And so Mood Media has, has sort of had several iterations. Uh, the most recent is that they were bought by a private equity firm called Vector Capital right after COVID.

Uh, and they realized that this retail media advertising opportunity was becoming a very real solution, that they wanted to bolt onto their suite of offerings that they built. And so they acquired a company called Omics, uh, omics, specialized in. Helping retailers figure out how to turn their inventory in store into a digital inventory pool that buyers could buy and sell the same way that [00:03:00] everyone else has done in programmatic.

And so omics really pioneered the idea of turning, uh, an audio spot or a visual spot into an individual number of impressions based on who and how many people are shopping in the store. Uh, and so they were acquired three years ago. Uh, I jumped on about four months ago after the founder decided he’d sort of gone through the process and transitioned the business and so they were looking for, to bring somebody else in.

And so been here for four months. Um, and it’s been fascinating. It’s been a lot of fun. Oh, that’s amazing. And I know that you have done a lot of work in this space. You’re kind of a rock star. You came from Fiber, right? Yeah. And you were Chief, chief Revenue officer there, and I think you grew it from a hundred million to 500 million, something like that.

Yeah. I mean, look, you know, so the, the quick background here, and I tell this you all the people this all the time, um. It’s really good to be very lucky about where you wind up when you wind up there. Um, I started in the notion in this whole idea of programmatic, before it was [00:04:00] called programmatic, it was called RTB back then, and we traded our first impression in February of 2009.

Um, and, and so. Like at the time, the market size was the value of that one impression that we traded. Uh, fast forward, you know, almost 20 years now, and it’s uh, you know, well north of a hundred billion dollar market, uh, most media is traded at least in part programmatically. Uh, and so what I’ve been very fortunate to have is a front row seat for how money used to go from hand sold through relationships and phone calls and iOS, and now traded through machines.

And I, I’ve had a lot of, uh, a lot of really interesting experiences. I’ve done it in the mobile app space with fiber on the sell side. I did, uh, demand side work with moloka, which was the most recent company I was at, which was a heavy machine learning based approach to, to programmatic. Um, I was at, I was running an online video business called a OL on after we sold a business called ible.

So I’ve done online videos. CTV, mobile app. Mobile web, [00:05:00] desktop web, and now retail media. And it’s fun because there are some things that are always different about each of these markets. But at the end of the day, the way this technology gets deployed and the way money moves through the ecosystem tends to follow very similar paths.

And so what we’re seeing in retail media right now is this maturation where. Programmatic is becoming more and more important and a lot of the, you know, sort of timeless rules around programmatic apply in the same way. It’s just that they haven’t been done before, and so we get to take a lot of what we learned in the past and apply it here and it’s, it’s been a lot of fun.

I. Yeah, I mean, let’s see, about three years ago. I do, I always do my top retail trends every year, and so about three years ago my, one of my top retail trends was retail media networks, but that was back when it was just a revenue generator, so retailers coming outta the pandemic looking for another source of revenue, they really kind of bought into this idea of getting revenue.

But this year. I kind of changed my, uh, retail trend and I call it retail media [00:06:00] empire because now they’re really what they, they’ve gone beyond just the source of driving revenue to actually evolving this retail media ecosystem that’s brand building, it’s partnership attracting, it’s an analytics gold mine.

And from my understanding, this whole idea of retail media networks is gonna account for 25% of total digital ad spending by 2026. Which is just next year. So, but you also, uh, it’s interesting you had mentioned that, um, tariffs are accelerating retail media strategies. So tell me a little bit about that.

Yeah, well, so it is interesting the, the, the. The KPI that I keep in mind when I think about market sizing is the revenue of a, of retail media networks compared to the total gross market value, the total transaction volume of products sold in the store. And you know, we heard a great data point coming out of the Ascendant Network event last week where.

It used to be that if you were at 0.1 to [00:07:00] 0.3% of total gross market value, right? So if you sell a, you know, a hundred dollars of of product in store, you’re selling 30 cents of stuff. That was sort of the, the benchmark. That number in a lot of cases has now crossed over 1%. And if you look at the margins that are associated with selling products in the store, we’re talking about low single di digits.

But if you look at the margins coming out of retail media, you know, oftentimes they’re well north of 50%. And so from a profit driving, uh, opportunity, retail media. Is like one of the most investible opportunities in the market. And the funny part is, you know, if these were independent companies, they would be being funded by VCs and Growth Capital left and right.

It’s just they live within, within these retail media groups. And so, you know, there was a great quote by, um, by Quentin George from McKinsey where he talked about like, if your CEO is not jumping up and down about the opportunity with retail media, he’s missing the vote. Uh, and so I think there’s a [00:08:00] really interesting opportunity here.

I think there are two trends that, that, that are in line with what you asked. Um, the first is that COVID scared the heck out of every retailer and, and there, and every, and every advertiser, every brand that sold in store, and there was this panic rush into, I have to be building my business with an e-commerce shelf strategy first because the vast majority of products will be sold.

Online and I need to be able to figure out how to merchandise myself in a digital environment. Don’t get me wrong, e-commerce is is really important. And if you look at the customer base by segment, e-commerce has an increasing importance, the younger you go. That said, when you talk about retailers. On average, 80% of products are still sold in store.

If you talk about grocery particular, it’s over 90%. And so what’s interesting is there was this, the bit of a whiplash for these poor retail media networks where it was like build, be an [00:09:00] e-commerce strategy. Everybody needs to build, sell products on digital shelves. And now the pendulum swinging back because what they’re realizing is that all the investment that’s gone into driving e-commerce sales.

Is nowhere near as profitable because you’re still having to sell products in store. And so there’s this pendulum swing back to in store and that that’s really gonna drive, I think 2026. You know, we talk about it, we think about retail media. 1.0 was all about onsite search. Onsite display. Retail media 2.0 was about accessing those data that those users, offsite social, CTV web, retail media 3.0, which is 2026, we think.

Is all about accessing users in store at the point of purchase. And so there’s this really interesting dynamic where brands are starting to figure out the right balance between those strategies. Uh, and so we’re seeing a massive amount of, of capital investment from the retail media networks and in particular the retailers and store experience folks in building out these better shopping experiences [00:10:00] that also help highlight the brands that are ultimately responsible for the sales volume that go through those retailers.

That’s amazing. And the other thing, Dave, I just wanna point out is that, you know, you talk about the profit obvious, you know, 70, 90% margin on the retail media networks revenue, but also it’s driving loyalty. It is driving deeper loyalty, which feeds that, you know, customer lifetime value. Yeah, right. You know, it, it, it’s interesting, right?

Because mood media does not have a background in advertising or retail media. They’re much more about experiential, uh, shopping, uh, shoppers. There is an opportunity. For retailers to fundamentally rethink what it means to provide a shopping experience. And you know, we’ve seen dozens of examples where retail experiences are more popular than like big concerts.

You know, if you remember back in the day when Nike opened their Nike flagship stores, there were lines around the, around the block to get into those stores. [00:11:00] If you think about the, you know, one of my favorite, one of my favorite Buildouts is we, we did this, uh, this build out for. The Ferrari location in the lawn, and there’s this massive LED screen, uh, that people can step into and feel like they’re inside of a car.

There is often an hour long wait list to get in there. Now granted, Ferrari is a brand that everybody wants to go experience, and, and there are those things, but the fact is that if I think about in a grocery store, those big physical pallets that have, like, you know, in the summertime it’ll be a Coors Light sponsored pallet and you’ll have some cooler and some sort of like beach thing.

There that’s possible to turn into a digital experience. The, the, the, the, the tech exists, and frankly, the cost isn’t that much more than what you’re already doing. Well, Coors Light is more than happy to create and be associated with a very cool execution in store. And so is it like, is it the same thing as like, you know.

A massive, you know, 20 tall, tall, uh, LED screen like we have on Victoria’s Secret Store on Fifth Avenue. No, but you can [00:12:00] absolutely create these experiences and now that brands have shown a willingness to invest alongside these digital experiences in store retailers have the opportunity to fundamentally rethink what it means to bring their brands to light in those stores.

I love that and I love that. Um. That story about the Ferrari, because you’re right, people just wanna go and experience. And same when you’re in the grocery store, I think people are kind of getting a little bit, um, weary of all the LEDs and all the signs and all the digital, uh, billboards. And so whatever retailers and brands can do to kind of bring more of a experiential to that shopping journey is great.

Do you have other, um, data stories where retailers you think are missing a return on investment? Um. Well, yeah, so I think, look, this is, it’s, it’s not so much a, a miss. It’s part of the evolution, you know, so look, we, we power, um, a dozen or so retail media networks for their in-store [00:13:00] advertising experience, end to end from hardware all the way through, um, measurement.

And what I can tell you is that the, in what I’m seeing is very common for businesses that try and convert to a programmatic or. Basically stepping into the modern way of monetizing media. You know, I’ll give you my favorite example, because Disney is, is, I think, has done a phenomenal job, but the first five years of Disney’s transition from linear to streaming was incredibly bumpy.

Yeah. You know, if you think about it, and, and there’s a lot of parallels between the streaming and the, the TV space and the shopper marketing space. First majority of money is allocated on an annual basis. It’s not spot market based. It’s very, very heavy upfront plan. Second, there is a high concentration of inventory among a very few number of players.

There are six broadcasters in the US that basically reach everybody. If you look at the, the grocery chains, there are probably five or so grocery chains that [00:14:00] reach all of the us. Um, and so there’s a lot of, lot of consolidation and, and not much fragmentation. Um. Yeah. The third thing is these are businesses that are going from a linear or analog advertising model and trying to convert to a digital model.

What most re uh, what most publishers, if I can consider, retailers a publisher in this example, what, what most publishers miss is that they, they take protectionist attitude. Toward the inventory and try and create these strict business rules that look and feel like the business rules that were possible and held up their business on the analog side, but don’t work for digital.

And I’ll give you an example. Um. If you are a, a, an A brand who’s focused on buying from, uh, from a retailer, you have a joint business plan. That joint business plan often locks your inventory into, okay, I’m doing the, the trade marketing for couponing with this much budget. Then I’m gonna do my shut, my, my end cap digital or, [00:15:00] or, or physical print signage at this budget level.

Then I’m gonna do this tactic at this level. Well, that makes sense when you have to have a 30 day lead time to get posters into a store. But when you have the opportunity to change the impression on the fly in real time, those business rules don’t work. And so what we’re seeing is a lot of these retailers are now taking a much more omni-channel approach and coming to the brands with a much smarter question, which is how much money do you have to spend and what are your specific objectives for next year?

And you can imagine that if you’re Pepsi and you’re trying to push sell bottles of Pepsi, that’s, that’s a very different goal than trying to sell cans of Poppy. Poppy’s, a challenger brand doesn’t have the same brand recognition appeals to a different audience. So your objective with Poppy might be increased market share, increased total sales volume, and make sure you’re driving new customers to purchase.

Right? For Pepsi, it’s lapsed purchasers. It’s hitting your loyalty people. Well, the fact is that all of the inventory and all of the data that someone [00:16:00] like a Kroger or an ahold have make those strategies very, very possible, but they’re different. Yeah. And so what we’re seeing is less about like, Hey, go buy your onsite search two months later.

Okay, now come buy your, um, your, your offsite display two months later, okay, now come by, you’re in store. Instead, they’re approaching that upfront, that JBP negotiation from what’s your objective? And here are all the assets we have to make that objective come true. How do we map to that? And so once that’s done and you have flexibility of the budgets within them, then you can really take advantage of things like programmatic, of things like machine learning to drive technology, that use technology to drive activation of media, where it’s gonna make the biggest impact.

And so it’s just a part of the journey. But at the same time, what I will tell you is that this is so new for so many retailers. That there’s, there’s a bit of hesitancy and there’s a bit of trepidation around like, well, yeah, okay, that’s all great, but I still have my a hundred million dollars shopper marketing budget over here that I don’t wanna say [00:17:00] no to.

How do I do this without putting this at risk? Yeah, it’s so true. And retailers are kind of funny. They love technology. They’re not the quickest to jump on technology, but once one or two success stories happen, then the whole industry runs towards that technology. So you’re gonna see an avalanche coming up here.

Um, the other thing that’s interesting about what you said is that there seems to be more like flexibility, real time flexibility to change the in-store marketing experience plan. Yeah. As consumer trends are bubbling up, is that something that you would agree with? Yeah, absolutely. And I think, I think what’s, what’s really fun, frankly, about my job is that we get to stress to stretch the creativity of these retailers.

You know, one of the things I’ve been blown away by is the level of depth at which the average retailer understands the shopping experience. And I think what’s what’s been interesting is that we bring a unique set of capabilities. And we’re [00:18:00] really good at those capabilities, but I don’t know the shopper the same way that a Kroger or a, a Lowe’s or a Dick’s Sporting Goods does.

And so when I bring these, these, these assets to bear, the, what the retailers come up with is incredible. Uh, and you know, I’ll give you a great example there. There’s a fun little stat that, that we, we’ve pulled out. We’ve now run this study a dozen or so times. If you look at online advertising and combine it with.

In store audio. The results of those two things being run together is around 60% better than in just any one of them on an individual basis. If you run audio and visual in store, in conjunction, the recall is four x higher, right? So like all of these things that we, we know we can do audio, we know we can do visual.

It took somebody from a retailer telling us like, how do we time these things together? So that we can effectively do a store takeover and the results were [00:19:00] unbelievable, right? And so I, we, we love to see how these retailers can use their understanding of the shopper understanding of their store, the physical real estate, and then bring these technologies to bear and create a unique experience for the brands.

Oh, that’s, that’s amazing. And I love the fact that you now, you have tons of data and you can actually come away with analytics from those, that data real actionable things that you can do. Um, tell me a little bit about what retail media can learn from the mobile gaming data exhaust model. Yeah, that’s really interesting.

Yeah, so talk about getting lucky. I, uh, after we sold Fiber, um, I left and joined a company called The Loco. And you know, every once in a while you come across those people in your career that just fundamentally change your perception of how the world works. And Moloka was that, you know, to give you some context, when I joined Moloka was, uh, about 200 people.

When I left, it was about 800. When we were at 800 people, we had over 150 PhDs in machine learning and data science. This is sort of the gold standard [00:20:00] for how you think about machine learning. And I got to engage with these guys all day every day and, and some of the things that that stood out to me are that machine learning, especially with the deep neural network based models, not the LLM stuff or the generative ai, but really the sort of pattern recognition machine learning.

Is so much better at understanding the correlation between multiple data points at a given time, that really what you want to do is generate as much data and then figure out the value of that data as it relates to your end outcome. So for example, there is, you know, I’ll use Instacart ’cause it’s relevant in our grocery world and also one of the best apps out there.

If you think about what it takes to become a customer of Instacart through the app, you have to download the app, you have to sign up for it. You have to browse web products on the site, on the app. You have to add products to your cart. You have to check out, you have to come back, you have to add products to the cart, you have to check out, you can build lists.

All of those data points are [00:21:00] all indicative of what my ultimate LTV will be for Instacart. And so what’s interesting about the app world is every app developer basically tags every single one of those data points and then feeds it back to their partners for whom they use to buy media. And so if you think about this world, you have loyalty card data that you have, email address, phone number, physical address.

You know how often they’re purchasing at your store. You know what they’re purchasing, you know what their total shopping cart value is. All of that data is right now being used and packaged into these like static segments. So if you’re fruit loops and you want to conquest tricks buyers, you can do that.

And that’s great, right? But the fact is that there are nuances between people who buy fruit loops and people who buy tricks, and oftentimes it’s more than just what did they buy last time? There are probably different things around income, location, number of kids in the household, attention to health, whatever.[00:22:00]

All of that data being fed into ML models can ultimately be the thing that determines when to serve somebody, an ad and how, and, and what the objective is gonna be. And so if I look three or four years down the road. That to me feels like the, the ultimate holy grail where I’m a brand, I, I wanna sell, you know, a million dollars for the poppy in this quarter.

I put $800,000 into the machine. The machine figures out which ads to serve, when and where within the store, how to time it with audio and visual together, uh, and then spits out the result to tell you how much got sold. And so, like the, the part that’s really exciting about this space is that there’s very little development of machine learning in this space.

In the in-store space. Uh, but there’s a tremendous amount of data that’s very reliable, very predictive, and very associated with future outcomes. And so that to me feels like one of the natural places this market evolves. Oh my gosh. That’s like a big [00:23:00] goldmine there. Yeah, it’s pretty cool. Yeah, so my other retail trend that I know, uh, listed out for 2026, the evolution of AI powered retail analytics, and you just beautifully explained the power behind that.

So, I mean, it’s, it’s one of those things, Dave, it’s one of those things. My, you know, one of those things I think about if you, if you buy media from Facebook and Google through their AI tools, right? Um, it’s an interesting process because there’s a lot of trust. They basically say, gimme some creative assets.

Tell me your objective. Get outta my way that like that’s what makes a dollar in equal a dollar 30 out. And I think these retail media networks are in a position to offer similar products across their entire PO portfolio of offerings. Yeah. David, I think one thing you nailed on the head a little while ago is retailers are brilliant about knowing their consumer base and their customer.

And so if you take that knowledge with all this AI powered analytics, you can really. [00:24:00] Develop a program that is going to not only resonate with their customer, but continue to drive that loyalty. Yeah, absolutely. Always a pleasure. I hope you’ll come back. Thanks, Shelly. Appreciate it. Happy to. Thank a great day.

Thank you.