The Digital Shelf Institute’s Lauren Livak Gilbert - Impact of AI in Winning On The Digital Shelf

Lauren Livak Gilbert (00:00)

Hi everyone, I'm Lauren Livak Gilbert, Executive Director of the Digital Shelf Institute, and you're watching the CPG Guys Podcast.

PVSB (00:07)

Hello and welcome to the CPG Guys Podcast. I am your gregarious co-host PVSB. I also moonlight as Head of Industry Engagement at Flywheel Omnicom's Commerce Acceleration Division. My co-host, he is the father of pop stars Rhea Raj, who is now the face of both Toyota EV and MAC Cosmetics. And of course, Katseye's Lara Raj, the fans call him.

Papa Raj, that makes a lot of sense. He's also co-founder of Think Blue Consulting. Please join me in welcoming the man known as Sri Sri. So this past Sunday, after getting back from California, it was good to see you in LA. Nadia and I were watching the Disney Legend Award show on Disney+. And they were recognizing a woman named Susan Egan. She's a singer that made famous, she performed Bell on Broadway, and and she was also.

Singing the iconic song I Won't Say I'm in Love, which was from the animated classic Hercules. when who should come on stage to offer their rendition of it? But this brand new global girls band called Katseye. That's right. So my question to you is how how proud did it make you to see your daughter Lara and her band celebrated by the biggest consumer brand in the world,

The Walt Disney Company. That must have been a bucket list moment for you.

Sri (01:31)

Beyond. We were just at as you know, we were just on Fallon. Then we gotta go see both daughters opposite recording rooms at iHeart for different shows. Then I gotta trail Rhea to Vogue. I got a trailer to the Today Show. I mean, I can write a little thesis here, Peter, but we're lucky that these two are our kids.

And we're barely warming up. I already know what programming is in store for fall for Rhea what the next big brand announcement, which is gonna happen in the next two weeks. I I don't know what to tell you, man. We're living it. I'm going through the emotions every day. I'm honored that these two are my children. And I'm also honored that Bala came into my life.

PVSB (02:13)

Can't forget Bala She is at the top of the proverbial Raj Mahal chain of of hierarchy. Your dog is the center of the universe for for everyone in your family. but thank you, Sri as always, for joining me to our audience. If you're getting value from the conversations and bickering that Sri and I are wont to do on occasion, we have three quick asks. First, make sure you're subscribing to the CPG guys on whatever platform.

You're listening right now, whether that's Spotify, Apple Podcasts, Amazon Music, or anywhere else you get your podcast, hit the subscribe or follow button so that you never miss a new episode. Second, leave us a rating. It takes just a couple of seconds, and it makes an enormous difference in helping other CPG and retail media professionals find us. That's what drives the search order. It's not fundamentally different than a product detail page,

we built this community one listener at a time, and a five-star rating is the single best way you can help us keep growing it. it's like when you think of pop music Sri, it's not four for fighting, it's five for fighting, is you like to remember. So five is the number. Smash the five. And third, make sure you follow us on LinkedIn between episodes. We publish a lot of thought leadership, industry analysis, and content. You're not gonna find anywhere else.

All built for the CPG and retail community, they're already a part of. We got great stuff like the retail media minute, our buddy Jeff Malmad produces every week. a lot of great content. So make sure you search CPG guys on LinkedIn and hit follow, or just click the link that's in the digital show notes of this very episode. Subscribe,

rate, follow. Three simple actions that help us bring you more of the conversations that matter. Sri?

Sri (03:54)

Did I hear Peter I might have accidentally eavesdropped on a conversation that the CPG guys had is there new programming coming sometime in the fall and November?

PVSB (04:03)

Well, Sri, the CPG guys are getting ready to debut a second podcast platform.

Sri (04:08)

Whoa.

PVSB (04:09)

It will be coming up this November. Now I'm not gonna we'll tease it a little bit, but it's gonna be a different focus. it's gonna be more on kind of the up-and-comers, the the emerging brands and service providers and ad tech companies.

really the entrepreneurial stories and what it takes to scale something from a concept into a real business. But more on that shortly, Sharia. Absolutely. But let's get to our guest, shall we? Generative AI is rewriting how shoppers discover, evaluate, and ultimately buy products. And the digital shelf, the brand spent a decade optimizing for human eyeballs and search algorithms is now being read, summarized, and recommended by machines.

Today's conversation digs into what digital shelf readiness means in an AI-driven world, how brands should rethink content for AI intermediaries and agentic shopping assistants, and where AI is already moving sales versus where it's still more hype than actual results. Today we are joined by someone who has spent her career at the intersection of digital commerce content and design and who now leads

The organization defining what best in class digital shelf strategy looks like. Lauren Livak-Gilbert is the executive director of the Digital Shelf Institute, a global commerce community advancing the industry through research, content, and education. She spent years owning the digital shelf for J&J's Consumer North America business and worked with dozens of global omnichannel brands at Salsify before joining DSI.

Where she leads original research. And she also co-hosts one of our very favorite podcasts, Unpacking the Digital Shelf. If you want to understand how AI is about to reshape shell space and share of search, you really want to listen to this episode. Lauren, welcome to the CPG guys. How you doing?

Lauren Livak Gilbert (06:09)

I'm great. Thank you so much for having me and for being a fan of the Unpacking the Digital Shelf podcast.

PVSB (06:14)

Well, I think you should

tell people, not only are we fans, where did you get your very first piece of unpacking the digital shelf swag? Did you get it from Salsify? Did Piyush send it to you, or did you get it from someone else?

Lauren Livak Gilbert (06:29)

It was from you, Peter. I got in the mail several t-shirts that have unpacking the digital. So I appreciate that.

PVSB (06:33)

I was I was very upset that you are not wearing

your own swag and I'm like, Sri, we've gotta do something. And Sri said, Well, why don't you just order her some swag for herself? And so we did.

Lauren Livak Gilbert (06:44)

I love it. And you'll be happy to know we made t-shirts and we are gonna be sharing out those t-shirts at DSS

PVSB (06:47)

Thank you.

Lauren Livak Gilbert (06:50)

Europe to some of our fans. So thank you for the push. I appreciate it.

PVSB (06:52)

We love that. We love that. That's good.

Sri (06:56)

I'm

wearing CPG guys. When will this say DSI?

Lauren Livak Gilbert (07:00)

No. I'll send you one, Sri. I will send you one.

PVSB (06:59)

well I I got one for myself, Sri. yours must have gotten lost in the mail. Must

Sri (07:04)

I want

the Lauren Livak Gilbert original one, not the

PVSB (07:06)

Okay. You want authentic.

you don't want the pirate merch? You want the authentic stuff? Okay. All right. All right.

Lauren Livak Gilbert (07:13)

And Sri and I go way back too because we worked together at J&J, which is a long time ago.

Sri (07:17)

J&J before it was called all

these fancy things like Kenvue and whatever else it's called now.

PVSB (07:20)

I know the artist formerly

known as J&J Consumer North America.

Sri (07:24)

I remember

Lauren Livak's first day coming into digital.

Lauren Livak Gilbert (07:28)

Do you really? I don't even know what day that was.

Sri (07:31)

I got introduced

to her that she's going to be the one focusing on the digital shelf.

PVSB (07:36)

We're gonna get more into that in a

Lauren Livak Gilbert (07:37)

Yeah.

PVSB (07:38)

second. I wanna say to our audience, please check out the digital show notes of this episode. You're gonna find a hyperlink to Lauren's LinkedIn profile, to the DSI website. check those out as we carry on this conversation, which I know you're going to enjoy. So let me get this, ball rolling, shall we, Lauren? How, from your perspective, is generative AI changing the way shoppers actually discover and evaluate products online? And is search

even the right term we should be using anymore.

Lauren Livak Gilbert (08:05)

we're in a wild time in our industry, right? Like it's a really exciting time to be in e-commerce, to be in commerce, because there's so much changing. But the way I like to break it down is you now have to think about three different shelves. So you had the traditional physical shelf, then the digital shelf came in about 31 years ago or so. And now you have the agentic shelf. So you have three different shelves as a brand that you need to worry about. And each of those shelves are

Catering to different people and or robots, right? So from a physical perspective, you're talking to a human who is looking for emotional content and wants more visuals. From a digital shelf perspective, you're talking to an algorithm who is looking at search and keywords and is matching. And then from an agentic perspective, you're working with an agent that's computing and reasoning and is looking at content in a machine readable way, which is very, very different.

From how we've created content and built out the digital shelf for the past 31 years. So, to your question about search, I like to think about it as for the physical shelf, you're focusing on the human who's thinking. For the digital shelf, you're focusing on the algorithm that's matching. And then from the agentic shelf perspective, you're focusing on the agent who's reasoning and computing. And each of those buckets have a very different approach.

Sri (09:26)

Now

let's talk to brands. the CPG guy's personal belief: brands are not ready. Because they've actually done a pretty good job of getting ready for SEO brands. You and I worked on this 10 years ago at Johnson & Johnson. It was a focus area. And I think brands have optimized pretty well, and then retailers have received it and have built great ecosystems. Now we're talking two different parameters here: AEO and then GEO.

Do brands even know what those things mean? And then now I'll layer something on top of that. Maybe this is a two-part question. Do they know what that means? Are they getting ready for it in your experience? And then with AI-powered shopping assistance, chat-based commerce emerging, how should they rethink the product content to really get ready for these three phases? AEO, GEO, and then agentic commerce.

Lauren Livak Gilbert (10:15)

Talk about a loaded question there, Sri. So let me let me start with the first part. I mean, I think that the challenge is most CPG companies, most brands in general are still going through digital transformation. So to your comment about them not necessarily being ready, I do see that there's a big gap, especially around data, right? Because it takes a long time for a large organization to figure out where does their data live? What are the systems? Who owns it?

And a lot of people are still tough stuck in that digital transformation world where we're now entering a gentic transformation that's happening in parallel to that digital transformation. So brands need to finish what they were doing before while adding this additional element of being AI enabled and getting ready for that world. So to your question about like AEO GEO, I think brands know what it is because everybody's talking about. They've heard it, they understand the definition.

Some of them, the more mature brands, are actually taking action. So I work with brands day in and day out, and a lot of them have adjusted their strategies and they're thinking about how to build out their content playbooks and what machine readable content means and what to tweak here or there on an Amazon or a Walmart. But I think the challenge is that they're not actually fundamentally changing the way they work. And I think that's the gap. So when I think about AI in general.

A lot of organizations are just tacking it on to a process that was already broken. So they're like, Hey, let's add AI to step three, thirty-nine, and a hundred and fifty. And yay, we're scaling and it's becoming more efficient. Well, at the end of the day, you didn't really change anything. What you really need to do.

Sri (11:53)

But they're taking credit

for it though and Lauren. Publicly they're saying I'm ready because they've done exactly what you said. You've taken a half broken process and you bolted something on and all of a sudden you're ready.

Lauren Livak Gilbert (12:04)

But they're not seeing the ROI, and that's the reason for it. So the word that I have keep using a lot is just reinvention, where I I use this term, like the clean sheet of paper approach, where you just have to say, get everybody in a room, put up a whiteboard and say, what would this look like in an ideal state? And then back it in from there instead of starting with, hey, this is broken and this doesn't work, and this person doesn't talk to this person. Like that's not effective. And it's not going to move you forward, especially in this next era.

Sri (12:33)

And then Lauren, if that's the current situation today, what are we gonna do about Agentic Commerce, which is setting foot upon us now?

Lauren Livak Gilbert (12:42)

I wanna define agentic commerce a bit. So I think the industry hasn't really decided on a definition. All of this is new, right? Nobody's really an expert. We're all kind of figuring it out. My definition of agentic commerce is fully autonomous shopping, where an agent is purchasing something on your behalf. I think there's a spectrum of agentic commerce that is on the very, very far right of the spectrum. We are kind of, I'm air quoting, we're kind of there.

We're mostly in the agentic discovery phase where consumers, shoppers, all of us are using LLMs, using agents to ask questions, discover products. And then we're probably going to our retailer of choice and purchasing it. Yes, there are some things with Google and what they've launched and other tools where you can have autonomous shopping, but I do not believe the trust is there yet from a consumer's perspective to have full agentic commerce.

But the way I look at it, Sri is it takes a long time to move a cruise ship in the ocean. And most of these organizations are large cruise ships. And so even if we're not in a world where autonomous shopping is happening today, it's going to take significant change in these companies to be AI enabled, to change your strategies, to make sure your content is machine readable, get all of your cross functional partners on the right page. And there is a first mover advantage, and we've seen that.

There's a lot of categories that are up for grabs if you think about ChatGPT searching specific products. And if brands can get in front of this and really focus on what matters, they will have that first mover advantage.

PVSB (14:22)

So Lauren, I was talking with the head of digital commerce for a very well-known global consumer brand the other day, talking about digital shelf readiness. And I think what he said to me was, Listen, I'm still trying to make sure that my thousand plus items in my catalog get

All of their content updated equally as frequently across every marketplace that we're sold in, rather than just the top 20 items across the top 10 marketplaces. If I could just get parity, it would help the tail of my industry. that's kind of like very rudimentary, but still very important objectives. So my question to you is now we're in this world of AI-driven.

consumer engagement. My question is what does digital shelf readiness look like now compared to what we're still fighting with, which was kind of the demands of five years ago. Just trying to get the basic block and tackling of content available.

Lauren Livak Gilbert (15:25)

I mean, the truth is to the point of that digital leader, we're still there. I think the difference in my answer today versus before is that I used to say, okay, 20% of your SKUs drive 80% of the value for your organization. Focus on those 20%, make sure they're best in class. For the rest, have like images and a description, but you know, don't really focus on them necessarily because you don't have the bandwidth. That is not the world we live in anymore because LLMs are

Also looking at your long tail products. So now digital shelf readiness means every single SKU is best in class. Period. And that from a resource perspective is very challenging, especially if you have not built the processes and the teams to be able to support that. If you don't have a PXM, if you don't have a source of truth, if you don't know where your data lives, that's not even possible. So I think that's the challenging conversation I have with brands these days, where it's

that you need to have this perfect content. And then from there, you need to optimize on top of that. Where it used to be maybe a hundred skews, it could be 2500 skews that have to have full enhanced content and videos and lifestyle images. And that's a really big challenge.

PVSB (16:34)

Lauren, you used a term that I want to make sure we cover with our audience, PXM. Can you explain to them what PXM is and why that's important?

Lauren Livak Gilbert (16:43)

Yeah, so product experience management. So think about all of the information about your product that needs to be shared with all of your retailers, from the images to the enhanced content to the supply chain specs to all of the context for LLMs and agents. It's your source of truth for all of your content for all of your products in one place. And then that gets shared out to like an Amazon, Walmart, Lowe's, Home Depot, all of your channels.

Sri (17:11)

Awesome. So I'm gonna now jump into the other side of the equation that without which we can't really sell product, which is the retail side. So Amazon's got its own wall garden, Walmart got its own wall garden, but the rest of the retailers so far have not built wall gardens. They're dependent on Chat GPT, they're dependent on Claude, they're dependent on Google Gemini, yada yada yada. But I'm pretty sure each of them is on the

verge of building their own tool for their app, their website, things of that nature. So are their own AI tools? Examples could be retail media search, generative product QA, creating new data, visibility gaps that now brands will have to start thinking about, my gosh, how do I, you know, like in the old world you have to create Amazon content, Walmart content, rest of the world content. Now do they have to create it for ChatGPT, for Claude, for Gemini, for Albertson's?

For AD USA, for Northeast grocers, heck who knows.

Lauren Livak Gilbert (18:09)

I think it's an excellent point. And I agree with you that retailers are probably all in the future going to have their own agents because if you think about it, so I'm actually presenting at a conference next week to regional grocers and been having a lot of conversations with them because I want to understand like where they're coming from and what their fears are. And obviously their fears are are consumers still going to come to my website? Are they still going to come to my store? How do I be that selected store when they're purchasing their product?

And if they're going through a ChatGPT and they're not being scraped from a content perspective, what's another way to drive them there? Well, that's an agent on your site and a loyalty program. And so I do think to your point, Sri, there is going to be a gap in terms of the data that retailers currently have and what brands need to provide because retailers are all saying, Hey, pick me, pick me, pick me. I want to be the one that you're purchasing my product at, but I haven't

Focused on my user experience. I don't have enhanced content. What are FAQs? Like, how do I go through this motion? And if I think back to my days from a digital shelf perspective, I remember very early on in my digital shelf journey, Costco was asking for content that was different than any other retailer site. And that was really challenging for us at the time. But that's a request that's coming from a lot of retailers more and more because they want to have the leg up.

So to answer your question directly, Sri, I do believe that there will need to be a specific strategy for Amazon, Walmart, Chat GPT, Claude, Gemini. But I do still think it takes this strategy. I like to call it the lowest common denominator approach, where you can look at the content that's required for your top 10 retailers and you can satisfy almost all of them if you really kind of build out a process and kind of define what that looks like. But Adobe just put out a study recently that said.

Agentic commerce and the agentic shelf is going to require 5x more data than brands have ever created. Honestly, I think that number is conservative, and I think it's going to be even more. And brands need to be ready for what that means, where they store it, how they get it, how they syndicate it. And we are a bit far from a reality where that is possible.

Sri (20:21)

And Peter's gonna jump into the economics next, but I want to point something out, Lauren. So if brands have to create 5x the content, the issue today is they haven't even reached a hundred index on existing SEO. And then I'm gonna put money on it with the exception of a few CMOs and one we know very well, like Ben Vatere from LiquidDeath. or Nick Hammett from

Newell brands, these guys are on top of what they're doing, with with AI. And they're certainly preparing for that volume of content. How does the average brand even fathom this? You gotta fathom it first before Peter gets to the economics element. How do you even tell a CMO this is what you're up against? A robot that can multiply. I'm thinking of the term the Terminator thing now.

Lauren Livak Gilbert (21:06)

Yeah, there's so many movies that like can depict this moment. I mean, what I say to them is that's why you need to launch agentic transformation and you need to be an AI-enabled organization because you cannot do this without AI. It you just it is not possible. You can't add 15 more people. We've never been able to do that. But I also try to flip it on. That's why this is so exciting. You know how many times I would have loved when I was sitting at J&J to have an AI agent who could help write the initial version of titles when I had to pay.

X number of dollars per hour for a contractor to sit in a room and tag things and physically write out ingredients. it's so exciting because we can finally do all the things that we've been talking about for so many years, but it does change your perspective and how you approach what your organization looks like. We're entering this hybrid world where humans are going to be managing agents and working alongside agents, but companies need to be comfortable with that.

And that means there's governance, there's approval, there's legal, there's everything involved in this.

Sri (22:05)

But hold on, Lauren,

that very word governance legal approval, right? And the level of risk aversion that large market cap brands are, this is gonna be an Achilles heel 'cause you don't you're not gonna get

Lauren Livak Gilbert (22:16)

Yes.

Sri (22:16)

sixty days to create the program with the romance copy and thirty levels of approvals you need up and down the chain. What do you do in that scenario?

Lauren Livak Gilbert (22:25)

But you're already see well, I I agree. This is what I say. I say, who's your biggest competitor challenger brand? They created that content in two days and it took you 200. The at the

Sri (22:35)

Lauren.

Lauren Livak Gilbert (22:35)

end of the day, that's the kind of catalyst that I always refer back to, and they're already feeling it, right? all these challenger brands are nipping on the heels of these bigger companies.

Sri (22:46)

I'll leave

it at this statement. I think if you're a large market cap brand, e-commerce was a choice. Being in SEO was a choice, and y'all stepped up to some reasonable extent. I'm not gonna give a single one credit for being an overachiever. But brands stepped up. It was a choice and they exercised the choice. Some didn't, some exercised it halfway, some exercised seventy percent of the way, and some have fallen behind. But the penalty you paid was deep, but it was not death.

Lauren Livak Gilbert (23:12)

It's not a choice anymore.

Sri (23:13)

But in the

AI world, you'll be invisible. You'll just be absent. You'll just be vanished from the ecosystem because AI search results give you five, not fifty like hundred like the store shelf or fifty like the Amazon page when you search on the search bar. That's my it's a warning to brands. I'll move on,

Lauren Livak Gilbert (23:31)

I agree.

Sri (23:31)

Peter. Economics, please.

PVSB (23:35)

Lauren, have I mentioned to you Shri's building an AI assistant to replace me as the co-host of this podcast?

Lauren Livak Gilbert (23:41)

no.

PVSB (23:41)

That th that's got me worried about the economics.

Sri (23:45)

the only

problem, Lauren, why I haven't triggered that is he opens every episode calling himself like the Gregarious co host. So how do I replace Gregarious with a robot? It's not possible.

PVSB (23:54)

You you refer to me as garrulous Sri. I don't know if that's anything.

Sri (23:58)

I'm

I'm practicing the vocabulary, remember?

PVSB (24:00)

you are you are working on the vocabulary, but let's I i the the thing that's the thing you mentioned 5x.

Sri (24:03)

You are no longer a Ludite He's the accoutrement on the show, by the way.

PVSB (24:09)

You mentioned the 5x the data. That is actually you know, in in the current zeitgeist is all these concerns people have over data centers.

how do you accommodate 5x the data requirements when you can't even get the town you live in to agree to build a data center? There's so much to unpack here.

Lauren Livak Gilbert (24:32)

So much complexity.

PVSB (24:33)

let's talk about how AI is actually changing the economics and workflow of content creation. Copy, imagery, A plus content.

at scale across thousands of SKUs. and honestly, how agencies and people in our space are fundamentally having to change how they bill out the work they've done. you know, advertising for years is predicated upon FTEs in seats in offices. And that's something that they've clung to.

And now it's being thrown up in the air. There's an enormous amount of tumult. What are your thoughts on the economics?

Lauren Livak Gilbert (25:14)

the big word that comes to mind is value, right? Like busy work used to be rewarded and showing numbers and and saying, hey, this was the output, but now it's about value. Like what is the value that you are driving for me as my customer, if you're on the agency side? What is the value that you can drive for the business? And when you when you talk about like the economics of AI and content and processes and workflows, I like to say the math doesn't math when you look at it today.

here's a perfect example, right? Let's say, for example, I'm making this up. You have 3,000 SKUs as a brand. And best in class, I'm air quoting, is to update your content four times a year. That does not happen for most brands today, but you know, best in class for most categories, it's even faster, but let's go with four times a year.

Sri (25:58)

She said four times a year.

Lauren Livak Gilbert (26:00)

Four times a year, let's go with that. It's honestly it's faster, but

Sri (26:02)

Am I allowed to smirk?

Lauren Livak Gilbert (26:03)

I'm being generous. You have one channel that's 12,000 updates.

To your to that retailer channel. Let's say it takes 10 minutes, let's say, to update that. That's 2,000 hours for one channel. For context for the audience, there's 2,080 working hours in a year. I had 32 channels when I owned the digital shelf at JJ. That is an insane amount of hours. If you're talking about with human resources, how you actually create that data, the math.

Does not work. It cannot be done without AI. And that's where the processes and the workflows really come in. And to our conversation in the beginning, you actually need to fundamentally rethink the workflow. It's not just tacking on these things and fixing these little broken parts. I work with this brand who's very digitally forward. And it was a really great conversation when I was talking to them. And they, their content team, their digital shelf team, is actually three people and six agents.

So they have an agent that does the briefing. So works with the marketing team and understands like the claims and all the information and what they want to target for the new product. That briefing agent then briefs the content creation agent who creates both the content and the imagery. That agent then briefs another agent that has synthetic focus groups for their specific personas to test out how that content will work. And then it goes back

to the human in the loop to review it. And then there's an agent that's actually pushing that content out live. That is rethinking your process and your workflow versus tacking on AI. And that's the reinvention that needs to happen at every single level of the organization.

Sri (27:53)

Wow,

I'm almost now smirking and scared to ask the next question. So this is what brands are gonna really thump their fists about. And here's the question. you guys measure success in this space. So are there any early use cases in digital shelf analytics that we have where we can say, hey, it's moved the needle on sale, share any other metric that a bronze ones, which is and so that this is not regarded as just AI hype.

Lauren Livak Gilbert (28:22)

I have seen some organizations be able to quantify it and say the impact on creating AEO content and focusing on our brand.com and citations and things like that, they've seen a lift on other retailer channels. The challenge with why you can't measure it, or there's a couple different reasons. Number one, it's very, very new. so we're all kind of figuring it out. Like early days of e-com when we were like, we know it's important. Please listen to us, right? So there's a little bit of that. I think the second piece is.

The data isn't being shared from like a Chat GPT or a Gemini or even necessarily from Alexa shopping. So we have some inclinations of things. And then the third being that most organizations can't actually tie together all of their data X AI, right? Like there's a lot of people I ask questions to at brands where I'm like, okay, do you know if this content impacted this channel? And they don't have their data aggregated to be able to do that. So

In test and learn environments and in specific pockets, I have seen a lot of data from brands showing that it is impactful. But the challenge is how you connect that to the broader business. The one example that I really love that I think shows the impact with AI visibility and AEO in general is there was a brand that had an influencer talk about their product, and they were like incorrect in some pretty bad claims about.

what the product did and and what it didn't. And the brand had a social listening tool and it saw what was happening. And they were like, okay, how do we fix this? We can't tell the influencer to take it down. They weren't someone that was affiliated with the brand. So they went on their brand.com and they created FAQs and blogs to say, hey, this isn't true. Here's all the scientific information. Here's the claims like here's all of the correct information about the product. Within two days, Chat GPT was then sorting the sourcing them

As the expert from a brand.com perspective versus the influencer. So to me, that's very, very strong and very, very impactful in a world where brands want control of their story and their voice out in the world. So that's one example I really like to point to because you can see the impact when you're focusing on AI visibility.

PVSB (30:38)

It's great story on authorities and how information can be distorted and you can work to correct that. so thank you for that. let me remind our audience that today Sri and I are speaking with our dear friend and industry luminary, Lauren Livak Gilbert. She is the executive director of the Digital Shelf Institute.

now the question that Sri and I get from so many brands and even retailers in the industry, right? Which is how should brand and e-commerce teams be restructuring roles or skill sets as AI takes over more of the manual digital shelf tasks? What do I do about my organization?

Lauren Livak Gilbert (31:23)

This is my favorite question, just because I spend a lot of time talking about org structures and I've gone really deep with a lot of brands around this. The first thing I will say is that the differentiating factor for brands in this era of commerce is going to be how they're doing the work and how they're executing. Because every brand now has access to the same tools. Everybody can get Claude, everybody can get ChatGPT, everybody can get insert LLM here, right?

But how you're actually executing that inside your organization and how you're structured is the differentiator. And I do believe that goes very clearly to your organizational structure. Now, you do not have to completely blow it up and change it, but you do need to take some very strategic moves. The first one that I've seen to be really successful is putting an AI task force in place. Sri, I think back to like our days at J and J when the digital COE came out because

Digital was new, and so you needed someone to focus on it, and it needed to be kind of pulled out and then eventually got democratized back into the business. Creating this AI task force gives accountability to someone in the organization to pay attention to this and brings in cross-functional partners. It has to be cross-functional because otherwise you might have like 200 AI pilots happening and you're spending all this money and nobody's talking or sharing learnings. So

From an org structure perspective, that's a great way to start. Have someone accountable, have an AI task force, and make sure that's set up so you're sharing learnings cross-functionally. The second piece I would say is rethinking job descriptions. So I think it was, was it JP Morgan that put out that study that said like 80% of most people's job most people's jobs will be automated in the next X number of

Sri (33:10)

Yep. It was JPM.

Lauren Livak Gilbert (33:12)

years? Yeah, there you go. So

That study came out and everybody everybody was scared that they were gonna lose their jobs. But I actually think it's exciting because it takes away a lot of the hands-on keyboard work that none of us actually liked doing. So when I'm working with brands, what I encourage them to do is look at your role, look at your team's roles. What are the things that can be automated? And then what are you going to use that time for that's more strategic or more cross-functional? Or maybe you can actually go talk to your retailer partner or go talk to your wholesale team that

needs to learn digital and educate things around the broader space, right? So rethinking job descriptions on what can be automated, what can be taken out, and how can you create more of that collaboration. And then the third most critical piece, which was always critical even before AI, is creating shared goals. So Sri and I lived this. If the sales team and the marketing team have different KPIs and our their bonus is tied to a different goal and objective, why would they work together?

So organizations need to think about what gets measured gets managed. And if you do not have shared goals across your organization, everybody's not marching towards the same North Star. And especially in a world where every single second counts, every single interaction from a brand perspective counts, when you're talking about the consumer, you all need to have the same shared goals. And if you have not implemented something like that, it's gonna be really hard to win.

PVSB (34:41)

I look forward Lauren to the day where I walk in and just like George Jetson at Spacely Sprockets, I just sit down at my chair, push that one button, and my job is done for the day. for those Gen Z and and younger, I suggest you just Google the Jetsons, you'll understand. Sri.

Lauren Livak Gilbert (34:57)

Well everybody

always uses up to is is is it is it up where they're all like r like really big and they're in chairs and they're in space and they don't walk around? Like are we gonna become that? I mean, I think that's a little dystopian, but you know, another reference.

Sri (35:11)

I'll

be back! Well, that's

Lauren Livak Gilbert (35:13)

Ha ha ha.

Sri (35:14)

a that's it, Lauren. Let's go to a complete different piece of this conversation now. We've been very rah AI gotta do brands. Now let's look at the side where AI does commit hallucination, some misinformation, maybe poor product claims, maybe a drift away from traditional old school brand voices, or maybe there's a time to rethink how brand voice is written. So when I when I put all of these in the mix.

As AI generates more customer-facing content, are there very early rules y'all have thought of for your listening to brands or s socializing or advocating for in the ecosystem?

Lauren Livak Gilbert (35:52)

Yeah, I think it's an interesting topic for a lot of different reasons. One being there's a lot of regulations that have come out recently around disclosing that you have AI content or AI was used. So Amazon actually requires now for you to check a box that says this is AI generated or the model was AI generated or you used AI when creating this imagery. And then LinkedIn has that new feature now. You can mark it as AI slop. TikTok has one.

I think YouTube also has one. There's a lot of new tags that are kind of coming out around AI created content. And the reason why I bring that up is because I don't know where this is going to go as an industry. I think we can swing this the pendulum way too far where everything becomes AI generated and then the differentiator is non-AI generated content. So I think you just need to be careful as a brand and need to be very intentional about where you use it and how you disclose it.

But you also need to have social listening and monitoring to the story I just told around what people are talking about for your brands And you need to make sure correctness, which is one of the five C's of the Agentic Commerce Framework that I recently worked on with Azoma, is really, really important. You need to make sure what you're putting out there is actually what's showing up. And you can do that with a PXM, you can do that with social listening tools, you can do that with an agent.

And so correctness needs to be really top of mind for brands where social listening was usually thought of from the social team, or maybe if you have a consumer care team, they were thinking about that. Now it's imperative because you can't control all of the places that it's being talked about, and it can really snowball if that information is incorrect. And you need to be able to quickly action those changes. So if you know something is incorrect, you can't take 30 days to go update a claim or

Go update your brand.com to dispel the myth that was out there. You need to have a process in place to be able to update that incorrect information.

Sri (37:51)

Do I have to put a tag on every C P G guys episode now that it's AI because my co host is an AI robot?

Lauren Livak Gilbert (37:57)

Well, not yet. So no tags quite yet. But

Sri (37:59)

Not yet, okay.

Lauren Livak Gilbert (38:00)

but actually, Sri, I think you make a really interesting point because people say they're not using AI. I'm air quoting it, but most people are using AI and they just don't know it. So where is the line that you draw? Right. Claude just put out that new watermark on the content that's written through Claude on the back end, but there's robots that can get around it, right? But

I'm just, I don't know where this is gonna go from a regulation standpoint, but I think it's important for brands to really monitor that and for us all to understand that there's a lot of amazing pros to AI, but there's a lot of things that you can lose by using it, right? You can get further from the from the consumer. I was talking to someone the other day, they use like an AI note taker and they don't aren't retaining information as much. So I just think we need to be real about what you gain, but also what you lose when you're using AI.

PVSB (38:54)

Lauren,

from your perspective, how are you seeing AI reshaping competitive intelligence gathering and price promotion monitoring on the digital Shelf?

Lauren Livak Gilbert (39:04)

I was talking about this yesterday with someone. I mean I'm excited about it because from a scraping perspective, right, it's it's really challenging to get all that information. But I find that it's not always accurate. And a lot of the brands that I'm working with that are using it for like a price from a price and promotion perspective, the information that they're getting is very unclear and/or inaccurate. And so it's not super helpful. I think it can be a great monitoring tool if it's trained correctly.

You're using it on specific sites, that it is actually able to scrape that data and information. I think from a competitive, that's from like a price perspective. From a competitive intelligence perspective, I do think it's really valuable. If I take Amazon, for example, if you go on any Amazon PDP, underneath the carousel, they have those like blue pills, like the circles they call them. Those are all the questions that consumers are asking about products. So if you go to your competitor's product,

And you know what consumers are asking, you know the claims about that product, you know what is potentially a differentiator in your niche. So it's giving you information that you might not have had before. And I think that can be a leg up if it's used correctly. And also you can scrape competitive PDPs or imagery or look at things that you've never been able to do. So I do think it's very valuable if it's used correctly.

PVSB (40:25)

Lauren, I I used to think that some of the most valuable information to

Sri (40:29)

Yeah.

PVSB (40:30)

scrape was actually product ratings and reviews because you get verbatims. you could run those through LLMs and and there was a research done by a couple of professors at Northwestern and MIT that hypothesized that almost everything you get from an expensive focus group you can get by reading ratings and reviews, and they proved that out significantly. But the problem is that a lot of these

Big marketplaces understand that now and they're putting up mechanisms to prevent that content from being scraped. You can't scrape Amazon reviews the way you used to. Fundamentally, this is how most open information sources are now treating it. they are restricting the big LLMs from being able to query their data. And that that's putting a bit of a challenge into what's going on there.

Lauren Livak Gilbert (41:18)

And LLMs don't read JavaScript, right? And so ratings and reviews on a lot of brand.com or even D to C sites are using JavaScript. So they're not able to actually scrape that information. So it's not being pulled into the conversation. So I think you make

PVSB (41:29)

Yeah. Yeah.

Lauren Livak Gilbert (41:31)

a really valid point. And I think that's also why Reddit became such a great source. I know this week there's been a lot of controversy about like how much LLMs are actually looking at Reddit, but I do still think it has an impact. I think that's why it's been

Really important for brands to understand what people are saying about it. But also, Peter, to that point, I always say most large CPG companies have a consumer care center or have a call center or call line. People call and tell you about your product. Like if you've never met those people, go talk to them. They have so much data that you can use to share about what kinds of products you should be making next or feedback about products. You have a lot of that that lives inside your own company.

Sri (42:13)

I think about this and I was talking to our one of our team members, Suzanne, this morning, and we were talking about this very topic about call centers and things of that nature and how calls come in about product complaints. Which CEO last looked or took a care of any large market care company to really look at any of that stuff?

They only care about it when it's become completely out of hand and corporate comms as to jump on something. I mean, but you're right, the data sits right there. That's where the future citations are actually gonna come from. So I don't know, Lauren. That one is a problem.

Lauren Livak Gilbert (42:43)

Yeah, do you remember they were downstairs? They were downstairs

in our office and I used to go down there and just be like, Hey, can I like listen to a call? And I was like, Wow, these people are

Sri (42:48)

Yes. Ugh ugly.

Lauren Livak Gilbert (42:52)

calling and telling us about our products. Like, you have so much great data. Just go find it inside your own company.

Sri (42:59)

Ugly. Let's go to the future. So in latest research done on behalf of Coca-Cola by Think Blue, we've kinda learned that consumers are voting saying, I want to co pilot. I don't want AI to take over the world for me and make my purchase decisions on my behalf. But that's today. Things might completely change eighteen months from now, where for certain categories, replenishables maybe health and wellness, vitamins, pet food, but once you make a decision you stick to the same product, the same exact skew again and again.

This might be a perfect fit. We'd love to, you know, kinda pick your brain and how your how you see it when it comes to AI agents, parts that shop, compare, purchase on a consumer's behalf, you know, how's this headed? And if it is gonna head down that way, first of all, let's ask you, you know, how what do you think about that? And if so, who the heck is preparing for that world?

Lauren Livak Gilbert (43:52)

I don't know when we will get to a fully autonomous world and I don't know if consumers really want that. I also think it's easy for us who sit in our little bubble of e-commerce and talk about this all the time to talk about like people using AI and wanting a shopping agent, right? But like if I went and stood in Times Square for four hours and I polled 500 people, I don't know how many of them are actually using ChatGPT to buy their groceries or are

using an LLM to discover products. Maybe all of them are and maybe I'm incorrect, but I really don't think we are that far down the road. I think it is a potential in the future for specific categories. But the challenge is that there was a story the other day I was reading that someone gave their credit card to an agent to like replenish their toilet paper and ended up spending like $500 on toilet paper because there was no guardrail for it. Right. So like those are the things that you don't really want to happen.

So what I'm hearing from consumers, and we just did a research project on this as well, that most consumers do not trust an agent to purchase. They trust an agent from a discovery standpoint, but they then go and take that information, pick their specific retailer, and then go and purchase the product. I think the other piece there is retailers have still have a very strong foothold.

If I have Amazon Prime, I'm probably gonna use that. If I have Walmart Plus, I'm probably gonna use that. If I have Target Circle, right? Insert name of loyalty program, I'm probably gonna choose that retailer. So I don't know when we're gonna get to that place. I think we're gonna stay in the AI discovery phase for quite some time. and it will get much better in terms of the the data we have to do that. But to all of our conversations today, most brands are not even ready for the discovery phase. So

If that information is incorrect, why are you going to have an agent take your credit card and purchase something? And then the biggest question is: who's accountable? If they spent $500 on that toilet paper, are you gonna go yell at your LLM? You can't, right? Like, can you return it? I it gets really murky in terms of kind of what that looks like. And most organizations we pulled, we did a study with NIQ and the DSI, and 66% of them said they're.

somewhat ready for the agentic shelf, but most of them have data silos and they don't know how to get there. So I just think we're far from that world. but there are some use cases for super users who are doing it today.

PVSB (46:20)

Lauren, you poll a hundred people at Times Square, most of them will tell you they're there to look for the naked cowboy. So I don't know if I I That's okay. That's okay.

Lauren Livak Gilbert (46:26)

Yeah. I that was the only example I could figure out. But pick any major city and see.

Sri (46:31)

I used to go for

the Nissin Kapur Noodle Raman with the little smoke coming on top of the one

Lauren Livak Gilbert (46:37)

Okay.

Sri (46:37)

time squared. But you know, Ria and Laura

PVSB (46:38)

Very nice, Sri.

Sri (46:39)

have been on the billboard now.

PVSB (46:41)

very they

Lauren Livak Gilbert (46:41)

That's awesome.

PVSB (46:42)

have been. They are they are time. I've been on the I was on the billboard of Times Square last year, Sri. I was going for a conference and they put they put my name up.

Sri (46:50)

Same more, how what say more though?

PVSB (46:52)

no, it's a company that's no longer in business. They just they they just they just went out of business, Sri. You know which one we're talking about. in any event, Lauren, you obviously have an opportunity to see

brands and how they are operating in the age of AI. Would love from your perspective what you think is a big common mistake you see these brands making when they're adopting AI focused tools for digital shelf management, either too quickly or without a strategy in place.

Lauren Livak Gilbert (47:24)

Yeah, I think the biggest challenge and biggest mistake I guess I see is they say, let's do AI without saying what problem are we trying to solve? What does this actually mean for our business? And listen, I if e-commerce was on the lips of every executive the way AI was, we would be much further in this process. But I think that the challenge is that most people are getting directives from their leadership to say, we have to do AI, we need to be doing this, we need to fix this. And

Most people take that as okay, okay, like we're gonna go use Claude or we're gonna go add AI to this process or we're gonna go this do this this way, instead of taking a pause, taking a step back and saying, What is the problem I'm actually trying to solve? What outcome can this AI help me achieve? And then going from there. And I understand it's challenging inside the organization. There's pressures from your leadership, there's all the hype around it.

But if you can do that and really define what you are trying to do with AI, you will actually be able to see return. And if you actually reinvent the way you're working, it will be impactful.

Sri (48:35)

Wowza, wowza, wowza indeed. So I want to jump now three years out. What do you think just one single big way AI will change our brands, compete for shelf space and share of search?

Lauren Livak Gilbert (48:49)

We've already seen it happen. The funnel is collapsing. So you used to maybe take 15 different tabs, read reviews, go on TikTok, do all these things to research a product. The work was done by the consumer. And now all of that can be done in five seconds with an agent. So I think less options for consumers, which is something that they actually crave, which means less brands are going to be showcased and recommended.

And the competition is going to get even more challenging. I think we're going to see a lot of consolidation on the brand side for large organizations to be able to compete. And I think we're going to continue to see more and more work done by agents from a discovery standpoint and less and less brands available for consumption.

PVSB (49:37)

Well, thank you for this conversation, Lauren. Sri, what's your big takeaway?

Sri (49:44)

it's not one Peter, it's several, but if I had to summarize it, say AI is already upon us. Large market cap brands, you have to act now. Your inspiration should be the challenger brands. You can't go back to the age-old story and make excuses of the challenger brands don't have this sort of regulations, they don't have that sort of governance. Make your own. Lauren put a very important fact out there midway through the show of the 5x or more effort and energy required, the amount of content required in an AI world versus where we are today.

In a world where brands itself have not over excelled in content to begin with in the first place. I'm not saying that's everybody, but a large portion of the brands. Then the next thing is the disintermediation that could happen, Peter, which is if you're not present in the AI world, unlike the e-commerce world, you can half in, be half out, and you can limit which QCU, you'll just be absent and you'll never be found and eventually you'll vanish, which is a extremely scary concept. And I'll give a fourth one as well, Peter, which is education.

This needs to be the top, the leadership needs to get educated because the investments required have to come from a check from the board. This is not small $20K and $30K effort. You know, unless you have five SKUs, then maybe it is. But for most large large market cap brands, they're in the thousands of SKUs. So education on AI, et cetera, needs to be right up top in the first place. Otherwise, good luck. I'll leave it as good luck. What about you, Peter?

PVSB (51:09)

My big takeaway, Sri, is all around organizational structure. Lauren covered that. She said, listen, you you need to build excellence around this, whether it's formal or it's a center of excellence and it's cross-functional. Everyone needs to be involved in this. if you don't do that, anything you invest against, any efforts you make are going to be substantially less productive than they otherwise could be. So that's it.

Lauren, way too long having you on the podcast. We're grateful to have you join us. It means a lot to Sri and me to have you here speaking with us and bringing your incredible expertise. Congratulations again on the big news for Salsify, the parent company, and the acquisition. We couldn't be happier for you and your colleagues. Please give our best to

Both Piyush and of course you know, Peter, your co host, and of course, let's not forget the second most favorite bald Cuban descendant living here in the United States, Mr. Gonzalez. Thank you for joining us, Lauren.

Lauren Livak Gilbert (52:08)

Thank you so much for having me. This was a great conversation.

PVSB (52:11)

And if today's conversation sparks something within you and our audience, maybe a new idea, fresh perspective, or just a reminder of why this industry is worth showing up for every day, share it. Tag us, tell a colleague. The more voices we bring into this community, the sharper all of us get to our over 47,000 followers on LinkedIn that trust us to inform, educate, and entertain you about our industry. Thank you. Thank you. Thank you.

For your trust and encouragement. Till next time, keep pushing the conversation forward. I am, of course, PVSB on behalf of my co-host Papa Raj. This has been the CPG Guys Podcast. Goodbye.