Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Mind the Model: Episode 010: Kiri Masters
38 Plays25 days ago
Hosts: Nathan Guerra & Emmalee Crellin
Guest: Kiri Masters (Retail Media Breakfast)
Retail media is booming, but are AI shopping bots about to disrupt the entire ecosystem? We brought in the big guns. Kiri Masters is the host of Retail Media Breakfast Club, a daily podcast and newsletter that provides sharp, actionable insights into the fast-evolving retail media landscape. Informed by a decade in the industry, she built the award winning agency Bobsled Marketing before its acquisition in 2022. She writes regular columns for Forbes, The Drum, and WARC, and has been named a Top Retail Expert by RETHINK Retail for four consecutive years.
She drops in fresh from the NRF Big Show in New York City to break down the future of retail media, "agentic shopping," and exactly how brands need to adapt.
Quote of the Week:
"Look at this agentic pathway as a new surface to reach customers."
Mentioned Links & Resources:
- Check out the current state of AI and frontier model intelligence over time here: https://artificialanalysis.ai/trends#frontier-language-model-intelligence-over-time
- Google’s recent release for UCP (Universal Commerce Protocol):
- https://blog.google/products/ads-commerce/agentic-commerce-ai-tools-protocol-retailers-platforms/
Find Kiri at:
https://www.retailmediabreakfastclub.com/
https://www.linkedin.com/in/kiri-masters/
Mind the Model: Got questions or feedback? Email the show at mindthemodelpod@gmail.com
Recommended
Transcript
Introduction and AI for Small Businesses
00:00:07
Speaker
Hello and welcome to Mind the Model. I'm Nathan Guerra and as always I'm joined my co-host Emily Krillin. Em, how's it going? Hello, hello. I'm good, Nate. How are you? Very well. What have you been AI-ing recently?
00:00:21
Speaker
Don't really have much to report. I've been a bit slack this week, but it's interesting because we had a very fun phone call. The other day whilst I was doing my weekly shop and you mentioned a very interesting application for AI that is great for small business owners, startups, entrepreneurs, and just the everyday layman or woman. So please tell our listeners, how have you been getting into
Upgrading to Gemini Pro: Emily's Experience
00:00:49
Speaker
it? Well, so first of all, I um umm im i know I'm really late to the game, but I've just finally upgraded to a paid version of um AI.
00:01:00
Speaker
You're finally paying for AI. I'm finally paying for AI. ah you know And that's because I've been running a startup and you know not having a salary, and it didn't feel like something I had to do. And so, um but I basically got a free year of ah Gemini Pro. So I've been using that and it's been absolutely bonkersly amazing. I'm un loving it. So i can I can attest to the value of a great AI. But um what I did is I've been having to do my Australian taxes And I, being a small business owner, didn't realize that I had to do things like calculate GST. And I'm very lax when it comes to taxes. So what I did is I took all my expenses, threw them into ChatGPT, or Gemini actually, and said, help me categorize these based on the Australian taxation system.
00:01:46
Speaker
um And um tell me which of these are likely to have GSTs so I can go back and look at the receipts. And so I got a really good... kind of first pass at my taxes that will both make my accountant's life easier as well as my future life easier now that I know how to categorize things correctly.
00:02:05
Speaker
So I was, you know, again, continually impressed by the power of ai I love
Privacy Concerns with AI
00:02:12
Speaker
that. And one thing i haven't done it yet, so I'll have to do the exercise and report back, but i and taking a look at our household budget, we're looking to sell our home upgrade from unit to a hefty mortgage home in the Northern beaches of Sydney. And i really want to understand the budgeting of where our money is going, because I find so often, especially with the cost of living creep, it's the the money's going out.
00:02:41
Speaker
We don't always know where and why. and So yeah, I, I, very interested to hear about your categorization exercise because I think that would be really interesting to apply on more of like a personal finance budget lens. Yeah. However, and I pay for chat GPT, but something just gets me a bit, I'm not fully bought into the idea of turning over all of the dirty laundry of my expenses to AI.
00:03:08
Speaker
So I would love for someone to come and tell me like, no, it's okay. It's safe and you know privacy protected. But I don't know. I haven't done it yet. And I think that's my hesitation for why I haven't yet jumped jump in. Now, this brings up an interesting point because, and again, ex-Google drank the Kool-Aid.
00:03:26
Speaker
But having read so much about it, it does seem like Google has more guardrails around their AI than maybe people like OpenAI ai do. Is that a consideration for you as well? I would feel more comfortable in Gemini if I paid for it.
00:03:41
Speaker
Okay. But I don't currently pay for it. I only pay for a chat GPT. yeah So do I look to, you know, shift away and move and pay for a different model? I think we're kind of coming to a point where, and we've discussed this before, all models are pretty much the same.
00:03:57
Speaker
They, you know, that output is. Yeah. Yeah. And they're all going through really rapid development, but they kind of leapfrog each other incrementally. I'm sure there's a graph we could put up um or link it in the show notes.
00:04:10
Speaker
So if I paid for and had like ah a paid subscription for a platform like Gemini, I think I'd be a bit more comfortable because Google does know a lot about me already and I'm integrated into their ecosystem quite well. So yeah, that's it's an interesting point of like the the trust of the models and where where we trust um it to look at our deepest, darkest secrets. And for me, that's financials. Well, you know, the good news is that there's a ton of apps out there that do that kind of stuff as well. So you probably don't even
Guest Introduction: Kiri Masters
00:04:38
Speaker
need AI. You probably can just...
00:04:40
Speaker
you know pay for a one-time app that would like help you track all that stuff um nate i am stoked for this episode today i self-professed fangirl uh of this upcoming guest and as we were uh talking about before we press record it's a bit of an inverse we are two american born um australian citizens podcasting in Australia about to interview a green card holding Aussie living in East Coast of the US. So it's a bit of a flip flop.
00:05:14
Speaker
I won't yet um spoil who we have, although for those that are listening and watching, obviously you can see the title of this episode so you know who it is. um But our guest is an independent commentator, speaker, advisor, and podcast host extraordinaire covering the retail media industry. She produces a near daily podcast, which just hats off to you. That is some um epic content coming out at such a fast clip because I know sometimes we do struggle on our end. and And she also has a newsletter called Retail Media Breakfast Club. It's informed by a decade in the industry as the founder of an award-winning agency, Bobsled Marketing.
00:05:51
Speaker
Love to know how that name came to be. It was actually acquired in 2022. She writes regular columns for Forbes, The Drums, and Wark, and she's been named a top retail expert by Rethink Retail for four consecutive years. Kiri Masters, welcome to Mind the Model.
Retail's Big Show and In-Store Media
00:06:05
Speaker
Hello. great to be with you. You're you back home now, I imagine, of a very fast and furious few days in New York City from the National Retail Federation's Retail's Show. yep I've never been. I actually didn't know it existed until a couple months ago. I'll say about the the Big Show, it's interesting. It's a very much an in-store media ah sorry, in-store retail based show. There's a lot of um other conferences out there that ah that focus more on the digital or e-commerce experience like a shop talk or other shows of that nature. But NRF's big show is all about the store. um So that's interesting for me as someone who's primarily focused on
00:06:58
Speaker
my career so far has mostly been on the digital side, but what's where the intersection there is, is ah in-store retail media is a big, um is a, is a booming space within retail media.
AI in Content Creation
00:07:13
Speaker
So there was a lot of content around that. And then another sort of, area of interest for me is agentic shopping and how retailers are thinking about that, how they are planning for that scenario. So there was a number of sessions I sat in on from, with panels from the CIOs of retailers like Home Depot, Wayfair, Lowe's, et cetera, talking about like how they're thinking about philosophically and how they're building toward it
00:07:44
Speaker
the The model of our show is we typically running through a couple quickfire questions. So just first thing that you can think of, and then we'll get into the meaty stuff. ah What is your preferred Gen AI model that you reach for most right now?
00:07:58
Speaker
But what's one job that you refuse to let it do? I definitely reach for ChatGPT a lot. What's one thing I refuse to let it do for me? You know, I produce a lot of content. I rely on AI a lot for that. I'm not going to, I won't, you know, mislead anyone by saying I write everything my everything myself or I edit everything myself. But I have to admit, I've pulled back a lot on getting AI to help me write things. I just see like such formulaic drivel coming from everyone about the, you know, the,
00:08:34
Speaker
masterclass that X company delivered on this or whatever. I'm like, oh my gosh, I just feel like um to to stand out, you need to do something totally different, non-formulaic written by AI these days. So I'm really taking taking like ah you know taking taking the reins back, I guess. And is that in part because it just hasn't captured your voice? I want a couple of things. one I find myself getting lazier cognitively because I'm not really thinking things through. and It's a common theme. Other people have admitted to it. I feel like, I mean, my memory is shot. i don't remember things anymore. i
00:09:18
Speaker
I'm outsourcing my thinking to it. And i I am just like, you know what? Enough is enough. I need to really think this through myself. Otherwise, I'm not going to retain it. I'm not really going to make the the connections. I will still...
00:09:32
Speaker
get the AI to maybe like flesh things out. What else haven't thought of? But I want to take at least the first pass at a concept and use my brain. okay So that's one thing. i mean, i think it can kind of capture your voice. I would just find so much time. I'm very, I'm not a perfectionist in most most areas of my life, but like when I would go back and look at them,
00:09:55
Speaker
no, if that's wrong, that's wrong. And i I would spend so much time rewriting what the AI model had already written. I was like, you know what? yeah I think I need to get 80% of the way there and then ask it to reorder things. And it does a great job of that where i'll be like, you know, this is a much stronger opening to your article if you start with this. I'm like, actually, that's a great call, but then it's all what I have already come up with.
00:10:21
Speaker
Kiri, do you have like ah a prompting methodology or approach that you bring to in AI? I've started to use those projects. um ah I've been using those for a while, you know, where you can create a project in Claude or chat GPT where you sort of upload a whole bunch of stuff. So for example, one thing that I have done that's been very effective is I, you know, creative person when I, when a partner approaches me, I'm like, oh, we could do this and this and this and this, all these ideas. And so I've put all of those, all of my proposals, all of my contracts, things like that into a chat GPT project. i' like, this is what I've done. So it gets a sense of like scope deliverables, timeline pricing. And so now when I'm sort of talking with a new potential partner or sponsor,
00:11:11
Speaker
It's got all this context about what I've done in the past and what like some good pricing models might be. And so that's actually been really great. So I recommend those project capabilities, especially in chat
AI Prompting Techniques
00:11:24
Speaker
GPT.
00:11:24
Speaker
Prompt engineering was almost like the 2024, 2025 era, but we're entering into a new phase of as much context as you can throw at something to say, if you want to think like me,
00:11:37
Speaker
You need to have the full background story to be able to deliver something that's quite valuable. yeah Although I did just read something yesterday that apparently the most, the the new prompting trick is to repeat your prompt.
00:11:51
Speaker
So you type the exact same thing twice and that's because the way AI works is it reads from like character one until the end. And so it doesn't have the context of the entire thing until it gets to the end.
00:12:05
Speaker
And so if you give it the whole context twice, apparently it's much more effective. I also like to tell it that it's wrong. It'll give me an answer and say, no, you're wrong. And then it will, it'll either backpedal and totally capitulate, right?
00:12:17
Speaker
Or else it will be like, no, I'm pretty sure I'm correct about this. And that's when you know you really nailed it. That's when you've got it. Okay. I like to threaten models against each other. I was like, it's, I was experimenting in with, like, cloud code, fantastic, but 21 different versions of the artifact and kept breaking. And I finally wrote to it and said,
00:12:38
Speaker
at what point are you going to give up? And at what point should I give up and leave this and go ask ChatGPT? And they're like, I am so sorry. This is just not and not not sufficient.
00:12:49
Speaker
I will make it, I will completely rewrite it. And then the next output, fine. I was like, a little bit of threatening work somehow. I don't know. So we've moved from like prompt engineers to autocrats.
00:13:03
Speaker
Well, it seems to be the way of the world right now. um um So, Kerry, question for you because, you know, it made our listeners know that we're an AI and marketing podcast and a lot of what we do is specific to AI and marketing and your your expertise is retail media.
Kiri's Journey: Banking to Retail Media
00:13:23
Speaker
But when you go out for you know a dinner party with friends and they say, oh, Kiri, what do you do? Do you just say retail media and people are like, oh, or like, what's your what's your cool pitch to get people like excited? Like, oh, wow, you're changing the world. Depends how quickly I want to move on from that conversation. Like if I want to just If I was at like a a neighborhood barbecue a couple months ago and someone asked me what I did and I just said journalist. um I think like a real journalist would be offended by that characterization. So that like, you know, if people are in if people are interested, I'll explain a little bit more about it. But I i don't usually, ah a more succinct thing is like, i mean I'm in advertising or I write about advertising. So you you don't go into the depths of that retail media world. Not unless someone really presses it. Is interested. Yeah, most people aren't. Shocking. But I'd like to say like, oh, you know, when you go to Amazon.com or you go to a retailer and you see the ads on their website, like, oh, yeah, well, that's what I that's what i do. Thanks. It's probably the response you get most often. Right. Exactly.
00:14:37
Speaker
So, Kiri, in a lot of ways, like retail media really is the kind of new hot kid at school. um How did you get into this room? Totally fell into it. i was It 2015. I quit my job as a commercial banker at JPMorgan Chase in New York because i had this little e-commerce business on the side that I had been running for a couple of years at that point. And I had learned along the way a little bit about Amazon.
00:15:07
Speaker
And I realized, oh, I know more about Amazon than some of these clients that I have at at the bank who are real manufacturers. So maybe I should go and like consult for them.
00:15:17
Speaker
And that's how it started in 2015. So that was early days early days, certainly with retail media. Amazon had only recently launched sponsored product ads on on site then. And just, you know, and I was in the right place at the right time, truly with building that agency. Over time, we...
00:15:37
Speaker
expanded out from just focusing on Amazon to then also working with consumer brands on their Walmart.com, Instacart, presence, Target ads. And that was my sort of entry point into the world of retail media.
Agentic Shopping and E-commerce
00:15:54
Speaker
Agent shopping. Want to hear your take on this. Is it another buzzword? Do you think it's going to change the world? I don't think it will be the the major entry point for a lot of e-commerce for a while. I think that this year is going to be very interesting year with Google's UCP and a lot of these retailers getting on board and actually starting to build that connection with LLMs and opening their product, like Walmart opening its the product catalog to to Google. They've really jumped in with both feet. And so I think when you look at the the adoption curve of of new technology, there's there's five sort of things that need to be present. One is pain, which is is the current way that we do things painful, which most people wouldn't characterize e-commerce as painful when you step back and you look at like the the friction points of yeah price price shopping in comparison is actually kind of difficult to do that manually. You need 12 tabs. all these different retailers, who has a discount, what's the shipping cost, what's the return policy and like figuring all of that out.
00:17:07
Speaker
There's, you know, the problem of fake reviews and and sort of commodity products that are of questionable quality. And, you know, there's there's a number of different frictions that we don't really recognize as such because we're just that's just the way that we've been shopping online for 20 years. We've become accustomed to it.
00:17:27
Speaker
Exactly. um another Another aspect is, is the technology actually there yet? Which in the case of agentic shopping, I would say no because we still don't have integrations with loyalty programs and past purchases. ah you know there's ah There's a bit of a ways to go there. And then sort of economically, is there a is there an incentive for the retailers to do this? Is there an incentive for shoppers to do this? um What's the incentive for the AI companies? I think we've seen with like open AI kind of like launching into shopping and then kind of stepping right back because they're like, oh, we're overwhelmed. We can't actually allocate resources to this. So it's not, you know,
00:18:17
Speaker
Google has sort of stepped in to fill that void. So it's a very complex thing when we look at adoption of new technology like this. And I think that most consumers are going play around with some of these features, like the price tracking feature on Google Shopping and things like that. And be like, that's kind of cool. But where I think the big advantage that these LLMs have in adoption is just how consistently people are actually using these tools every day. And you look at the daily usage of these tools, it's just creeping up and up across all these different segments. People are finding that they're relying on it every day. And it's not just like, you know, weirdos like us. What are you talking about? It's like across every cohort. Like my mother, I had to explain what ChatGPT was to her in 2024. And now, you know, she's using it. So across all of these cohorts, people are using it more and more. And so that gives the LLMs an opportunity to kind of keep nudging you, hey, you want to, you know, incrementally making that shift. Yeah, exactly. So that frictionlessness is another really important criteria of lasting change in that
00:19:38
Speaker
are we already using this device, this software? What is it? Are we actually using it for other things? Not just like a ah new type of shopping behavior, which of course with AI we are.
00:19:52
Speaker
Full authentic shopping doesn't feel like a 2026 thing then. Yeah. I mean, is it a 2030 thing? Is it?
00:20:00
Speaker
Dude, I don't know. We're like two weeks into 2026 and all this crazy stuff's already happened. I just, I think like this, this is the year that we throw projections out the window. I just, that is such a good point. first Who knows? Fair enough. So it's definitely not today. It's definitely a future thing, but when you're not going to put a stake in the ground on that, i get that. Recent news out of Woolies just announced that their little chatbot assistant, Olive,
00:20:28
Speaker
I have used before. and It will now start to plan more for that, but it won't it won't fully go into agentic shopping. I think that retailers will have a lot to prove in terms of the utility of these assistants before people will actually start to meaningfully engage with them. And it's only once they have been ah assured through a number of different interactions that, oh, this is actually ah this is a good tool, this is better than searching, this is maybe a way I can save a little bit of time with my meal planning and recipes and adding things to my cart.
00:21:05
Speaker
um You know, by by the 10th time i've I've used that tool, yeah, I would trust it enough to like put my payment details in there and link it up to my everyday rewards account and um just let it go go wild with like reordering my core essentials every week.
00:21:24
Speaker
I think that's a great point. And kind of going back to the earlier chat Nate and I had about the financials, it's so important that trust that you have to, that's like the the last, um the last hurdle for LLMs in terms of, I hate to be a Debbie Downer, but looking at the threat of agentic shopping, do you think it's an opportunity or more of a threat for retail media
LLMs and Retail Media Networks
00:21:46
Speaker
networks? So if we,
00:21:49
Speaker
If we look to what you've written previously about how LLMs will intercept queries before they reach, you know, just shopping directly on Amazon. Is that where the the onsite ad inventory and, you know, all the millions and billions of dollars that's attached to that? Is that where it shrinks? it Is it a nightmare scenario if LLMs point consumers directly to the product and, you know, cut out that middle ground of the funnel of that discoverability.
00:22:18
Speaker
So I think that when you look at the landscape of retail media, there's three major categories of retail media, you know, activations. One is the onsite inventory that you mentioned, which is the majority of retail media spend. It's like 70, 80% of retail media spend is these ads on, on the website, sponsored products, banners, things like that, which are sort of that they're particularly helpful at the bottom, that the the latter stage of the per purchase journey.
00:22:51
Speaker
And they have a really good ah ROI that will on paper, a really good ah ROI. um And so if people are doing more research and consideration upfront, and then just going to a retailer.com to transact once they have decided, what widget they're going to buy, then yes, there's less time on site. There's less keywords that you're using to search. There's less pages that you end up browsing before you make a purchase decision. And so therefore less in ad inventory overall. So again, coming back to the point of like, is Agenic going take over? Look, if it's a percent, a smaller percentage of, of e-commerce this year, it's not going to kill on-site retail media.
00:23:35
Speaker
and anytime soon, but it will start to erode at that. The second bucket of retail media is offsite retail media, which is when a retailer sort of ah allows advertisers to activate on their audience offsite. So on social media, on connected TV, things like that. So the risk I see here is if these LLMs create their own media networks.
00:24:07
Speaker
Why would they do that? Well, one, it's a pretty neat way to monetize all of these users that they have and all of the things that we're feeding in about like you know, our hopes, dreams, desires, wishes, stuff we want, whatever whatever, like they know a scary amount of information about us, even if they don't have your tax planning projections, like they still know too much about than you'd care to admit.
00:24:36
Speaker
So if I was an LLM, I'd be seriously considering that because that would mean i could monetize my user base without ever showing them an ad in the chat interface. which is what a lot of people have been scared about is like, I don't want to see ads in chat GPT.
00:24:57
Speaker
It's going to like, at I feel like the utility of chat GPT is that it gives me an honest appraisal of what my options are. And if it starts showing me ads, I just feel like it's trying to make money off me.
00:25:11
Speaker
Well, so off to me, offsite kind of seems like a really great business model for that because yes, they're going to make money off of you, but it's going to be because When you're watching Disney+, plus you see an ad for, i don't know, Crackers that because, you know, the Crackers company knows that you're going to be hosting a ah ah barbecue soon or something like that.
00:25:33
Speaker
So that is, um again, if you're a retail media network, that you've you've sort of charged a premium CPM for your audience data,
00:25:45
Speaker
where an LLM is gonna have ah also a very compelling audience data set that could now sort of in theory compete with your audience um data set.
00:25:57
Speaker
That's really interesting because I mean, of course Google already has that. yeah You know, they already have a huge understanding of everybody on the internet. So that's a really interesting kind of build, easy build for them. It's interesting because we just recently spoke to um Adam Goodman who works at Microsoft here in Australia. And he was talking about the fact that they're definitely bringing ads to their AI. Copilot?
00:26:23
Speaker
In Copilot, yeah. They've had them for a long time. Yeah. And, but apparently right now you cannot target Copilot ads. um They are part of a bundle of advertising that you can buy across MSN. And that will hopefully, that will change eventually to be specifically targeting co-pilot.
00:26:41
Speaker
We had a discussion at South by Southwest Sydney, RIP, recently saying that it's not gonna be up again. But we recently chatted to Paul Blackburn about that, of like, you know, the data that retailers had, that first party transaction data, like so, um it's like the golden nugget of why they exist.
00:27:01
Speaker
And the way we were phrasing the conversation with Paul was, you know, ads within, ah you know, the likes of ChatGPT. But I think your point about off-network, it is such an easy win for them. And yeah i I know people are talking about advertising and LLMs as, you know, a display banner alongside the chat.
00:27:20
Speaker
Do you think it will erode consumer trust and they'll say, oh, well, no longer want to be interacting with that LLM if I know I'll be targeted for advertising? This is just a this is a reading the tea leaves thing. I don't have any sort of insider information about any LLMs doing this. But when I look at the, I think of you all all the things that these LLMs are doing through a retail media lens and like where could there be challenges just to round out that those three buckets of retail media spend the other, which is still very, very small, In the US, I think, I'm not just so sure about Australia, but in the US, only 3% of retail media spend goes here, which is in-store retail
Resilience of In-Store Retail Media
00:28:07
Speaker
media. And specifically when we're talking about that, there's a little bit of fuzzy definitions. People love to argue about this. But when we're talking about that, it's just specifically the digital signage and things that can be, ads that can be purchased programmatically rather than the paper,
00:28:23
Speaker
um activations and things like that. So it I think that that is extremely resilient with all of this agentic commerce behavior, because people will still go to stores to shop. There's lots of missions that are best executed by going to a store, including the let's get out of the house on a weekend um mission, which is like, where where are all the third places these days? There's not so many of them. And so for better or worse, going to going to a store or a mall is like the third place for a lot of individuals and families. so We will still keep going going there. And I think that that is actually part of the resilience of retail media, even if an LLM is seeing a, they're making a, that they're sending you to a transaction, it's still the retailer who is going to be that merchant of record and actually see the in-store transactions as well. So even if some offsite spend moves to LLMs, this crazy, you know redistribution of that,
00:29:35
Speaker
the retailer is still going to be the only one that truly, truly sees that actual transaction and then also what happens in the store as well. So that's sort of like the bull case for retail. retail I still think the LLM will be to see, will be able inference, infer a lot yeah on activity, but they won't know for sure.
00:29:57
Speaker
um if you bought that item later on. Unless you're like M and uploading all of your household receipts to LLM.
Non-Traditional Entrants in Retail Media
00:30:06
Speaker
Bit of a wildcard question for you, Carrie, only because I'm passionate about retail media. Are there certain stores that are getting into retail media that cause you to just stop and think like, but don't know, is a local like hair salon starting to get into thinking it can, you know, push into retail media or coffee stores? Yeah, I mean, there's a sort of outside of,
00:30:28
Speaker
retailers, there is the commerce media networks that are starting, you know, getting, getting a bit of traction. So you think about airlines, um, banks, things like that. I mean, also fitness like gyms and fit. These have all existed as well. So the idea of an airline selling ad space is not a new concept, but the way, you know, the sort of, sort of getting, uh, lumped into the same consideration set. So, um Yeah, it's it's it's expanding in terms of the the definition and that the the way to think about it, I think. But to me to me, retailers have something very unique, which is knowing at the skew level what you bought. So ah a bank or a credit card company will know you spent this much at this retailer on this day. And they might be able to, again, infer something about that, but they don't know of like to the skew level, what you actually purchased.
00:31:29
Speaker
How do you see ads fitting, I guess, into the agentic shopping universe? um Do you really see that AI is going to be doing all the shopping? Are we going to be advertising to bots or are we advertising to humans or what's your take on it?
00:31:42
Speaker
We might see more experiential and creative collabs between brands and retailers that don't involve a onsite ad unit, which could be ignored by LLMs or not. We don't know. And a retailer, a retailer can't control that. So things like in, in-store retail media, product sampling, post, post-purchase offers, these are all things that They exist today, but I think that there could be a more compelling use case for those in in the future as brands wanna push past the traditional sponsored product ad unit that may or may not get seen or may or may not be clicked on by a real person. um I think that that brands will have more appetite for really creative and disruptive ways to get to consumers beyond the sponsored product ad unit.
00:32:46
Speaker
People will still travel store. We won't be migrating purely to an e-com world. and And I think an interesting part about the experiential, because we had a previous guest on the pod, Natalia, who spoke about that that human connection is still so um valid. It also ties in the importance of branding.
00:33:04
Speaker
So brands being able to show up and remain top of mind in areas and connect with consumers so that if consumers are then asking LLMs for recommendations and it comes back with three, oh, actually, you know, I had a really good experience with brand X. I'll go ahead and transact with them. Mm-hmm.
00:33:23
Speaker
For marketers who are listening right now and who are quite scared of the future, if you know that agent at commerce is the next wave, what is one skill that you recommend their team really needs to dive into this year to not drown?
Agentic Pathways as Engagement Opportunities
00:33:38
Speaker
I can't point to a specific skill because things are moving so quickly, but it's probably more of a mindset thing. And that is that look look at this agentic pathway as a new surface to reach customers.
00:33:53
Speaker
So just like the early early days of e-commerce and it was weird, like, why would you buy a textbook on the internet? That's wild. Like, how could you put your credit card into the internet? And then like stuff just started showing off on your doorstep. You're like, wow, this is this is kind of crazy. What else could we sell on the internet? um so it's when you think of it as this is a new channel it's a new surface it is a new way of engaging with customers in it like and this maybe the behavior and utility is a little bit different but it's not maybe think of it less ah of a threat and a disruption to your e-com site or your retailer relationships and more of like this is a new uh
00:34:36
Speaker
a surface, a channel, and a door, a way to connect with our customers in a way that they're starting to adopt now. um i think that reframe is really helpful. And certainly what I was hearing last week at and NRF from these retailers is they're like, well, of course we would rather people buy from us directly,
00:34:58
Speaker
but We also understand that they are using LLMs for this stuff, so we're going to build it that way too We've got to do both. So I think that that is like the reframe that marketers need to have.
00:35:09
Speaker
I wish we had more time, but I do want to do a couple of quick, fair questions. i'm I'd love to get into the dichotomy between Amazon's point of view about how they're approaching into shopping and everyone else, but there's some amazing articles on your website about that.
00:35:24
Speaker
um But just to kind of close things off, ah do you consider yourself to be an AI optimist, realist, or pessimist?
Future Disruption in Retail Media
00:35:32
Speaker
Probably objectively an optimist. um would That would have been my guess from this conversation.
00:35:36
Speaker
And then if you could fill in the blank for us, Kiri, will be blank for retail media in three years time. I'll say disruptive. But I'm taking that as like disruptive.
00:35:48
Speaker
like opportunities. Positive disruption. i got Positive disruption. Look, it's all about how how do you interpret that? Go back to that mindset shift. Love it. Yes. Well, we'll leave that for the yeah the listeners to kind of interpret as they do.
00:36:03
Speaker
Yes. As they do, Keri, thank you so much. That's been a really, really interesting conversation. I appreciate your time. Thank you so much for having me. I really enjoyed the discussion.
00:36:16
Speaker
Amazing, Nate. Great conversation with Kiri. thought it was an amazing chat. Yourself? Personally, i am a fangirl of her hers, as I have stated. I'm also a fangirl of retail media. So retail media plus AI, this I've been looking forward to this chat for quite some time. um The thing I really ah heard from her is, yes, there's in retail media, there's so much that is online for you know e-commerce and for retailers that have an online presence. but we can't discount the opportunity for more of that one-to-one experience in store.
00:36:48
Speaker
I also really liked the ah her kind of, and I want to get from her so we can add it to the the notes. She talked about like a list of five items that is on the tech adoption curve. I know I've seen the C curve before. So I'm kind of curious to see what what those five items were. And well we'll put a link to that in the show notes as well.
00:37:04
Speaker
What else did you take away? You know, it's funny because it reminds me of a lot of news that I've been reading about how the the chatbot interface was very much 2024. know, agents were very hyped in, 2025, but it's almost not regressed, but kind of progressed in a different manner that 2026 is more of the year of automation.
00:37:24
Speaker
so not full agentic workflows, but finding automation and just making things so much quicker. i feel like we could have had so much more time with her. I'd like to say, we didn't really get into Amazon and what they're doing and how it's different than other shoppers.
00:37:38
Speaker
um I've been reading a lot of her stuff and reporting her around Costco as a retail media network. Like there's just so much more we could have done with Kiri. We'll probably like, we should definitely have her back. Agreed. And, you know, to the point about Amazon, i it's it's still relatively new. And um I mean, obviously, Brennan-born company, and they've been really good to adapt to all of the changes and really help, I think, shape and shift consumer behavior for how they experience shopping online. One of the articles that, I mean, like you said, there's so much we could have asked Keri about. One of her articles recently was talking about how The sponsored product ads on Amazon, like they don't see it as having a ceiling. They can just keep on adding more and more because they can. And we just think of like, you know, and I know from being on the other side of it, the margin that's attributed to that, it's just a spectacular business model that they're continuing to push. But I don't know. I think...
00:38:36
Speaker
It'll be very interesting to see how Australia's retail media ecosystem develops over the next year. I think there is still such a importance for that in-store experience because that's how a lot of Australians tend to interact with shopping. Yeah. I mean, you know, I shop almost every day at Kohl's because we don't have any space for bulk, you know, like I live in a relatively small apartment in a very urban area. So,
00:39:05
Speaker
Thanks for a great year of podcasting behind us. and First couple months, very new to the game, but we really appreciate our listeners. If they can share any feedback or if they want to ask us send any questions, and send an email to mindthemodelpod at gmail.com. Be sure to like, share, and to subscribe to all of our episodes wherever you watch and or listen.
00:39:26
Speaker
And remember, the intelligence might be artificial, but the wins are real. See next time.
00:39:34
Speaker
This episode was brought to you by Testchimp. Testchimp is a web developer's best friend. Testchimp uses deep runtime context to automate your testing workflows, generate scenarios, write scripts, and fix bugs all in one place.
00:39:50
Speaker
Stop wasting time and start shipping quality code faster. Bugs suck, but finding bugs doesn't have to with Testchimp.