AI is transforming industries worldwide, but how is it reshaping the retail experience? In this episode, we sit down with Lisa Avvocato and Mika Yamamoto to delve into the role of AI in retail and marketing. Lisa and Mika are seasoned marketing and business leaders specializing in AI, customer experience, and enterprise transformation. Lisa is the VP of Global Marketing and AI Community at Sama, where she champions ethical AI development through high-quality training data. Mika serves as the Chief Customer and Marketing Officer at Freshworks, a company driving global growth with AI-powered solutions that enhance customer and employee experiences. In our conversation, we unpack how AI is transforming retail operations, customer service, and online shopping experiences.
We explore the current trends of AI-based solutions in retail, what has driven its adoption in the industry, and how AI-based customer service technology has improved over time. We also discuss the correct mix of technology and humans, the importance of establishing boundaries for AI, and why it won't replace humans but will augment workflow. Hear examples of AI retail success stories, what companies got AI wrong, and the reasons behind the wins and failures. Gain insights into the value of copilots, business strategies to avoid investing in ineffective AI solutions, and much more. Tune in now!
Key Points From This Episode:
Quotes:
“I think [the evolution] in terms of accessibility to AI-solutions for people who don't have the massive IT departments and massive data analytics departments is really remarkable.” — Mika Yamamoto [0:04:25]
“Whether it's generative AI for creative or content or whatever, it's not going to replace humans. It's going to augment our workflows.” — Lisa Avvocato [0:10:46]
“Retail is actually one of the fastest adopting industries out there [of] AI.” — Mika Yamamoto [0:14:17]
“Having conversations with peers, I think, is absolutely invaluable to figure out what's hype and what's reality [regarding AI].” — Mika Yamamoto [0:30:19]
Links Mentioned in Today’s Episode:
Mika Yamamoto: Right now, we've got the co pilot trying to make jokes, and they're pretty terrible. And the language actually isn't on point with what we want to say, and sometimes it's technically inaccurate. And so had we actually just completely trusted the system and not trained the model with human beings, we would have done our customers a disservice and our company a disservice.
Rob Stevenson: Welcome to How AI Happens, a podcast where experts explain their work at the cutting edge of artificial intelligence. You'll hear from AI researchers, data scientists, and machine learning engineers as they get technical about, the most exciting developments in their field and the challenges they're facing along the way. I'm your host, Rob Stevenson, and we're about to learn how AI Happens. All right, hello, hello, everyone out there in webinar land, in podcast land, anyone who is tuning in, whether it is via Zoom or LinkedIn, wherever you like to consume your sweet, sweet AI content, boy, are we happy to have you. I'm your host, Rob Stevenson, and under normal circumstances, I host and produce Sama's podcast, How AI Happens, where I sit down with leaders in the artificial intelligence space and we speak about the challenges befalling them as they bring this awesome tech to market. And we are doing a very special edition. We are doing it live. You may be thinking, Rob, if you normally do the podcast asynchronously, why are you doing it live? Well, because there'simply no time like the present. And you know that it's gonna be a good one when I bring out the big guns. I have done so today. We're gonna meet our awesome panelsists here in just a second. But we are gathered to speak about an amazing, exciting new, or at least developing, use case in AI, which is of course retail. And specifically, we're going to reflect on what is the Super Bowl, I guess, for retailers, the back half of Q4, the holiday season. We're going to look at some of the winners and losers and sort of reflect on who did and did not leverage AI to their lasting business glory. So let's get to know our panelists here. Joining us from Sama, she serves as a VP of global marketing. Give it up for Lisa Avvocato Lisa, welcome to the podcast webinar live stream. How are you today?
Lisa Avvocato: Thanks. yeah, doing good. I'm super excited to be here today. I have a ton of experience in retail tech. My entire career has basically been but around helping retailers accelerate their digital transformation strategies, specifically within AI, trying to improve those average order values and streamline operations to increaseit margins. So obviously have a lot of opinions on the winners and losers this past year.
Rob Stevenson: I wouldn't have you any other way. The more opinion needed the better I think when we were doing these things. So I'm really happy to have you, Lisa. Also joining us and waiting patiently in the wings is a woman with a ton of experience in the space. She, has served as the Chief Digital Marketing Officer over at SAP, was the SVP of Go to market for marketo. Currently she serves as the Chief Customer and Marketing Officer at freshworks. Mika Yamamoto. Miko, welcome to you. How are you today?
Mika Yamamoto: Thank you, I'm doing really well. We're just finishing getting pelted here in Seattle with some snow. So just trying to stay warm and stay safe, but doing really well. Thank you.
Rob Stevenson: Precipitation is a real and ever present danger in the Pacific Northwest. I understand. So I'm sure you're well used to it by this point.
Lisa Avvocato: Yep.
Rob Stevenson: Did I do your curriculum vite justice there? Is there anything else that we can add for the folks at home?
Mika Yamamoto: The other thing I'd add would be that I actually have worked in retail for a bit of my career as well and working at bringing physical retail to Microsoft. So opening up the first, actually 15, 15, 20 Microsoft stores and then starting with Amazon and having worked on Amazon Go and Amazon Books and so bringing physical retail to Amazon. So I've spent my career in both rolling out retail solutions and then also enabling them with the B2B SaaS companies that I've worked for as well. So I've had similar perspectives to retail as Lisa has as well.
Rob Stevenson: Wonderful. Yeah. And you've been monitoring the space for a long time. You kind of live and read this throughout the last several places in your career. So that's why I was excited to have you on and speak with you about some of this stuff. So I am curious to hear from you, kind of at the top, just what you've seen in the retail sector. If you look back over the last one to two years, as obviously with Generative, more recently with Agentic, how are you kind of seeing the trends and emerging technologies of artificial intelligence sort of embed themselves in retail?
Mika Yamamoto: I think that what's been amazing to see in terms of this evolution because I mean, it's not like AI is particularly new. What is new is just more of the democratized access that we can have. So you don't necessarily have to be an Amazon or a Walmart or a Target to take advantage of AI and your retail environment, whether you're an E commerce provider or whether you're, a brick and mortar or do both. So I think that that evolution in terms of the accessibility to AI solutions for people who don't have massive IT departments and massive data analytics departments is really remarkable.
Rob Stevenson: Lisa, is that kind of been your experience too?
Lisa Avvocato: Yeah. And to add on to that, I would say, I think also just the pure technological advancements, I think LLMs obviously have just catapulted us into a new realm, but AI in the beginning was really more around solving business outcomes, like business process outcomes. It was built more for the business, less for the consumer. And a lot of these advancements are actually flipping the script and going to provide much more tangible and real solutions for consumer problems. And that's ultimately going to increase adoption, number one, but also drive much more positive customer experiences throughout the lifec cycle.
Mika Yamamoto: And I think the fact that we've had after the release of chat cpt, people became more comfortable. I think people from a consumer standpoint have been far more comfortable with dealing with more conversational experiences that they know aren't humans, but that are actually increasingly human. To your point, Lisa, in terms of the innovation that's been taking place. And so I think the comfort factor with people, whether they're consumers or if it's in a B2B scenario, has helped a ton in terms of the adoption or the acceptance of the technology itself.
Lisa Avvocato: Yeah. And, I just think the naturalness of the responses. Now I remember the beginning chatbots where it was basically just a menu tree, right. Where you're clicking on buttons and just frustrating right out of the gate. Now you can start to have a real conversation and you know, it's not with a human right now, but I actually feel like I'm getting help versus just kind of stuck in this menu or like death by menu, death by clicks design only it was like taking.
Mika Yamamoto: The you rem the when it was on a VRU and it was over the phone where you just wanted to zero out just because you wanted to get to the live person. I think Operatorator. Yes, operator. The same held true for the initial chat bots. She was like, how do I please speak to an agent? Please speak to an agent or at least chat with an agent on in chat. I think now the conversational tone and the personal nature of the interaction is so much better that you feel like now, now I'm comforted by how fast it is to get my solution solved. And I actually would prefer to go with the AI agent versus someone who can deal with me. In person, because I feel like it's faster.
Lisa Avvocato: Yeah. And you have it more on your timeline. Like trying to do something in between business hours is typically impossibleible. So usually when I'm trying to get support, it's like 9 o'clock at night when the kids have gone to bed. So being able to have that flexibility with an AI agent that actually knows what they're doing and talking about is far preferable than sitting here trying to figure out, okay, how do I schedule in 15 minutes to call so and so to an automated menu Y.
Mika Yamamoto: But sometimes it's not 15 minutes. So I think that previously you used to have to mentally prepare for a time slot and frustration of how to get your question solved and now you can do it in the day while you're maybe possibly doing other things, which of course we're not doing right now. But I think that how effective the AI agents have actually become is so much nicer than having to again carve out a chunk of time and get yourself much you prepared for frustration. So it's a true testament to how far we've come from an, industry standpoint and how much we can get from our AI agents.
Rob Stevenson: Now that definitely was my experience over this holiday season. I had much better experiences with the chatbots, for example, than previously. But it's not all good, is it? What comes to mind is some of the backlash that Coca Cola faced in response to their AI generated ad. There's this very obvious moment in the ad where part of the logo is just smushed and weird and there's no way that it was an accident. So there was some backlash to this. And so I'm curious to hear you both weigh in on. All right, yes, the tools are, they are expedient and they are great in some ways, but it's not necessarily like, okay, just the AI is going to be great and accepted no matter what. M Mik, what do you think was responsible for some of the backlash with Koch'ad?
Mika Yamamoto: I think that we're sitting there and we expecting so much and we expect AI to be absolutely perfect and there just aren't perfect scenarios, regardless of whether you're looking at human beings doing a task or computers doing a task. And I don't think that, you know, I mean, you saw presuppositions when you started to see companies start to use AI and customer support solutions. You started to see analysts and other pundits say, like, that's the end of the agent and all customer service people are going to be going away and I just don't see it as being, a zero or a one. And it's a comulbination of how people are going to engage with AI just as they did when the Internet was invented. It wasn't like all jobs changed. We just didn't replace all human beings. It's just that jobs changed. And so I think that the backlash was just going, ah, it's just not as perfect as we thought it was, but nothing is. And so now I think we're trying to settle in, in the industry and figuring out, look, what's the right combination of technology with human beings? And they're complementary, for sure. But understanding where those boundaries sit is what we're all figuring out as we adopt AI, whether it's to help with marketing, messaging, or whether it's to help with customer service scenarios.
Rob Stevenson: The curious case of the ad, too, was that the ad itself was fine. Like, it worked, and it was more about the fact that it existed at all. And so the AI, as it was intended to use to have this generative video experience, it did work, but it was more about how it was wielded, which is the human decision that was made like a person, a team of people more likely decided this was going to be good and they missed the mark. So, it didn't feel like it was a javelin through the heart of AI hype, you know, the backlash. It just felt like it was just a one swing and a miss amongst what it's going to be a million swings. Sorry, at. Lisa, I kind of interrupted you there. What were you going to share?
Lisa Avvocato: Yeah, I was just going to say I think part of the backlash too, is it just. It felt like it was a step back. Right. Like, when we think about AI, we think about all of these technological advancements, and we think of moving forward, moving forward, moving forward. And that specific, I just felt like a little bit of a step back. And I think it really goes to show the importance, as Mika touched on the importance of having humans in the process. Right. Like AI, whether it's generative AI for creative or content or whatever, it's not gonna replace humans. It's gonna augment our workflows, at least for now. Right. Like, who knows what their future is gonna hold? But we're constantly going to be evolving, and humans are constantly going to be thinking one step ahead, and it's really up to us to figure out how we can train and guide AI to maximize these creative outputs. Right. So we're talking about retail and advertising right now. It's really great as kind of a foundational storyboard partner or a thought partner and helping get those creative juices flowing. It's really great for taking your mind content dump. right. And putting it into just the concise like impactful sentence. But it's never really going to truly replace that kind of strategic aspect behind it. And I think this was really where Coca Cola's big miss was right is I think they were so focused on let's be the pioneer. Let's be the first one to get this generative AI ad out there that they didn't really go back and do that quality assurance check. And looking at, hey, the logo is not right here. Like that's Coca Cola's brand 101 is logo placement. Like I'm sure they have a 20 page guidebook right on what their brand guideline is just for that logo. And then you know, some of the other pieces that just felt a little unnatural. Right. That's I think where the humans could have stepped in a little bit, refined those pieces and the ad would have been received in a much more positive fashion.
Mika Yamamoto: I think just turning it on and saying, look, we're just go going toa put AI at the wheel. We're just not ready for that yet. And I think that you're right, Lisa. in terms of the training of the models is key. We're looking at right now a copilot scenario and using copilot to be able to help enable our sellers inside freshworks. And we are training the model and we're seeing we've got a bunch of subject matter experts that are looking at what the output is had. Do we actually just gone completely with the model and allowed it to actually push communication to our customers? Right now we've got the co pilot trying to make jokes and they're not very good jokes and they're pretty terrible. And the language actually isn't on point with what we want to say and sometimes it's technically inaccurate. And so had we actually just completely trusted the system and not trained the model with human beings, we would have done our customers a dissevice and our company a disservice. So there is this learning curve in terms of at what point do we actually push the AI agent to act on our behalf and it can't be too early otherwise, you know, there's reputational risk, there's experience risk, there's customer satisfaction risk and at the worse end there's obviously security and privacy risks. So I think there definitely has to be some oversight and checks. But I think like with any new technology, we're all figuring out what that is. And so this was kkeh's foray into the. We see the same thing happen with self driving cars that get into accidents. Bad things happen when technology as they do to, you know, just when humans act on their own accord. So I think we're all testing and learning right now. And that's really true with AI for sure.
Rob Stevenson: I love the idea that the model is making bad jokes. And if you think I won't report an AI to hr, you got another thing coming. I can do that so easily and quickly. In any case, I'm curious to hear from you Mika. Ah, when you think about the ways that your customers or freshworks are deploying AI in the retail space, are there ways that are exciting and particularly interesting or maybe surprising to you?
Mika Yamamoto: You know, when we look at and when we survey our customer base, retail is actually one of the fastest adopting industries out there in adoption of AI. And there are so many vectors that we can go in in terms of fraud detection, in terms of inventory management. I think the biggest, most exciting part is, is you talk about, you know, Rob, you were talking about just the, you know, the super bowl of retail is actually during the holiday season and this is when you get a lot of seasonal workers. There's a ton of competition. And this is what makes or breaksail the season, is what makes or breaks retailers. And so making sure that no one abandons their cart either physically or digitally is really important that they fill their cart with as many things as possible. And so I think the thing that we're seeing is that the adoption of AI and retail is incredible. The ability to actually serve customers very well in a time when retailers have to rely on seasonal employees helping out with that is actually quite remarkable. And what I mean by that is that with copilot scenarios, if you're dealing in digital scenarios, but even in physical scenarios you can have a copilot scenario that can have access to inventory, that can be personalized based on the customer who's interacting. And so you can be faster, you can be more consistent with the customer that you're dealing with. And that helps in terms of bringing someone up to speed. That again, frankly, is a seasonal worker dealing in an environment where the pressure's on to perform not just for the worker but for the retailer themselves. And so the ability to handle more customers far more effectively and more consistently is really impressive. And we're starting to see that. We saw that this season More than ever. And I think that as we see innovation evolve, as Lisa said, it's happening quicker and quicker. I think we're going to see that happen even more in 2025 during the holiday season.
Lisa Avvocato: Yeah, I think there's really two brands that stick out in my mind that to me are some of the clear winners in AI for 2024. And one is Target, which launched the exact application mica that you were talking about with their co pilot in stores. And it's great for these seasonal workers to just be able to go onto the device and get the information that they need to help their customers. I think Nordstrom is the other one who launched their new app right before the holiday season and just really double down on hyper personalization and gift recommendations. To your point, Mika, like we don't want carts to be abandoned. We wanna maximize order value and we really wanna make sure that people are getting the recommendations that are best suited for their personality or what they're looking for. And so Nordstrom's app just, they spend an incredible amount of time, you can tell on building out the models behind their search relevance or their personalized recommendations, but also their style swipes tool. Right. And just leveraging all of this trend information from some of their stylists internally and then being able to make that available to the masses in more of a personalized way through generative AI. And I think as we get into 2025 we're gonna see more of those types of applications coming to market. And I as a shopper am here for it. Right. Cause don't always have time to just be parsing through the websites to look for whatever outfit that I want. So the easier it is to search and find like one of the things that I'm really excited to eventually have happen is like I'm going to a New Year's Eve party. It is semi casual. I want something that is black and a line frame or whatever and you just kind of put your criteria up and then all of a sudden you get these five recommendations versus having to go through and like click the decision.
Mika Yamamoto: Trees to narrow and knows what you've bought before. So and this goes really well with this necklace and this pair of shoes that you already own. I think is brilliant as well. And it can pattern match based on what you've already bought just to be able to start to be your personal stylist. I think also on the back end is where we see AI play a big role also is in terms of the ability to actually process if things didn't work out. Like, I hate shopping, but I love doing it. I hate shopping physically, but I order more than I will keep. But the process, I think on the back end has become a lot more seamless in terms of what it's like to actually process a return. This is another instance where you typically have braced for impact and thought of this, the absolute hassle of returning things. And now it's become so seamless that in terms of both reering the for the retailer to be able to take return items and put it back into physical inventory is one thing, but also making it a lot easier for the customer to be able to return things even if things were a gift just makes it a lot more seamless for us as consumers, but also for the retailer themselves to have to not hire a bunch of seasonal staff in January to handle this onslaught of returns, because it's just become a lot easier for them to be able to handle the complexities of being a retailer, frankly.
Lisa Avvocato: And I mean, I think too just to kind of touch on managing the return process, right, like having AI help you with your demand forecasting and inventory and knowing how much that you're going to be coming back is gonna help significantly with the profitability, right? And when you layer in the better recommendations and even gift recommendations, right, you're gonna reduce some of those returns along the process. So there's just a lot of different ways that it can help ye.
Mika Yamamoto: And I mean, there's personalization as well. Retailers like Amazon process or consider like, how much has Lisa purchase versus Person X and if she is actually pursuing a return, we're actuallynna give her less hassle because we know she's valued at this level with us. And so we're just gonna say, you know what, don't even return it. Just keep it. We're just gonna refund your account and you can just carry on with your dayuse. It's cheaper for the retailer not to accept the merchandise back or even worry about accepting it. And they know that you're a frequent enough shopper that you know they'll get it up back in the front end by offering you that type of convenience. And so that type of personalized service on the front end to help you pick something, but also on the back end to help you when you have an issue just absolutely is something that AI definitely enables because it gets to know your patterns and what you normally do. So it helps both you as a customer, but it helps definitely on the back end just to reduce or eliminate m complexity from, a Retailer standpoint.
Rob Stevenson: Yep, that is well said. Mikei, I want to follow up with you on this notion of who is going to be experiencing these tools because there is of course, the co pilot, the use case for an employee, I think is well established that they can have more live update access on data that they can ramp up and have a faster onboarding. When you think about the way it's going to be experienced by consumers, I suppose the main thing is more personalized recommendations. But what can people expect to see in the digital storefront or maybe even in person as the AI tools become not just in back of house, but a front of house kind of experience?
Mika Yamamoto: I think that what we can expect to see is more of our scenarios or use cases getting handled by AI agents. So if you just think about how much we've evolved in terms of being able to both think. If you think of an airline scenario, typically you could just find out cursory information about your ticket using a chatbot. Now we've got AI agents that can then act on your behalf, cancel a ticket, book another flight and know what your balance is in terms of what's going back to your credit card without ever actually having to deal with an agent either on the phone or online. And so I think there's more personalized scenarios that are going to be made possible at 24 hours a day. To your point, Lisa, you know, you can do it when the kids are in bed versus having to wait when, you know, a human being can actually get on the phone with you. Which again, is a massive point of convenience and efficiency for the retailer. It's convenience for the customer point of efficiency for the retailer. And so I think that we'll see more efficiency gain on the back end. I think customers will be able to have more choice in terms of what will happen autonomously with an AI agent. I think what we're seeing now is more multimodal connections with customers. And what I mean by that is that if I think if we continue this example of a return is that I can, you know, what we're going to anticipate and what we're actually starting to look at now is if you can take a picture of like a smashed vase or you can take a picture of a broken watch, you can actually upload it into your conversation that you're having with an AI agent, it can recognize that it'something that you've previously purchased, or it can recognize so it might be damaged or it might be the wrong item, and it actually can recognize that, recognize that in your order history and then process your process what it believes your request to be without, again, ever actually talking to another human or without you having to describe what it is your scenario is. And so I think that it'll be more effective and efficient. And the fact that it'it's multimodal and can start ingesting pictures and video is going to be extraordinary for, again, the customer itself, but also driving efficiencies for the retailer.
Lisa Avvocato: Yeah, I think we'll see a lot of. With the growth of multimodal and a lot of these models in various different ways, too. Right. Like, when you think about from, the recommendations standpoint, I could just upload a video explaining what I'm looking for instead of trying to type it. Or actually was talking to someone the other day about the fragrance experience and how you could answer a few questions about how your week is going and that you can digitally create this fragrance that's suitable for your mood. So if stressed out, maybe it's, you know, some combination of lavender, if you're happy, you get something else. Right. And so you can get these more personalized fragrances that are built for you, for your body, and no one else has that kind of same formula. So it's going to really push us into this new realm from apparel standpoint, but also from a beauty standpoint and our products and getting things that are unique to us.
Mika Yamamoto: You on the beauty front, there's the sense there's also a look. So if you follow a certain influencer, but that has a different color profile than you. But it's like, I love Lisa's look. I just. I'm going to screenap her picture on Instagram. I'm going to then upload it to my app with the retailer and I'm going to upload my own picture to be able to see that look. I've got dark hair, I've got darker skin. How do I get Lisa's look with the profile that I have and help me pick all my products so you add them to my cart for me, versus having to go, what color is that of Lisa's? And trying to pick the right color and tone. You can just, immediately say, hey, here's what you need. Here's your kit of parts that you need to get this look based on what you've bought before and then based on what you look like today. And that can happen without having to search and enter a ton of information just by uploading two pictures. And that's not too far away in terms of the ability to enable that to happen. And not just for the massive retailers that have a ton of IT staff and that have a lot of people who work on a ton of economists and data scientists, et cetera. This is becoming more accessible to just to small businesses, which I think is extraordinary just because it kind of levels the playing field in terms of what these small businesses can make happen for their customers, which is, I think, really exciting.
Rob Stevenson: As a consumer, it is definitely very exciting. You imagine instead of having to, I don't know, hire a personal shopper or spend hours and hours sifting yourself, you could, I don't know, plug in your Pinteresto board into a Nordstrom, for example, and then I'll be like, oh, here's, here's how to look like Ciza. Or in this case, like, here's how to finally get that Lisa avocado look that we've all been finding for. So it's definitely very exciting. We are creeping up on optimal podcast length here, but before we wrap things up, I just wanted to ask a less technical question and a more just organizational one. And here Lisa and Mika, both of you sort of op pine on it. We are in this region of just complete AI hype, obviously. And it'the most obvious thing to imagine is a CEO or a board coming to you and saying, you know what, but where are we going to implementing AI? No one wants to be caught out not trying and not attempting to implement AI into the organization. So I'm just curious for both of you, how do you go about being consultative to the business and making sure that you're not on a wild goose chase when it comes to where to invest?
Mika Yamamoto: It's a very real scenario. Just because I do deal with our board all the time and I'm accountable for marketing and sales and support and actually putting input into our products development. And so there's a lot of vectors that I'm asked about in terms of how are we leveraging AI because we offer AI solutions. But the question I get a ton is how are you using AI to be able to drive more, to be more effective within the domains that you're accountable for? And so I'd say that one thing is that I've realized is like mandating use of AI isn't particularly effective, but enabling people license to be able to drive proofs of concept and experiment with it. And so it's just fine line that we're need to strike between how much do you regulate use of AI and how much do you actually allow for experimentation and so people are going to experiment in your environment whether you let them or not. And so allowing for that and making sure people are driving proofs of concept and making sure they're doing so in a secure way and providing guidelines on what to do with customer data and what not to do with customer data is important. And so the accountability to not mandate, I think mandating use of AI, it feels somewhat punitive and doesn't feel very inspirational. So what we've done is just said, look, go figure it out, but work it within the constraints of not allowing access to personal or private information. And that's actually played out really well for us in the sales environment and the marketing environment as well as among our engineers and our support individuals to say go be more efficient and effective, but not mandate it has helped a ton and we've seen a lot of, a lot of uptake and excitement for taking advantage of AI throughout all of our functions as a result.
Lisa Avvocato: Yeah, I think one of the things that we've seen just with the retailers that we've worked with is picking the right type of AI for your application. And this is, you know, especially when it comes to the C suite and the executives and they're kind of all about the generative AI hype and the agentic AI hype. And we need to do this and this and this. You don't wanna spend a lot of resources building these highly complex models when just a traditional model or like traditional AI that's more outcome based will do. And so that's, I think one of the biggest cautions and, or mistakes that we'll start to see coming out not just in retail but in any industry is over complicating the models for a simplistic use case. Now there's obviously a lot of times when you need these highly complex models, but it could be a huge amount of resources, a huge amount of budget, a huge amount of time. It can cost you that competitive advantage from a time to market standpoint by over complicating your model versus working with the right partner to help you hone in on that model strategy, that data strategy and say, okay, you really need this small piece right now. And if we wanna build down the road and get a little more complex, you can, but, but we're just, I think gonna see a lot of even startups, right, that are overco complicating certain models and not surviving because they're too big for the value they provide.
Mika Yamamoto: What's hard though is that everybody is because the hype is so high is understanding what's reality and what's hype.
Mika Yamamoto: Some providers are actually really good at keeping that range tight in terms of they will say what they do and mean what they say versus overselling or overshooting. And so there's so many, there's nary a bendnder out there that isn't saying that they're driving and delivering AI. So really understanding what is complex and what is more simple and what's necessary based on a business's needs is really hard to navigate. And so I think a lot of it is talking to peers and having idea exchanges. And if you're dealing with a vendor asking to actually get peer references to say can I talk to another retailer and can you tell that is of my size or that does a similar business model to mine so that I can actually hear what corners they bonk themselves into and whether they actually are realizing the value that is being espoused by ex vendor. So having conversations with peers I think is absolutely invaluable to figure out what's hype and what's reality and what's too complex or what's not complex enough to be able to deal with a certain scenario, because we're all trying to figure it out, vendors as well as just practitioners and operators. And so I think these conversations are really necessary to judge what's helpful.
Rob Stevenson: Absolutely. And that is great advice, particularly for the really technical folks out there listening. The days of you coding or doing your own data science in a, vacuum are long gone. You are now a diplomat for AI. I hate to tell you, but that's just their reality. You want to have influence and make sure the tools are developed and deployed correctly. You need to go out there and conduct some education, speak to some stakeholders. So that is great advice from both of you and I don't think we're going to find a better way to book in this. So at this point I would just say thank you, both of you, Lisa and Mika, for being here and sharing your experience and expertise. I definitely learned a bunch from you both today, so I really loved having you on. And I would just say thank you also to the folks out there in webinar podcastland who tuned in. Thanks for being here one more time. Lisa Avocado, Husband Lisa Avocado, Mika Yamamoto Hasband, Mika Yamamoto. I've been Rob Stevenson and you've all been amazing, wonderful AI practitioners. Have a spectacular week and hey, see you next time. Thanks for being here. How AI Happens is brought to you by Sama Sama's agile. Data labeling and model evaluation solutions help enterprise companies maximize the return on investment for generative AI, LLM and computer vision models across retail, finance, automotive and many other industries. For more information, head to sa.com.