IN CLEAR FOCUS: Lisa Avvocato, VP at Sama, explores AI in retail and how it’s reshaping customer experiences. She outlines five key areas where AI drives revenue and cuts costs: personalized recommendations, search relevance, virtual fitting rooms, inventory management, and AI agents. Lisa contrasts luxury retailers Nordstrom and Bloomingdale’s with Amazon, emphasizing brand-aligned implementation. Hear why Lisa advises marketers to prioritize solving specific problems over chasing AI trends.
Episode Transcript
Adrian Tennant: Coming up in this episode of IN CLEAR FOCUS
Lisa Avvocato: This bridge between physical and digital is happening, right? We still want to go shopping, but there are a lot of times that we don’t have time to do that. And so creating this metaverse almost where you can go in, you can try on clothes, is a huge area that AI is going to help not only drive revenue, but then decrease the cost of returns.
Adrian Tennant: You’re listening to IN CLEAR FOCUS, fresh perspectives on marketing and advertising, produced weekly by Bigeye, a strategy-led full-service creative agency growing brands for clients globally. Hello, I’m your host, Adrian Tennant, Chief Strategy Officer. Thank you for joining us. According to a January Salesforce report, consumers and retailers leaned into the use of AI and agents to enhance holiday shopping experiences. AI-driven product recommendations and personalized order support influenced $229 billion, or 19% of all online orders in the US. During the 2024 holiday shopping season, retailers saw chatbot usage increase 42% compared to the previous year. However, a study from Riverbed found that only 40% of retailers feel prepared to implement AI, citing data quality and scalability as key challenges. Our guest today is an expert in AI implementation, retail technology, and marketing strategy. Lisa Avvocato is Vice President of Global Marketing and AI Community at Sama, a leading provider of data annotation solutions trusted by retail giants, including Walmart and eBay. With over 15 years of experience in enterprise SaaS technology, Lisa specializes in the intersection of AI, machine learning, and digital transformation strategies. Her insights on AI in retail and marketing have been featured in publications, including WIRED, Advertising Week, and Total Retail. To discuss the evolving role of AI in retail and marketing, I’m delighted that Lisa is joining us today from Nashville, Tennessee. Lisa, welcome to IN CLEAR FOCUS!
Lisa Avvocato: Thank you. I am excited to be here today.
Adrian Tennant: You’re Vice President of Global Marketing and AI Community at Sama. Could you tell us a bit about what Sama is, and what your role there entails?
Lisa Avvocato: So Sama is a leading provider of data strategy and data labeling solutions for enterprise companies, but specifically around retailers. So, what that means is we help retailers build their AI models to drive various customer service applications like search relevance, personalized recommendations, and virtual fitting rooms. All of these are based on data. Without the data, you’re not going to have a great customer experience. And so we work with customers to build all of those components. And then my role as the VP of Global Marketing is obviously to foster conversations around how AI is transforming retail and what retailers can do to drive a competitive advantage. We’re constantly seeing bankruptcy announcements or closures in the news. And we’re really at a crucial time in retail. You either innovate or you’re going to start liquidating. And AI is the game-changing technology that will be the difference between those two scenarios.
Adrian Tennant: In your career, you’ve had a front-row seat to the evolution of AI in retail and marketing. How has the landscape changed from the early days of personalized advertising to where we are now, would you say?
Lisa Avvocato: So I think everybody inherently can see some of these changes, right? When personalized advertising specifically first came out, it was that same pair of shoes that just followed you around the internet because you took a look at it on your favorite retail site. So Nordstrom is obviously one of my favorites. So I would have these shoes that would just follow me around. And we evolved a little bit to having recommendations based on what others were buying, right? It wasn’t just, you looked at these shoes, here’s these shoes. It was others are looking at these shoes. And now today we’re moving beyond that to being able to create these hyper-targeted personalized advertisements where it’s not just a picture of the product, it’s the picture of the product in that environment. So when I’m looking at these pair of shoes that maybe is for a wedding or maybe it’s for the beach, I am seeing them on a model on a beach or at a wedding or at a fancy event. And so I really start to understand and can get that view of how these products are going to fit into my life. And that ultimately is just a much better experience as a customer. And ultimately for the retailers helps generate more revenue because you’re really getting into that hyper targeting and the next round is just adding in this additional layer of data and moving from still that kind of one to many segmentation to, I can build the same advertisement for five different people, swapping out a different model, swapping out a different product and getting that extra layer of granularity. So it’s really exciting to see how far AI has come specifically in advertising and personalization over the past relatively short amount of time.
Adrian Tennant: Yes, interesting. Now, you believe there are five key areas where AI offers benefits for retailers. Lisa, could you walk us through these areas and explain why they’re significant?
Lisa Avvocato: So there’s really two key buckets that these all fall into, right? And the first one is driving increased revenue. That’s top of mind for every single retailer out there. And so there’s a few different applications that we are seeing a lot of success with through AI that’s helping drive this increased revenue. So the first one, obviously, is what we were just talking about, the personalized recommendations, right? It’s using AI to hyper-segment, hyper-target, but also build those personalized ads, generate models, generate voiceovers, generate backgrounds, to make a much more comprehensive experience when you’re providing recommendations. And then this is also for gifting and things like that as well. So Nordstrom did a really great job this past year launching their new app ahead of the holidays, that was just providing enhanced recommendations. They had their style swipes that was kind of taking recommendations from stylists and showing, you know, “Here’s some of the top trends,” and things like that. So that’s personally one of my favorites as a busy working mom trying to synthesize all of the fashion trends and just get very curated recommendations of “Here’s different types of clothing that you can buy.” The second one is search relevance. So again, you know, when I have my list of things that I need to purchase, typically looking for something very specific. And in the past, you would, you know, have to go on to the left side navigation and say, “I’m looking for maybe a shirt.” And maybe you would get that next layer of granularity between t-shirt three quarters length, long sleeve, things like that. But we’re getting to a point now where we can actually type in, “I’m going to the beach so I’m looking for beach shirts, or I’m going to a PTO meeting, right? And I need to look like a PTO mom.” Then getting those very specific search results based on what I’m looking for. All of this is driven through foundational data. As the machine learning models learn, they’re making those ties back to, this is beach attire, this is school attire or PTO attire, this is cocktail attire if I’m going out for the evening and things like that. I think the next big piece of driving revenue is then coming into virtual fitting rooms. So we just know that this bridge between physical and digital is happening, right? Like, we see it a lot with all of the bankruptcy announcements as well, that, you know, if you cannot truly bridge the physical and digital together – yes, we still want to go to the stores, we still want to go shopping. But there’s a lot of times that we don’t have time to do that. And so creating this metaverse almost where you can go in, you can try on clothes is a huge area that AI is going to help not only drive revenue, but then decrease the cost of returns, which kind of ties into that second big area that retailers are looking to do. But again, it’s not easy, right? You have to map all of these different body types. You have to map all of the different pieces of clothing so that you can overlay them together. And that just requires a significant amount of data. So that’s really the first bucket. The second bucket is reducing operational costs. We all know margins are super thin within retail. Even getting a half a percent reduction in your operational costs has a huge impact on bottom lines. And this is where things like inventory optimization strategies come into place. It’s advanced mathematics that are telling you how much to order for a season and what sizes and things like that. And then the next big area that we’re going to start to see is the AI agents. And I think this is going to play a really big component into returns management as well.
Adrian Tennant: Sama works with major retailers like Walmart on data annotation. Lisa, could you explain what data annotation is, and why it’s crucial for effective AI implementation?
Lisa Avvocato: So the data labeling aspect of data annotation kind of fits into this middle piece of building your models. So the first step that really has to happen is understanding what the business objective is that you are trying to solve. What tasks are you trying to automate? And that’s where Sama would help to build that foundational data strategy. Then you move into the data annotation and it’s kind of what I was saying in the virtual fitting room aspect is you’re looking at an article of clothing, for example, and you have to annotate all of the points. This is the sleeve, this is the chest, you know, this is the waist, this is the neck. and correlate them to different parts of the body so that when you take that same t-shirt and a person uploads their picture, you can overlay it and you can match the certain points on the body, right? Like you don’t want a shirt that is upside down, for example, right? Or off to the left a little bit. It’s very difficult then to see what that would look like on you as a person and your whole virtual fitting room application would just be a poor user experience. So that is why it’s so important. And that’s kind of the initial piece. The next piece is then evaluating those model outputs and making sure that this is correct. So we’ve went ahead, we’ve overlaid these images, it’s having somebody go in and say, yes, this is correct. No, this is not correct. This is what’s wrong. And that helps retrain the model so that these algorithms get even stronger. And ultimately, it just comes down to the customer experience, right? Like if the customer does not have a good experience with their virtual fitting room, they’re likely going to abandon their cart. And we don’t want anybody to do that.
Adrian Tennant: Generally not in retail, no. In a recent LinkedIn post, you discussed AI agents and their implications for customer experience. So I’m curious, Lisa, how do you see these AI agents transforming how businesses interact with their customers?
Lisa Avvocato: Yeah, I think there’s going to be a really a huge transformation. So if we look back, one of the examples that I really like to give to show the difference between a standard chatbot and then an AI agent. So we’ve been so used to this standard chatbot experience that’s basically just working off of predetermined responses. They are really meant to cut operational costs for the retailers. They’re not actually meant for customer and the customer experience, right? That’s not their end game. But AI agents, we’ve all heard about kind of this acceleration in large language models, you know, chat GPT, there’s so many more that are coming out. And that has kind of pushed forward how AI is able to communicate with individuals. They’re now able to understand what the problem is and then adapt to help. And this is where we’re going to start to see this next wave of AI agents support customer service. Because now customer service is going to be available 24-7. You’re no longer going to be limited to where I have a complex problem. I know I have to speak to an agent and I am blocking out an hour, sometimes more, on my calendar during business hours or when the kids are here and trying to figure that all out. Because these AI agents are going to be able to understand and respond to more complex problems, they’re also going to be able to understand your emotions, sense frustration, how you are feeling, and respond accordingly. You’re going to have a much more humanized experience and we’ll see people start to rely more on these AI agents and chatbots to some extent to solve their problems versus having to have a phone call and wait in line and that other set of frustration.
Adrian Tennant: Let’s take a short break. We’ll be right back after this message.
 | Adrian Tennant: Each month, in partnership with our friends at Kogan Page, the Bigeye Book Club features books by industry experts and thought leaders. Our featured book for March is “Effective Brand Building: Unlock Growth with Strategy, Insights, and Measurement” by Andrew Geoghegan. This comprehensive guide gives marketers practical wisdom for creating brand-building strategies that deliver short- and long-term business growth. The book includes real-world examples from leading brands, including McDonald’s, Apple, and Unilever, to illustrate how asking the right questions creates the optimal conditions for effective brand building. As an IN CLEAR FOCUS listener, you can save 25 percent on a print or electronic version of this month’s featured book, when you order directly from the publisher at KoganPage.com. Just enter the exclusive promo code BIGEYE25 at checkout. This code is valid for all KoganPage titles, including pre-orders. Shipping is always complimentary for customers in the US and UK when you order directly from KoganPage. So, to start unlocking reliable long-term brand growth, order your copy of “Effective Brand Building” today at KoganPage.com. |
Adrian Tennant: Welcome back. I’m talking with Lisa Avvocato, Vice President of Global Marketing and AI Community at Sama. You mentioned Nordstrom’s earlier, and you recently wrote about Bloomingdale’s success in luxury retail. How do you see AI enhancing the shopping experience while maintaining the human touch that luxury customers really expect?
Lisa Avvocato: Yeah, I think both Bloomingdale’s and Nordstrom do a really great job here in understanding what their brand values are and integrating AI into those foundational components. So they both have pretty strong personalization algorithms. Nordstrom has taken it one step further, like I said, with some of these Gen AI components into their app and using all of the advice from their stylists to curate these recommendations and share things on trends and things like that. And that’s really going to be crucial, right? Because people still want to go into the store. They still want to feel that human touch. But when they need that convenience, you still want them to be able to turn to Bloomingdale’s and not go somewhere else. And I think a really stark example is when you start to compare Amazon’s AI strategies for retail with Bloomingdale’s and Nordstrom’s, right? Amazon is all about convenience. It’s not necessarily all about cost, but it’s predominantly cost and convenience-driven, right? You want something quickly, you go in and search for it, and you get it. But it’s typically not going to be the highest quality that you would find in a Nordstrom or Bloomingdale’s. And so what they’re doing is they’re prioritizing the customer experience. They’re giving you those trends. They’re giving you those recommendations based on this is everything you’ve bought in the past. We know that you are a working professional and you typically default to working clothes, but we also know that you like to go out to fancy bars on Friday, Saturday night and get glammed up. And so those are going to be your top two recommendations versus, you know, they know me and I live in athleisure half the time as a working mom. And, you know, those are some of the recommendations that I get prioritized. So it’s really about creating that personalized, unique experience. You don’t feel like you’re getting the same information and the same recommendations that are served to a million other people like Amazon does. It’s Lisa, we know you, we know this is what you like, and here are these recommendations. And so that’s how I see that human touch still maintaining the luxury environment.
Adrian Tennant: Great. Lisa, you’ve been very vocal about the importance of responsible AI development. What ethical considerations should marketers keep in mind when implementing AI solutions?
Lisa Avvocato: One of the biggest pieces is having representative data sets and making sure that you’re avoiding these inherent biases that have just really perpetuated over time. And so, especially if you’re using foundation models, it’s really important to understand how they were trained it’s really important to understand as you are what’s called fine-tuning. So fine-tuning is essentially taking a base model and then making it specific to your company. So Nordstrom would have a different version of a chat GPT model than Bloomingdale’s would. That’s an oversimplification, but it’s kind of an important distinction. And so as you’re going through and using all of your data, it’s making sure that you don’t have major gaps in there. You know, we’ve seen just a lot of kind of high profile, not so much in retail, more so in finance of just inherent biases that are getting perpetuated, especially in high-impact decisions that is detrimental to people’s lives sometimes. So, the foundation of building responsibly is really digging into those biases, as well as just the trust and safety of the models, right? AI has grown so fast that there are still a lot of vulnerabilities or models are a lot easier to hack than cybersecurity right now. So, you know, making sure that you’re doing your due diligence and kind of putting all of those protections into place so that consumers are not getting served inappropriate content on your website is crucial. One of my favorite stories, which, thankfully not anything too terrible, but it was a car dealership in California that their chatbot got hacked and basically sold a car for a dollar and kind of basically ended saying “This is a legally binding offer, no takesy-backsies.” It’s just one of my favorite stories. But those things, that’s part of building your models responsibly is understanding that trust and safety aspect as well.
Adrian Tennant: Looking ahead, what emerging trends in AI should brand and retail marketers really be paying attention to?
Lisa Avvocato: So I think there are two big areas. We have seen the leaders in personalization and search relevance. And if brands are not really paying attention to these trends, investing in these areas, Accuracy of your search relevance and your personalization algorithms matters. It’s the difference between a sale or losing a customer potentially to another brand. And we have to get past these days of good enough, right? Our previous search algorithm was good enough. Our chatbot is good enough. Customers don’t want that anymore. And we’re seeing brands that are really pushing forward with these concepts, and they’re the ones that are going to continue to be the market leaders. Otherwise, you’re going to get left behind. The second piece is understanding AI agents is going to be another really big theme for retailers. But what is important to understand when you start to think about AI agents and what’s called agentic AI, it’s kind of the overarching umbrella term, is that it’s not always going to be needed. So don’t just rush your AI strategy and try and just hop on the next buzzword or trend. Really take a step back and understand what are the problems that you’re trying to solve and then build out your AI strategy. That to me is the most important piece of advice when it comes to emerging trends in AI because we see so many companies just trying to run and they end up wasting a lot of time and resources chasing something that is this shiny object when they could have actually gone in a more simplified direction, got it to their consumers faster and in a better state because they’re not wasting time on building something super fancy. And, you know, again, with the rate of bankruptcy announcements that we’re seeing here, you can’t just ignore some of these things that are happening, right? It’s really time to take a step back, set your AI strategy, make sure it aligns with your brand values, and then move forward.
Adrian Tennant: When we were preparing for this episode, you mentioned to me that companies really need to think about their corporate strategy before their AI strategy. So Lisa, could you elaborate on how marketers can ensure their strategies align?
Lisa Avvocato: Yeah, it really ties back to the number one thing that every marketer cares about, and that’s your brand and your brand experiences. And making sure that when you’re implementing AI strategies, they’re very much aligned with your brand experience and your brand values. And to go back to the Amazon versus the Nordstrom and the Bloomingdale’s example, can you imagine what the experience would be if Nordstrom just copy and pasted Amazon’s AI strategy? It would be a huge turnoff, right? Like it would completely change my view of the brand and probably prevent me from shopping, right? Because that’s not why I go to Nordstrom, right? It’s not, “I need something here in two days really quickly” for whatever need. It’s, “I’m going to an event like my maternity photo shoot,” for example, right? And I want the perfect dress. And that’s the something that I would turn to Nordstrom for versus just mindlessly scrolling terrible options from Amazon. And so, you know, it’s so important to make sure that you’re making decisions that are aligned with your brand. And then when you’re implementing those AI strategies, make sure things like your chatbot is responding in the way that you would want your brand to respond. How Amazon responds to customer complaints is gonna be completely different to how a Bloomingdale’s and a Nordstrom should be responding to complaints. That needs to be infused throughout those chatbots, AI agents, whatever type of AI strategies you’re using, They all need to connect back to your brand values. Otherwise, your customer experience is going to be affected, and that’s ultimately going to be translated into lost customers.
Adrian Tennant: Well, for marketers listening who feel overwhelmed right now by the pace of AI advancement, what advice would you give them?
Lisa Avvocato: I would say that you don’t have to do everything all at once. And especially from a vendor perspective, there are going to be 20 new tools every day that you can use as a marketer to implement into your marketing campaigns and programs and strategies. And it can quickly become overwhelming. So really this just goes back to the fundamentals of understanding what are the biggest problems that I want to solve. And a lot of it too ties back to am I trying to drive revenue? Or am I trying to improve operational costs? For marketers, it’s all about trying to drive revenue. And so it’s then connecting that to your brand values. How do I want to do this? What are the top three things that I am looking to do? Do I want to bridge that physical and digital experience with a virtual fitting room? Do I want to improve my recommendations? Do I want to have a better chatbot experience on the website? And so it’s starting to define those goals. And then it’s looking down into the tools that you want to do. Do you want to use a partner? Do you want to build it yourself? A lot of larger enterprise retailers are going to opt to build things themselves. And this is what creates that intellectual property. So one thing that I think is just as a quick important note or piece of advice is if everybody is using the same model, like if everybody is building their chatbot based on chat GPT or Claude, you’re all going to get the same answers, right? So it’s really important, especially for some of these larger companies, to take ownership over their model so that you can train it to your brand voice. Obviously not going to be as easy for smaller companies to do, but that is a point of consideration when you’re thinking about your AI strategy. But again, it just ties back to those foundational components.
Adrian Tennant: Great advice and a great conversation. Lisa, if listeners would like to learn more about your work at Sama or connect with you, what’s the best way to do so?
Lisa Avvocato: So you can always follow me on LinkedIn. We share a lot of content there. I love talking about retail and AI and retail. You can also visit our website at sama.com if you’d like to learn a little bit more about the products and services that Sama has to offer and how we’re helping a variety of different retailers.
Adrian Tennant: Excellent. Lisa, thank you very much for being our guest this week on IN CLEAR FOCUS!
Lisa Avvocato: Thank you. It’s great to be here.
Adrian Tennant: Thanks again to my guest this week, Lisa Avvocato, Vice President of Global Marketing & AI Community at Sama. As always, you’ll find a complete transcript of our conversation with timestamps and links to the resources we discussed on the IN CLEAR FOCUS page at Bigeyeagency.com. Just select ‘Insights’ from the menu. Thank you for listening to IN CLEAR FOCUS, produced by Bigeye. I’ve been your host, Adrian Tennant. Until next week, goodbye.
TIMESTAMPS
00:00: Introduction to AI in Retail
02:27: Overview of Sama and Lisa’s Role
03:48: Evolution of Personalized Advertising
06:04: Five Key Areas Where AI Benefits Retailers
10:18: Understanding Data Annotation
12:20: The Role of AI Agents in Customer Experience
16:04: AI in Luxury Retail: Balancing Technology and Human Touch
19:54: Ethical Considerations in AI Development
21:35: Emerging Trends in AI for Marketers
24:13: Aligning Corporate and AI Strategies
26:12: Advice for Marketers Navigating AI Advancements
28:13: Conclusion and How to Connect with Lisa
