The new engines of loyalty: Who will drive the next era of choice?

Loyalty is no longer just about points and perks. 

As AI‑powered agents influence consumer decisions, the balance of power shifts—reshaping how trust is built, value is delivered and relationships are sustained. This change is already redefining loyalty across industries. 

Play this audio version of our webinar to explore how traditional loyalty models are being challenged—and what it takes to build emotionally resonant, trusted and AI‑enabled strategies for the future. 

What you’ll learn: 

• The decline of traditional loyalty models: who’s in control—your brand or the machine? 
• What challenges are disrupting loyalty ecosystems? 
• Is loyalty fatigue real? 
• Are brands ready for the consumer power shift? 
• How should you design relevant loyalty strategies? 

Featuring insights from senior leaders at Arcos Dorados (McDonald's LATAM), Saudia Group and Dynamic Yield by Mastercard, this session uncovers where loyalty is headed, who controls the relationship and how brands can stay relevant in an agent-driven world. 

Download the slides and watch the video recording of this webinar.  

Lots of brands claim to be number one… but can they prove it?

At Euromonitor International, we help brands build trust through evidence-based research. Our claim validation service ensures your marketing messages are backed by real data. Stand out in a crowded market. 

Visit euromonitor.com/claims to learn more.

Introducing Opportunity Minded, a new series from Euromonitor International designed for forward-thinking business leaders like you.  Each episode tackles a strategic approach or topic on corporate agendas.  You’ll hear from our experts who share in

Transcript

0:22

Thanks for tuning in to this Euromonitor podcast. Today we're bringing you an audio version of a recent webinar on the new Engines of Loyalty in the age of AI, featuring expert insights from senior leaders at Dynamic Yield by Mastercard, Saudia Group and Acosta Rados, known as McDonald's Latin America. You might hear us talk about slides, charts or on screen moments. If you'd like to follow along or want to watch the webinar. You can find links to the full slide deck and video recording in the episode description. Let's get into it! I'm Charles Peterson and I will be your moderator for today's webinar. Thank you all for joining us today. We have a great conversation ahead with some fantastic panellists we've allowed to get through today, so let's get right into it. Today's discussion is about how traditional loyalty models are being reshaped, and in many cases, fundamentally challenged by shifting consumer expectations, AI enabled engagement, and the growing demand for emotionally resonant brand experiences. We'll be exploring a few key questions today. What are what challenges are disrupting loyalty ecosystems today? Is loyalty fatigue real, and how should loyalty strategies be designed in an AI powered world? But before we get started, I want to give a quick, brief introduction to your monitor and run through a few housekeeping notes for anyone new to your monitor. We lead the world in market intelligence and research into markets, industries, economies, and consumers. We provide global insights on thousands of products and services so executives like yourself can make confident business decisions and explore new avenues for growth, which is also our aim for you in today's webinar. Now, just a few housekeeping notes. You'll see a Q&A box at the bottom of zoom. Feel free to submit questions as they come up throughout the session. We'll address as many as we can. Towards the end, we'll also be running a few polls throughout this session and we would really appreciate your participation. This session is being recorded, and we'll share the recording and slides with you in the next few days. So please keep an eye on your inbox. All right. Let's dive into the insights. Today you're going to hear from four industry experts. First up is Nadia Popova, Euromonitor global head for loyalty. Then Diego, customer acquisition and engagement corporate director at Arcos for McDonald's Latin America. We also have Andre Orca, Cetus VP, global enterprise Unit at Dynamic Yield by Mastercard. And lastly, Cashmere Montoya, cheap loyalty officer at Alfredson Saudi Group, who will be joining us for the fireside chat in the second part of the webinar. But without further ado, going to hand it over to Nadia to get us started. Nadia, over to you. Thank you very much, Charles. As we move deeper into an age defined by AI, where customer behaviour is predicted, influenced and optimised in real time, I think one of the questions that becomes unavoidable is loyalty dead? My personal view is absolutely not. Loyalty programs are being redefined to become one of the most important strategic levers for growth, for customer retention, for creating a long term value. But a lot of these are in and boring. Traditional structures have been failing in meeting the customer expectations and very much achieving a competitive advantage if you want. And as a result of that, a lot of brands and retailers have been repositioning their loyalty programs from a purely transactional mechanism to an evolving a loyalty ecosystem, at the heart of which is real time intelligence, connected data. And going forward, expect it to be powered by the AI in order to deliver that hyper personalised and contextually relevant experience that those consumers are very much in, you know, demanding as part of their engagement with brands. And as these capabilities mature, a loyalty programs are becoming this very important connective layer across the whole path to purchase, creating a unified view, integrating data, interactions, engagements at the heart of which is very much to support that differentiated loyalty offer at scale. But one of the key key drivers behind that need for transformation is very much about understanding the consumer shopping preferences, the channels that they are choosing, and ultimately how they engage with these brands. An e-commerce has been the success story in the last couple of years. Eating share from the retail landscape and with our projections to reach 30% of the total retail value sales globally. Another telling sign is that Amazon replaced Walmart as the number one retailer in 2025. A lot of markets in Asia Pacific are pointing out that e-commerce is becoming a really important category in the retail space, but at the same time, we're seeing another niche developing and growing really strongly, and that is social commerce, at the heart of which is peer to peer interaction. And our projections from the back of our research is pointing out that sales for social commerce will reach $1.4 trillion by end of 2029, but a lot of you allow you to. Professionals would ask, but why is that important and why does this matter? When we talk about a loyalty space? And I think the answer is very much kind of driven by the need of the loyalty programs to become more rather than functioning as a standalone reward system. They need to be evolving into an integrated proposition, integrating transactional incentives with emotional engagement across social media platforms, digital storefronts, and live stream platforms. As we see that there's these are expanding in Asia Pacific and creating a new omnichannel flywheel, at the heart of which is a continuous loop where we have discovery, when we have interaction and purchase. And these are very much kind of driving the importance of loyalty to be embedded as part of that loop where tiers, points, miles are earned, tracked and redeemed in real time. And this is critical. Real time going forward will be the most important aspect of how these loyalty programs are delivering value for customers. Another important aspect that I wanted to kind of look into is the impact of generative artificial intelligence and how it is shaping the e-commerce scape. And the research by our e-commerce team is pointing out that there is an exponential growth. Of course, it's growing from a low base looking at the performance for last year in 2025, but definitely illustrating a shift away from the traditional product discovery towards a really strong referral traffic to e-commerce and very much securing strong conversion rates. We're also predicting that by 2029, AI influenced in commerce is going to record over 700 billion USD. And this is very much on the back of the accelerated traffic conversion. But at the heart of that is very much the importance of visibility, which is shifting from the search rankings from the paid media that we usually associate with the traditional legacy type of loyalty programs toward products and content that can be interpreted and prioritised by AI agents, and that is with the need to create an adaptive loyalty offering that is relevant and at the same time helps to deepen that customer engagement. And we've seen that in the last couple of years and in the last couple of months, a lot of significant macroeconomic technological disruptions have been impacting the loyalty space. Ongoing regulatory structural pressures, and in many cases, still, we're seeing the loyalty programs are still functioning as a peripheral rewards, too, if you want. Despite these pressures, we're seeing that participation in loyalty programs is high. 83% of the consumers aged between 30 and 44. On the back of our dedicated loyalty survey for 2026, point out that they're participating in free to register loyalty programs, compared to 56% in subscription based loyalty programs. So what is that telling us? There is high penetration but uneven monetisation. But participation alone is not a meaningful measure for success. We're seeing that a lot of these loyalty programs are challenged, and that could be costly because they're generating activity without changing the customer behaviour, without driving that frequency of shopping that is so important. And ultimately engagement without conversion impact. And that requires the need for differentiation of your product, of how you interact with your customers in order to deliver that higher value. And a lot of consumers, from what we're seeing, are becoming more selective, less responsive to blanket promotions. They're actively moving away from purely price led decision making. And I would like to open a bracket, probably for the exception of the grocery stores, where still this is quite important for obvious reasons. But going further afield, looking at other industries, that is changing. We're seeing that actually the top reason for participation in loyalty programs is discounts and offers. But when we look through the lenses of the different time frames from 2023. This is showing a decline, illustrating a consumer fatigue. And of course, that potentially is then influencing the margin erosion for loyalty programs. So our loyalty programs really missing the mark. And that seems to be the case because we're still seeing very low reward redemption rate on a global level. But of course, when we look that through the lenses of different markets, different brands, different categories, that varies as well. 19% of the consumers are pointing out that they're redeeming rewards weekly, and only 28% monthly. There is an engagement gap between that participation and the actual usage, and some of the barriers that these consumers are very much pointing out is a mixed bag. It's very much about lack of transparency, low awareness of the point balances. Points are expiring much quicker before customers have the opportunity to redeem them, to name a few. And if you notice, the majority of these can be really addressed with the help of technology of well orchestrated technological infrastructure. And I think this is where AI can have the opportunity, the potential through multi-agent, through autonomous frameworks to that are capable to discover, evaluate, to execute actions, to introduce a new interface layer for loyalty rather than loyalty, being solely confined within the individual app or the or the brand platform, but moving towards an AI mediated journey, which is, you know, delegating decisions, recommendations to these intelligence intelligent agents. And ultimately, it's very much about a paradigm shift where the focus is segment of one. You're not focusing on broad demographic groupings or static segments. It's more about dynamically tracking the consumer behaviour, intent, signals and contextual data. And what we've done in preparation for this webinar, we wanted to compare and contrast to different data points one relating to the loyalty value contribution. And if a loyalty program embraces a different AI tools, for example, for a comparison of product, how is that impacting the measurable value for that program? And we can see that there is a strong positive correlation between those two metrics. In other words, that loyalty programs tend to have much higher measurable value when they're embracing that going forward. And we're seeing also that actually those consumers that are members of five and more loyalty programs are stating that they're very much more open to embrace and operate and use these different AI tools in order to have a better brand trustworthiness, to have better promotions, to have summarised consumer reviews in a much quicker and efficient way. But look what's happening for those consumers that have only one membership. I do admit that is probably much rare profile that we can see today, but they are saying, well, they don't see a benefit of using AI, and this is where I think we can most certainly plug in the importance of AI. Loyalty programs can help present the brand performance. If these AI capabilities are being embraced in a proper way. But these need to be done kind of intertwined into the whole strategy overall, rather than being done separately. And what we see here on this slide is a couple of brands that have been actively embracing AI as part of their proposition. ARCore, which is one of the leading hotel chains, has been launching the launch their loyalty app on ChatGPT in January 2026. And look how much traffic they were able to secure on that platform, but also the active loyalty users they have been able to engage on a much more efficient way. We can see equally that Walmart, another big player from the retailing space, have been proactively introducing different developments in that space since 2025. Individual customer pages as part of the capabilities where you have the browsing, the shopping history, having and launching their own AI agents supporting that process. And, you know, compared to the brands that we have been tracking, look and illustrated on this slide, how much they're dwarfing in terms of the loyalty program traffic share, but also in terms of that active usage. In many cases, we're actually seeing the traffic through chat. GPT one of the kind of platforms is converting at approximately the rate of 1.5 to 1.8 times the rate of the traditional organic search. So the next year is very much that fight between or that battle, in a sense, between the agent of the customer and the agent of the brand. But that requires loyalty programs to also be a little bit more machine readable, that are enabling that autonomous system to discover, to evaluate, to compare these offerings on behalf of the users. And that essentially can help reduce friction, but also help in terms of lowering that card abandonment, which is so problematic for many brands. So coming a full circle at the beginning of asked this question is loyalty. That most certainly, in my view, not, but at the heart of which should be to develop and ensure that there is a strong, improved layer of trust with your end customer to create a foundational operational layer. If you're integrating AI as part of your strategies so that these consumers have the visibility, how you collect, how you store, and how you protect their data going forward. And now, I would like to allow Diego to talk about how must make downloads is operating and shaping their loyalty strategy. Everyone. And thank you, morning for the invitation that you showed us how loyalty is entering a new phase. And it's clear today that it is no longer just about points, value or breakage. It is about choice in an AI driven world. And before talking about this, let me give you some context about all in Arcos Dorado McDonald's Latin America, we are the largest franchisee of MacDonald's globally and the leading chain in Latin America. We operate in 20 countries with more than 2300 restaurants. So we serve millions of customers every single day in a highly competitive industry and in a very challenging region. And there is something really about the McDonald's brand in Latin America, inner markets, price matters and competition is fierce, so we have to double down in experience and brand connection while doing so. Millions of customers have built personal stories with their brand, from their first Happy Meal memory to even get it getting married at a restaurant. This is not an exaggeration. It actually happens. So in articles, we would like to think that we are not just a place to eat, but we are part of people's lives and this is what makes loyalty special for us. It is not just about transactions, but it is about making that relationship stronger. Most of you may have seen or slogan I'm loving it. What we have learned with loyalty is that a program should not try to reinvent or brand promise. It should try to amplify it. So my first advice is take whatever your brands are really stands for and elevated with your loyalty program. While doing so, the new engines of loyalty become powerful. They will allow you to extend your brand into something more personal and more meaningful. The role of loyalty in an AI world is not to create a brand promise completely new, but is to make the existing one even stronger. Let me bring this to life with one example. Last year we launched our Stranger Things meal, which was already embedded in culture. But we wanted to make that connection personal, not just targeting, but something that people could truly feel. So we use our loyalty, third party data and AI models to translate behaviours into hyper personalised narratives. Each customer experience a version of his orders, highlighting what makes each one of them unique. That specific order that only you make, or the fact that someone miles away ordered exactly the same menu than new at the same moment. We then frame these stories within a narrative inspired by the world of board games in collaboration with tasks, some customers became wizards, some explorers of stone warriors, any of multiple archetypes with unique traits. For each one of our fans, we use real behaviour to create something that felt personal, recognisable and emotional, something that felt uniquely Michael. And that is the shift. With AI. We are not just personalising content, but we're creating emotion at scale. We launch or rewards program in Latin America in 2023. We today we have over 31 million members across ten countries, and we track over a quarter of our sales. This is especially meaningful for us because our business was mostly physical and historically. Historically, transactions were anonymous. What has enabled this rapid progress is a combination of three things. First, a strong digital ecosystem where identification can naturally happen. Second, CRM automation capabilities, allowing us to activate customers across digital channels. And third, really integrating loyalty into everything we do, not just marketing or the app, but as a part of how we approach customers in restaurants. We are already seeing the impact in some of our most digital, digitally advanced markets, where approaching 50% of identifying sales. This means that we are no longer just managing some transactions, but customer relationship chips at scale. We already know that transactional loyalty works. It drives behaviour, but for us, we are seeing customers coming back more frequently. Active customers are up 77% and frequency is up 38%. This is the result of combining multiple engines working together. First, the transactional engine which creates this value exchange and drives repeat behaviour. On top of that, CRM, automation and app experience allow us to stay connected with customers across push notifications, email, WhatsApp and media. And third and emotional and personalised experiences. What we're seeing is that when these engines work together, they do not just right for us, but they drive lifetime value, and the bottom of which is there are two that have been especially powerful for us. First, the fear of missing out former and that exclusive access to new products, brand merchandise, limited time deals and unique experiences such as World Cup final tickets and formula one pattern passes. These create urgency. The second we call digital dopamine, and it is the satisfaction that comes from earning, progressing and being rewarded. This is what drives repetition and builds habits habits over time. When you combine this with a solid transactional engine, you can actually change behaviour over time. The evidence of this change translates into business impact for us. Customers who engage with the program behave very different. They do not just come back more often, but they generate much more value as we move from anonymous customers to those who redeem rewards. You can see a clear step change in behaviour, more visits, a much more lifetime value up to four times more. And that's the key insight for us. It is not about just the reward itself, but about the behaviour that it triggers. What we are experiencing. Inadequacies, the result of loyalty engines working together, transactional rates, action and emotional builds connection. And now this intelligent layer allows us to scale both as we saw in Nigeria. AI is becoming an enabler, especially reducing friction in decision making. And this goes both for brands but also members and customers. We have made progress, but we are only beginning to take advantage AI. As I showed earlier, we are using it to turn the loyalty data into meaningful experiences. We are also using voice AI agents to capture customer feedback at scale. That was not possible before ambiguous personalisation. AI is helping us predict and act on behaviour, optimising point exploration or reward for them. On my perspective, the opportunity for everyone in this webinar is to combine AI, emotion and transaction. And that's how I think that the loyalty system will evolve from programs to systems that shape behaviour and drive choice into better decisions. So far we have focussed on the what and why and now and there will take you into the how, how these engines are actually being built and scaled using AI and data. Thank you. Thank you so much for that very, very interesting explanation. So I am coming from Mastercard where we help thousands of brands all across the world with loyalty, but also customer acquisition and engagement efforts. We want to tell you, when we speak about AI for loyalty, what is it that we mean when it comes to loyalty and the use of AI? We need to understand there is a big revolution in which we are not only using models or predictive AI to understand what the user wants. We have more leverage than ever. Diego was mentioned in many of them. We can offer promotions. We can offer unique products. There's many more and more things we can offer to a user. And at the same time, we have more and more data on these users. Now the big revolution we see happening only this year is that we have generative UIs taking over loyalty. What we mean by this is that we have llms. We literally have a foundational model that build up page for a user in real time, based on which offer, which type of message should be delivered in this moment in real time. This, of course, is paired with the fall of attention span we are seeing all across the world, up to 59% in the last five years. Which means that showing the right example at the right time and not expecting customers to search, it's essential and more and more important. Now, of course, in order to deliver all of this, we need to have, let's say, a real time experience. And we need to provide a customer experience that is already immediate, that allows us to take all the information and execute it with a single view of the customer, a data treatment that many brands have been already working on for many years now. Of course, there are many headwinds, right? We want to focus mostly on the value proposition, ones that Nadia mentioned. Let's say when it comes to the C or CBS, as attention span drops, customers are less and less dwelled into what makes this loyalty program worth it. Now, this is why that true omnichannel engagement, being able to show the right experience to the right user, genetically, whatever they are building full pages for this user in real time is the most powerful initiative we are seeing all across the world and taking most loyalty platforms not only on retailers but also within banks by storm. And it really comes down to execution to be able to select properly an area without siloed data in, execute those pages in real time for each user to provide that conversion. Not only that engagement. Now where we are basically playing is on the orchestration layer, on the decision on execution layer of our loyalty system. We believe that, as we mentioned in this webinar, the core value, the value engine of a loyalty system is still exactly the same. The main difference now comes at the last execution layer. What am I going to show to this user right now? This on all those level levers based on all that data, based on agents that let's say, might be influencing the decision from LMS to to other all the other solutions. Now, this is where the big battleground is. And taking able to individualise this experience is already, let's say, taking many brands forward. Now, a few examples of this are, for example, even on the banking sector where let's say, for example, JPMorgan Chase is already personalising, let's say messaging through machine learning. So it's using AI in order to generate 1 to 1 messages for users at the right time, based on all the wealth of data and information available to them, not to provide a general offer or a general statement, but to have an area that is completely unique to each user. When we say it's unique to the user, we mean based on context, based on time of the day. We mean based on the weather. That was very weather at the location we mean as we are Mastercard, based on the spend of all the users at the zip code location or that user right now, and showing and generating the content that might resonate with that user in real time through LMS, when the users interact with our page, now we see a jump in performance from, let's say, a B, testing based personalisation efforts into genuine capabilities. And the same would be set by CD. That is even going to step further and even creating dynamic rewards. So taking all the possible rewards and not only the messaging or the experience layer, but even of course, being much more flexible on what could be shown to each user. Now, of course, this goes even deeper as a genetic AI is coming, but we need to take into account that let's say conversational uses of AI are just the tip of the iceberg. When we look at Bank of America with their AI financial assistant, or we look at Amazon and many others, let's say with their conversational experiences, these are experiences adopted more and more by the Gen Z, by the new generations in the RV perfected that the biggest revolution is on the usage of this. To understand the intent of the user and to be able to drive that semantic revolution. What does the user actually want? And here is where we see retailers leading not only on the QSR space, but also on the high fashion space. By being able to fully understand large prompts from users, what we see is that conversations of more than five words with brands have been growing more than 80% year on year. So the search of 1 or 2 items, let's say, for match is is gone pretty fast. Now, this semantic revolution means that users are know what they want and they're giving us all the information. But where we can, as stated, build all that together is on using all our models or the internal models built by the brands, which we always recommend brands do, as well as orchestration, in this case from Mastercard to build whole new pages, to be able to bring and create user interfaces that respond to the user in real time. Now this is creating a revolution in terms of CTR. And of course conversion because allows us to show something to the user where they don't need to browse, they don't need to scroll. We are, let's say, showing a much more real time experience. And this is where we see that revolution. Now of course, the final point is as this becomes, let's say, more and more scalable, and as we can generate 1 to 1 experiences, we do see a myriad of brands also adopting internal agents, agents that help them run their own loyalty campaigns. This is for example, this is an example of experience OS in the case of Mastercard or let's say many others in the market in which what we are doing is providing agents that will help the team ideate and plan new loyalty campaigns, generate creative or even development resources to launch these campaigns on any digital surface offline, online, pre-launch, IT or even more importantly, analyse and optimise which type of user wants to see which experience in real time or where should we actually generate more copy, let's say per user. So as you can see, the loyalty program means AI towards the user, but increasingly more AI towards us, towards those running the loyalty programs to make that scalable and to be able to respond to users at scale. Thank you so much. Thank you everybody. That being said, we wanted to open up a poll. And so what you see here, you'll see a poll box in the bottom of the zoom screen. And we'd love to hear your responses on exactly where you stand when it comes to leveraging AI. And while we're feeling this poll, we wanted to now hear from you, the audience. We'd love to hear your questions, your thoughts as we move into the fireside chat portion of this webinar. So in the bottom centre of the screen you'll see a Q&A box. Feel free to share any questions you may have, and we'll open those up during our fireside chat. So we'll we'll wait just a second for questions and everything to come through, and then open up the second part of the webinar which is the fireside chat. So thank you. So with that, I'd love to open up the second part. Now our fireside chat. And we have a number of questions that we'd love to get through. And I'm going to kick it off with you can see our panellists here. You know welcome Kashmiri module. We'll be participating in the fireside chat. And I would love to open up the questions. I'd love to open it up to Kashmir at the start. What does what does loyalty mean for you in your organisation and what are the fundamentals of your loyalty ecosystems, and what fundamentals do you think will change? And what are some of those that you think will fundamentally stay? Stay the same? And I'll open this up to mirror and we'd love to hear from the other panellists as well. Great. Thanks so much for having me on the on the panel. Going to your questions in terms of what is loyalty really mean? I think loyalty at the end of the day, is a value exchange between the customer and and the brand. It is when a customer believes that you're driving an emotional connection, that they will choose you time and time, time and time again. They will forgive you for the mistakes you make, and they believe that it is a relationship that is worth nourishing and nurturing and taking forward in terms of fundamentals. Charles. No, I don't believe the fundamentals of loyalty have really changed. It is still about trust and is still about value, still about recognition, simplicity. And I think consistency. Consistency is a is something that we often tend to miss out. And I don't believe AI changes those human needs. But what it can do is it can give us the ability to deliver at scale responsibly, that is, and by embedding it thoroughly within the organisation. I, I go back to Diego presentation where you had the, you know, the the three elements. You had the emotional layer, you had the rational layer and then an intelligence layer. And I would say that good and strong loyalty programs do all of that. And I would add one more. And that I think is the operational layer, because if you're not consistent and if you're not reliable and you're not creating ease of use, then we're missing a very important element of how we deliver loyalty and how we deliver value and emotions to our customers. Thank you for that, Diego and Nadia, is there anything you'd add to the, you know, fundamentals of loyalty into that? Yeah. Just to reiterate, I'll say that the essence of loyalty programs doesn't change with AI. It just makes the decisions that you make easier from the brand side and also from the customer side. They will soon come the moment where you will be asking ChatGPT or any other agent, where was the best way to use your loyalty points, right? So in that context, I will say that from my perspective, that emotional component taking that transactional engine, granted as, as as a pillar of what you need to do, that emotional component in combination with what others showed is what really makes the new the new engine supplier valuable for both for brands and customers. Thank you for that. And we have we have a number of questions that have come in. And you know, I may I may pose this question to you. We have a question from the audience focusing on how we prioritise use cases within loyalty. And if you're prioritising use cases and loyalty, would you think of it as value versus ease of implementation when it comes to those core use cases of, you know, of loyalty? It's a matrix. So usually those are the two axes. So we have an X and y. We have value. And we have this of implementation. We do advise though exactly. Looking at it the way Diego and Cashmere looked at it. So we do recommend looking at what is the core, what is going to drive more trust from the customer. What is the most central to building an AI driven loyalty program? And then based on that, we can look at based on that value. What is the, let's say, the low hanging fruit and what we can do when we connect the whole data. We do see both types of bad practices. One is focusing only on completely superficial use cases. Without that, don't connect to all the data of the of the or almost any of the data of the company, which is not scalable, and the other one is trying to build the perfection from day one. So building a 2 or 3 year plan, within three years we will launch something perfect, right? So we do advise a middle way where we take a side, we take a loyalty program. We take either a brand or a market or something. We actually bring the data together and we run, let's say, a low hanging fruit program. So that's what we usually advise. Yeah, I'll let you jump on that because we made our own share of mistakes on these went big. And that I think that happens because AI is super overwhelming and is very tempting to do and very sophisticated things for us up there failing a few times, I will say that it goes back to the business question in hand. What is the one key thing that you are trying to solve? And then you need to look at AI on a way to solve that business question. That is paramount, because it's not about the technology itself, but about the challenge that you want to overcome. So I will say that it goes to a mix as under said between background implementation about what your is, your biggest business pain. And the other thing I had to that is sometimes you need to think about scalability and how how you can take a pilot and make it scalable in your company, because otherwise it's going to end up in something very glamorous, but not really with traction in your business. Yeah, I tend to agree with what you were saying there, Diego. I think it's not AI for AI sake. It is really about solving real challenges and real business questions. And therefore the use case is important. What is the right use case? And to your point, Diego, what is the problem you're trying to solve for? And I think it's not all or nothing, as in that AI needs to improve both the member experience as well as internal efficiency. Most of the time right now, you know, there is a polarity between those two elements. And what we've got to find is a middle ground that does both. And the other the other thing that I would say is that it's about incremental. It's not about trying to go, as Diego pointed out, all, all or nothing. But the one other element is that is very difficult. I think implementation is the hardest part of embedding AI and using AI. When you think about some of the stats, I think it was MIT that said 95% of AI use cases fail, only 5% have really been successful. And so you have to ask yourself, what is it that you need to do differently in order to actually have something successful? Start small and to Diego point, then find a way to be able to scale it if it turns out to be successful. But you've got to do it in a manner that it doesn't necessarily completely disrupt your operations, and more importantly, it doesn't disrupt your customer trust and it doesn't disrupt your economics. I think there's all it all AI is still glamorous, but we need to all to make sure that we understand the economics behind adopting it and using it. What does it do to the economics of a loyalty program, for example? Does it, you know, because you can get run away or you can run away with something. So I would imagine that there are lots of considerations. It's not quite an easy question or problem to to to solve, but it is one that definitely has to be done. The great point and it leads into another really good question we got. And it goes to the fact where loyalty, there's always a push pull when it comes to loyalty, loyalty in the finance departments of organisations when it comes to showing the value and just demonstrating that loyalty is worth it. And one of the great questions we got is around, you know, how can the ROI of the loyalty programs get past the CFOs? What are those critical KPIs needed to really showcase the positive results of loyalty programs? And I think that's an interesting question to think about from a one year, 3 or 5 year plan. So I'd pose that maybe cashmere at the start. How how have you thought about that, that question? I actually don't think you skip the CFO. You make the CFO your best friend and and you you know, you you make sure that you because once you've mastered the art of selling to the CFO, you can sell to anyone within the organisation. And and there are various KPIs. But as far as the CFO is concerned, they want to see the return on investment. So you've got to find a way of understanding what that means for your organisation in airlines, you know, and a lot of other loyalty programs. One element that is traditional is partner partner economics. How does that how does that stack up. But other than that it is about being able to. Diego showed some stats earlier, which was about what percentage of the people, what percentage of customers are now coming in through the loyalty program. There are so many different KPIs that loyalty can use to demonstrate different areas of the business, how they are contributing and how you're contributing. And I think I think I'm I think it was an issue that said it, but it is how do you make sure that a loyalty is embedded at the core? Loyalty is still too often something out there, but it is a key growth and strategic engine. How do you ensure that you align with all the various stakeholders, understand what they desire from loyalty, and find ways and means to be able to deliver on that and demonstrate that? But definitely don't don't turn away your CFO. Certainly. If I can add to that, Charles, I think for a very long time as part of these legacy structures, loyalty was very much kind of in between. People were associated it as part of the marketing department. Others were saying, well, it should be part of another team. And it was very much operating in, in, in isolation and was considered as a very cost generating team. And I think this is changing because a lot of examples and I always kind of refer to that cool story of Alta in in the US, where 95% of the revenues are coming through the loyalty program. And to the point of what cashmere said, it's what exactly you want to achieve as a business. Do you want to have a higher retention of of your customers? Do you want to acquire the acquisition aspect? Is that the economists that are not working, where and how you direct that, it's very much, you know, it might be all of the above, right. And it's where you kind of add all these different components to sell that to your leadership team in order to advance and create a really strategic lever out of this loyalty program. So for me, it's very much the core of what is important for your business and where you're heading and what you want to achieve. And let's remember from the AI standpoint, you actually do need to measure your program and your campaigns for the AI to learn. So this is not not this is not something that is optional anymore. So you need to build a sample for each campaign you're going to run. You need to have control groups. You need to be able to detect the attribution so that the AI engine itself needs to know and understand per user what is working. That's how the AI learned. So you actually will have a lengthy attribution or ROI reports. If you don't do this, you are just running a few AB tests, which is not going to give you the full ROI of the program. It was really interesting. And yeah, thanks for everyone's input on that. We have a bunch of questions that we've come through, and with the time we have, I want to I want to make sure we get to as many as we can. And, you know, one came in for Diego. How do you do 1 to 1 personalisation. Is it in-house or via third party providers like a Mastercard? If the latter. How long did that integration take? So I think the the proper way to do this is actually a combination of both. We also use internal capabilities together with the strategic partners such as Mastercard and other platforms. And and of course you can build this yourself, but it's going to take you years and experience and data and statistical knowledge that you may be able to acquire. But there are partners that already have that over and it is their job. So if you're concerned about time to market, I'll say that you should go with a partner in combination with with your own internal structure. What I've seen is that the biggest challenge is usually is not usually the model, but actually defining the business question that you want to solve and maybe organising the data you have and unifying customer data towards that. Right. So because I know from experience that partners like Dynamic Yield can deploy personalisation agents within weeks, but that's after you have defined your business problem to organise your data. Building an agent is not particularly hard. Once you have all these information set up and that's then you need to apply everything. And that's actually more difficult than the model from my perspective. But I'd like to relate to to to under and see what's his point on that. I think definitely the where you got the points nowadays, we always work with brands that are building their own models. So not only, let's say deep learning, TensorFlow time models, but even Genii based models. Now, when it comes to Mastercard, we are working with an orchestration of models at the at the at the end of the when the user is coming to a site or interacting with it and connecting existing and own models, we only work and we advise you on how to actually first put some order to the house as they go is mentioning. So we need to look at what do we want to improve. And we need to make sure that we actually have the data to create that impact. And that's when we can actually develop and run a posse that will deliver results and will allow us to scale. So I think keeping the fundamentals well, first is something that it's essential. And that's most brands, what most brands are doing right now I guess also those that we're answering on Nepal planning on implementing, I guess that's the stage at which many brands are right now, and we can of course, help with that. But as the goal was mentioning, it's important to launch once we have that not, let's say, waste time on deployments which barely use data. And another question that came in, and I think I'll pose this to pose this to the group, if you had to pick one, what is the most effective lever to improve loyalty, which in turn can lead to an increase in sales? If you had to, if you had to pick one. Cashmere, I can start with you. Just give me a minute to think about that I. I wouldn't pick one, I would pick two. The reason I say that is we have we often have this debate in the loyalty world about real loyalty versus deal loyalty transactional versus emotional. And I really believe that it's important for us to get the customer value proposition right. And that customer value proposition is not just a currency, but is not just an invisible loyalty mechanism or inversion mechanism. I think that those two elements, if they are finely tuned and when they come together, that's what will really create the stickiness. I don't think one without the other works. Currency has always a loyalty. Currency has always had a bad rap or a bad rap, whichever one you choose to use. But it's the glue. More often than not, if a customer can't see tangible value, they may not necessarily want to engage with you. And if you blend that tangible value with the intangible emotional elements, then I think that's what will get you there and you will be able to create stickiness. You will be able to drive acquisition, retention and growth. From my time, if I had to answer across verticals, the only something that we see across verticals is simplicity. So we do see more and more offers more and more mechanisms. We do see all types of campaigns that are being run when attention spans of users are dropping. So being able to keep it simple and show something to the user that makes sense in real time is what we see working across the board, from airlines to retail to QSR to to banking. I agree with you. And and I would add with that simplicity, it's more than simplicity. It's ease of use. I would think that, you know, that's really the crux of the issue. It has to be easy to use. For me, and from my interactions with other players, is very much about understanding really. Well, your customer here, you're targeting what they want, how they're engaging with your brand, and only then once you've collected that big data, how you're actually using it to tailor and move the needle if you want, and adjust your strategy towards that. I think a lot of companies are sitting down and saying, well, that has to be our strategy, but they kind of forget the customer and that is a disconnect that still is is out there. And I see this quite as a big challenge for many companies that they kind of put forward their financial kind of focus rather than the customer. And I think they need to work together. When you understand your customer, the financials become better. Thank you all. And I know we're almost up on time. See if we can get to one more. One more question to leave everyone with. I guess you know, a question that you know I'd like to leave everyone with. We've gotten a lot of questions on the biggest barriers to implementing, you know, AI into loyalty programs. And if you guys were to leave, kind of one thought on like, what are the biggest barriers that were, you know, you're seeing everyone wrestle with what would what would that be? And, you know, after that we can we can wrap up the show and any other questions we haven't answered, we will get to and address as a follow up. So not maybe a start with you on that one. Well, for me it's building trust. There is still very, very much lack of transparency. Well, if something goes wrong who handles that? How are the customers being protected? Hue steps in to kind of cover their losses. There is no regulatory framework that is kind of, you know, unilaterally supporting these customers irrespective where they're based. So I think I've looked up my presentation with a point about trust, and I really stick with that. I feel like creating stronger trust and that transparency of of what you're doing. You know, the Eagle pointed out they started there were lots of mistakes, but they were finding it. If you're transparent with your customers and they stick with you, then you're on to a winner. But trust for me is essential. If I had to think, if I were to say one, it would be a trial and error mindset. So as mentioned, we will see many programs that start with more of a two long term vision and take ages to implement. Or they are too small in which even if it works in one scale. So we are trial and error mindset. That always takes the core customer value into account. That already leads to everything, leads to be able to acquire the data and leads to experiments that allow companies to move very fast. And within six months, 12 months, we see that's how we see immense moves when it comes to the loyalty program. I think in terms of the barriers, I think when you're operating in a legacy environment, it's how you layer AI into what already exists. I think it's finding the way to embed it, and that's often the challenge. You can you can see it as a standalone technology. You can see it as something that's going to solve a use case. You can do a trial, you can run a POC, but then how do you operationalise it? I believe that's still one of the hardest barriers, one of the hardest things to to to achieve, and probably why so many people haven't succeeded. Yeah, I like to comment on that. Beyond everything that you have mentioned, one of the things that we have worked with, dynamic specifically, is sometimes you overlook the creative challenge that it means to do a recommendation model, let's say, for loyalty recommendations. But sometimes you overlook how are you going to communicate and make sure that the customer knows about that recommendation? Is that going through an email, WhatsApp media? Because that changes. And if that if you don't have potentially a system that can support variables within emails or different creatives, because sometimes most of these recommendations, you have to personalise the the image that goes along with that. So on, on on the operation itself, there's many challenges that you only realise when you're trying to do it. Like, okay, you have 1 million recommendations, how I am going to send 1 million emails different for each one of our customers with different creatives. And sometimes that I feel it gets overlooked. And on how to actually deploy that. Once you have the recommendation at a customer level, how you are going to communicate that to to a customer, so you can measure and everything that goes out there and maybe testing and all that, but you need to figure out a way that customers actually learns that you have that recommendation for him. And that can vary business by business. And with that, we will we're going to have to wrap up. I really appreciate all the great feedback and all the great conversation from our panellists here. A few just points. You guys will be receiving the slides and recording over email over the next few days. There's a feedback survey that will be sent out after this. Please fill that out. We'd love to know what what we can continue to provide. And you know, finally, you know, huge thank you to our panellists today. This was a great opportunity for all of you to share your expertise with us. And most importantly, thank you to everyone that joined us today. We'll get to every question that was answered as a follow up, and we look forward to seeing you again at future Euromonitor events. So thank you all very much and have a great rest of your day and week. We hope you've enjoyed this audio version of our recent webinar. As a reminder, you can find links to the full slide deck and video recording in the episode description.