Market Logic at DMWF Global 2024

Good morning. Thanks for joining the session. Unlocking consumer insights with generative AI. I mean, generative AI, that part is obvious, I guess. Everything is AI today. But for me, the key here really is the unlocking in the message, and it’s about how can we use technology to make sure that we use all the lovely data and insights we have dormant in all our information in the moments and for the use cases and on the occasions where we really needed to make a difference in the business. That’s ultimately what we wanna achieve to be more successful with the marketing outcomes. So I’ll speak about the problem a little bit. I’ll show you how a AI AI can solve it, how we solve it. But then I’ll also, share a couple of learnings we have from our customers where we have deployed that and also brief glimpse into what the future will bring around this topic. But starting off, MarketLogic, who are we? We’ve been in the market for quite a long time, for more than fifteen years now. And what we do is we provide a software platform to our customers that they use to connect all the data sources they have around consumer insights, markets, brands, categories, to bring it all together in one place and then be able to serve a consolidated view on everything a company knows around these topics. Now we’re not a CDP, and we’re not a machine learning platform. Think rather an aggregator for all the in house relevant information and all the paid and public information that is also flowing into the system. And the data can be unstructured data like PowerPoints or PDFs. It can be structured data like data tables, but can also be other tools can these these days even be other AIs we interrogate to share their information on demand to feed into this consolidated view. And so the ultimate mission of why we do that, obviously, is to help stakeholders, to help marketeers in the moment when they need to understand the situation, when they need to make a plan, when they need to make a decision to have the best possible understanding of their consumer and the world out there to be informed and make a winning decision, really. And this is where AI can really make a difference and help. And to bring it a little bit more to the point, what we see and what we hear when our customers is if you are, for example, in marketing, in this situation, you need to now develop maybe material, you need to develop creative, you need to develop a message, you’re working on innovation, Whatever it is, you will need also understanding about this human element, about your consumer. And unless it is a very strategic and high stakes decision you’re about to do, more often than not, people when faced with a decision of spending the effort and trying to find all that relevant information versus maybe taking a shortcut with all the business pressures they’re exposed to, many do the latter. And what we really wanna make sure and what AI and technology can now make sure and help with is to have a third button, if you will, which is press a single simple button and get the information in no time and be sure you have everything that the company knows and you don’t have to sift through endless data sources, platforms, tools, reports to try and piece together the picture of what’s going on out there. And Joe, therefore, what we have developed, we call it DeepSights, our product. What it does for you is to read all the material, to read all the reports, to read all the trackers, to connect to all the other tools Joe you don’t have to do that work. And instead then it can answer for you what these, what the actual insights to your questions are. And what you get is you get your trusted answers to the questions, to the things you need to understand, and you get them in a few seconds. And you get them grounded in your own insights and data of what your company knows. So you can be sure whatever you get as a response is deeply linked and traceable to everything you know from all of your systems and sources and also that it is really everything. You don’t have to second guess if there’s other information maybe sitting somewhere in the dark drive that nobody knows about. Everything is coming together. And then, of course, also, this is available twenty four seven, unlike maybe some colleagues you may wanna ask. Otherwise, this can be used day and night. It doesn’t give you an odd look if you ask silly questions. You can use it for any kind of exploration you need it. And very, very importantly, you can access it through managed channels. So, yes, of course, you can go through a web interface. I’ll show you a peek at that in a second. But also you can use other channels, established ways of working that you have, like you work in Teams or Slack or maybe you still prefer email. Or you even work on Copilot or Enterprise Jet GPT, and it can be integrated there to make sure, however you work, you can rely on the information about the insights from the consumer, from the market, from the brands. They are always coming from your trusted sources in your working process, And you don’t have to change your habits in terms of, again, remembering there’s a system, remembering there’s another source. Yes. I could go there. Yes. I could ask that, but reality is most people don’t. And lastly, also, you can integrate it anywhere else with APIs. And we have customers who now integrate this as a data source for other downstream systems, for other AIs, for example, who then start generating, for example, product concepts driven by inside consumer understanding that comes out of such a consolidated view. And let me give you a brief clip what that looks like. So you would go into the platform. You would ask your question. You have some sample question. Consumers react to automatic personalization and digital experiences. You shoot the question. It goes through all the sources you have and comes back with this consolidated answer, which gives you the exact quotation of all the underlying pieces of information down to the page level wherever possible. So it’s trivially easy for you, to read up more, to learn more, but also to go back and then, for example, understand and verify is that number correct, where does it come from, it came from here, so we can maybe trust this. You can see more. You can dive deeper. It branches out into the entire data universe, and you can even run a report as we say, meaning ask the AI now. Take ten minutes and look much deeper and broader and come back with an even more comprehensive result set, that gives me a broader narrative, that gives me all the sources with extensive bibliography, and that also allows me to export everything and go away with the report. And so within a couple minutes, you’re gonna have the the work done. That usually took a day or so by manually triangulating everything. Plus, you can be sure that everything’s in there that’s relevant, and you don’t have to worry about forgetting something. You can also, as I, said, integrate that into other systems, into other places. So if people don’t wanna come there, just ask the question in your Teams chat, for example, and get your answers there and break out from here to understand sources to create reports and drill further. Or you can also integrate it, for example, into enterprise check GPT if this is the tool of choice where you do a lot of work now where maybe company has already also other capabilities integrated. So you can, integrate, DeepSites, integrate this AI. And then whenever a user of GPT asks a question that is relevant to insights, it will automatically delegate, pull out that information, and bring it back into, the chat GPT environment where then, of course, you can do whatever you wish to do in terms of further working with it, in terms of now drafting an email, writing a report, or whatever the other workflows are. So very flexible way of, again, making sure wherever your users, wherever marketeers work, they can leverage all the information effortlessly without explicitly having to go somewhere. Having seen this, ChatGPT demo, of course, always begs the question, why can’t I do that just straight off out of the box with Copilot or ChatGPT or a similar solution, which is a fair question. And to a certain extent, of course, you can always load documents there and work with them if you know which documents you want. But what you’re missing out on is these systems don’t have this aggregation capability of all those disparate sources internally, externally, paid, public, and so on. And also, they are not optimized for these use cases when it comes to understanding deeply insights, consumers, etcetera. And to illustrate it, taking a piece of unstructured information, which is a typical specimen of what we have on our platform. Here’s a chart from some market report. And now looking at that, there’s a couple of questions one should ask themselves about how can an AI work with this information. The first one, of course, being, can it read the charts even or does it only look at the text? Joe, usually, a great proportion of the information residing in these reports is actually visually represented and not written out in full sentences. So that is a huge limitation for pure text based approach. Also, you need to understand whatever there is, does that really apply to the specific context of the problem at hand? Does it speak about the market, about the target group, about the situation, and contextual, parameters that I really need to make sure it does in order to avoid wrong advice or wrong information. How trustworthy or reliable is this force, for example, in the first place? Where does this come from? Is it is it a report? Is it, maybe a marketing white paper somebody uploaded and found somewhere? Does this uploaded and found somewhere? Does this all go out to third party tools to maybe all other lovely tool providers we see here who also provide relevant information that should be integrated into this consideration? Can it do structured data? Like, if you have trackers, if you have panels, if if you have any other kind of structured data information source, can it use that to pull out answers and information on demand as needed? And can it even go down to the individual, a single respondent, or transaction level when it comes to analyzing information like surveys, like transactional data you may have? And lastly, very importantly, can it even say I don’t know if there’s not enough information, which is a big weakness of many of those AI systems and so far as they usually are very much inclined to give an answer no matter what. And, of course, we’ve taken a lot of efforts to address those questions to to enable proper processing to ensure trustworthiness and robustness. But all that’s not only us. That’s also, of course, building on the capabilities of all the partners we’ve integrated with, which are content partners like Mintel, Kantar, etcetera, who provide content into the platform, but also technology partners who then, for example, like Experian help us do survey analytics and transaction level analysis in the broader scope of things. And this solution is something we’ve we’ve initially brought out more than a year ago, and now we have already many customers that we’ve rolled out to and that we’ve learned, from and have had experience with. And and a couple of thoughts in terms of what are best practices, what are learnings to be kept in mind here in this context. First of all, I think, initially, there was a big uncertainty that everybody thought, oh, is that now going to replace my Joe? Or what where will this go? And it kind of emerges that the best way to think about it is that the AI can help you to get seventy percent of the way very, very quickly, but the last thirty percent really, really require the expertise and the experience of the human. And you can rather use this as an enabler, as a multiplier to do more, but maybe to do the same things with much, much higher quality than it was possible before. So it doesn’t mean we’ll all be out of our jobs. I would say we’ll all be just more productive and produce much higher quality results with the AI. Also, of course, the question always is, how do I start with all these AI tools? One obvious, inclination is to say, let’s start where the risk is low. Let’s not go maybe customer facing. Let’s not take high stakes applications to begin with. Let’s take something where there’s a human in the loop anyway. Let’s take something internal, maybe an efficiency case. That all makes sense for risk control. But also very important is that, we see broadly a lot of Gen AI initiatives also fail, and many of them fail because the success criteria were not really clear and are hard to measure and are a little bit, based on vague assessment and impressions. So efficiency is also a great case to have a clear ROI driven metric where you can measure what do I expect this to deliver and what does it deliver even though we all believe effectiveness is going to be the ultimate game. Focusing on this to proving the point has has proven to be very, helpful. And then, of course, AI is exciting. Everybody is super amazed that the dog can talk now if we type something into the system, but, real live adoption and usage is still a different story. Everybody has now dozens of AI tools and is inundated with these things. So how do you come to the point that people really come back and make use of it to make their decisions and understanding better? And what has proven to be key so far is, on the one hand, find ways to integrate into where people are today, not expect them to come to the information, but somehow bring the information to them, to integrate into the tools, into the flows where they work, and to somehow surface content there. But also now emerging ways that you can even technically automatically begin to integrate so that there are certain workflows, like maybe concept generation, where then information from AIs can directly be fed in. So where possible, of course, given all constraints and considerations in terms of risk and review, it is also extremely beneficial to be able to integrate directly different technical components. And then very importantly, and I still don’t see a big consensus approach emerging there yet, is if you deploy these AI solutions, they are all super dazzling and great to look at to begin with. Very impressive. Also fun to work with, but it’s really difficult to judge at the surface level whether this is robust and reliable or it’s just a great story may made up by the AI, maybe not grounded in facts. And so it’s extremely important to to be conscious of what the real business need for accuracy is, what the real business use cases are, and then to have a systematic program to validate that objectively and not by just looking at three things and saying, looks good, sounds cool, we’ll do it. And lastly, also important to ensure that you have guardrails, in education in the company. So guardrails especially also meaning that you’re able to configure or to to instruct the AI to be working in line with what your best practices and what your ways of working and what your constraints are, for example, so that it becomes part of your team in a sense and it not is not a black box that is working in an uncontrolled way in a very central function. So these are a couple of the main themes we we hear. And I wanna conclude by a few thoughts on where we’re going, with this in the next step by looking metaphorically into the crystal ball. And there’s a few very interesting pieces and areas of activity that that we’re engaged in also, that I I find very exciting and that I will that I think will come to fruition in the not so distant future. One thing, of course, is what we call attitudinal personas. So using the AI now, again, to draw on all the information that is available from all these sources and be able to more and more impersonate a certain, target group, a certain profile, which you can then use to interrogate and get feedback from. Of course, there’s very specific solutions to that out there, but we believe an approach that is more integrating across the different sources will have have great benefits and be much more flexible. And that will especially allow us then to do this this other thing, the audience specific guidance and and critique of for content generation. So imagine you can not only use all the information from a virtual target group representative basically to help define what a new piece of creative or content should be, but also take any such piece back, feed it back to the AI and say, take it apart from the eyes of that target group and give me all the feedback, that I can then reflect on and and think about to improve and iterate on my content. So that can be like a like a virtual pretest, if you will. Of course, there’s a number of, activities also around using the AI to drive further learning from activities that have been done, be it activation, be it campaigns, be it innovation, be it launches. Typically, we attempt to do systematic reviews, but typically that also falls short in practical reality for a variety of reasons. Again, using AI to pull together data from what flows in any way across the board to make sense of it and to provide more learnings. And lastly, early warning opportunity detection, again, the mere fact, everything flows into you here being able to find new signals that are business relevant and proactively flag them or being able to find signals that vanish that used to be business relevant and, again, figure out and alert and enable quicker reaction times, by having this three sixty degree view. So these are areas that are currently working on, to drive further proactive value on top of the more reactive. Of course, we can answer any question and provide you with the data you need. And that’s what I wanted to share. And with that, I guess, we Joe open it for questions if there’s any. I had a question around what type of, clients do you find leaning mostly into your technology? And secondly, how long does it take to get to some value, post integration or once you’ve got one of the client? Yeah. The types of clients are obviously are those that do have a certain existing data estate that can be connected and onboarded. Typically, they would have either heavy investments in internal data that we can bring onto the platform or they may also, depending on industry, be more focused on, for example, syndicated content. I would say Rini needs to be a certain initial scope of content that that the customer needs to bring for it to make sense. But then in terms of rolling out, the beauty of the AI is that there is no deployment project. It is really a matter of, switching it on, connecting the sources, and then you can start asking questions after a few hours. Surely, you go through some data governance, some onboarding, make sure it’s in a safe environment, all of those sorts of things. So how quickly, have you found? Typically, that’s well, in reality, of course, what I said is, like, the technical, piece in reality to go through all the interim steps. It would be on the order of maybe two weeks or so if you have the data available. If we need to go hunting together, of course, it takes longer. Yeah. Yeah. Hi there. Hey. I just had a question around the translation element of this and whether there’s a tool that translates automatically within the technology. There is no explicit translation stage, but the AI per se is now at a level where it would support out of the box any major language. So you can ask the question, for example, in French and we’ll be able to use all the existing evidence to give you a French answer based on English, German, whatever contact. Okay. Okay. Great. Thank you. Yeah. Great. Then thank you very much. Be happy to see you again tomorrow in the in our automation panel. And in the spirit of don’t trust the vibes, please see us at the booth. And we also offer free trials. So if you wanna convince yourself that it makes sense, we’re we’re very happy to prove that. Thank you. Thanks, Olaf. Thank you.

Let’s get started in in having a conversation. But before we get into our topic, let’s get a a a brief overview, maybe thirty seconds to a minute of every person here on this panel. And let’s go ahead and start down at this end and work our way down. Sure. Morning, everybody. My name is Ayesha Khan. I’m the, digital transformation and strategic marketing director for Teva. For those of you that are not familiar with Teva, we’re a global pharmaceutical organization. And more locally, we’re proud to be one of the largest supplies medicines, to the NHS. This is a topic very close to my heart, actually, because I’m living and breathing marketing and AI as a challenge at the moment. Challenge and opportunity, I would say. We, are currently focusing on how we can leverage AI to really enhance our marketing, marketing effectiveness, looking at channel optimization, and also enhancing the customer journey as well as colleague and employee experience as well. Very nice. My name is Olaf Lenzmann. I’m a cofounder of Market Logic. We are a software provider. We, give our customers a platform where they can connect all their data about consumers, markets, regions, brands, their insights and intelligence, and then use AI to tie that together to help answer questions, marketing questions, tactical, strategic innovation, activation questions. And, of course, we also have a lot of customers right now who are taking the steps to to bring this to the next level, to embed this deeper in their, marketing practices and both in terms of using, and unlocking it for, the the the folks in the marketing, but also by integrating it even into other AI marketing systems. So that that is a whole interesting journey to see how how all these technologies grow together now. Hi. I’m Sabrina Godden. I’m global creative director at Vodafone, one of the largest telco suppliers globally. I’m heading up the in house creative studio. We’re working across twelve different markets, leading brand and marketing campaigns. All about bridging gaps between technology and creativity and ensuring that AI is not pushing us out of the way, but making sure that we’re using it in the best way possible to really elevate our creative and marketing output. Thanks, Ming. Hello, everyone. I’m Asavari Moon. I have spent over the last fifteen years, in tech and mainly in FMCG. I’m a computer science engineer who worked in AI decades ago. And then for Ellen Love in marketing, did my MBA, and I’m back in AI again. I’ve worked with the likes of L’Oreal, Estelota, Uber across different sectors of marketing, like brand management, digital marketing, etcetera. And now I’m in tech again. I’m also founder of Future Female Marketers. It’s a community for women to learn about new era of marketing with AI, AR, VR, Web three, Metaverse, and all of that. I know you today, you’ll hear about AI a lot and how it’s gonna transform the marketing, industry. But if you don’t know where to start or where to upscale, don’t worry. I’ve got you. I also run an academy called FutureEdge, Leadership Academy, where I’ve recently launched my AI and marketing course, which is basically an hands on experience with strategic thinking. So, I’ve got you covered. Good morning, everyone. My name is Qaiser Bachani. I work for Mondelez International, the makers of Cadbury, Oreo, Toblerone, all the great snacking brands. I look after the consumer experience function, for Mondelez, for Europe, including working with our global brand teams as well for some upstream planning. I’m also part of the AI center of excellence team, that’s been currently working globally for us to be ready future ready for the organization in that space. Really looking forward to this conversation, sharing what we have done, and also learning from all these experts in terms of, you know, how what’s the future holds for AI and, how can we leverage, this new tech or tool for for to build brands in the future. So looking forward. Beautiful. And I am Jimmy Newsome. I am your host and and, super super dad that’s not so cool. And, you know, I’m here to to to hang out here and and really find out what’s going on. Now the key here is I have a number of questions, of course, that I’m gonna ask. But the the real thing is when we open it up, you wanna make sure that we’re hitting all of your questions as well. And I’ve had the opportunity to to chat with a a few of you already in advance, so I’m excited to get your intake, but, of course, everyone’s as well. Because I think AI is a lot more than just most in most cases, generative AI, what people talk about. So I really want to hopefully, we can expand on that and go further. So I guess the first question becomes, you know, how, you know, what can be optimized with AI? What how are you leveraging AI in addition to, of course, generative AI? And this can be open to anyone to kinda jump in there and just start talking about the creative aspect of how it really is making a difference. Maybe I’ll start with that, and I’ll start with a right list answer. AI in marketing is not new. I know a lot of people have been talking about it, and they feel like this is something new and shiny. But we’ve been using AI in marketing since decades. You know? I started my career in what we call as broadcast era with newspapers and, you know, radio and TV and holding. I’m old. But then, yeah, we came to social media where we had more of a precision area where we had one to one communications. We could target a message that we want to send someone. And then we have what we call as a predictive era where we have, you you know, the likes of Spotify or Netflix or even Amazon, which predicts what could be your next purchase, and that is all AI. We’ve been using AI for media, media buying as well. So it’s not something new. But what’s changed is the rise of LLM and the rise of generative AI, which has made content creation simpler, which had made understanding of text and output, into, like, test and images easier. And that’s what’s gonna change marketing. And I also feel like the growth is gonna be has been linear for a certain extent. It’s now gonna be exponential. And lots of things with the help of Web three and Blockchain is, again, gonna help us reach our consumers in different areas, in different, platforms as well. If we look at marketing per se and where AI can play bigger role, I see three use cases. One of them would by content creation, of course. I read a stat somewhere by Citista which said by twenty thirty, which is not too far from now, ninety percent of the content will be created by AI. So that’s a big opportunity for us. The second one that I say is gonna be around chatbots. Chatbots will be integrated in most of our customer experiences, also customer support. That’s another area of marketing where we say AI playing a big role. And third one that I feel would be predictive analysis. So sort of learning about our customers a little bit better, designing our workflows to speak to different persons, different personalities, different consumer segments based on very targeted approach. So these are the three areas that I feel AI will have the biggest impact. Yeah. If I may chime in there. Thanks. And and I do agree AI, of course, is a broad broad field and has been around for a while. But when it comes to LLMs and Gen AI, I also see that, there’s a more mundane aspect to it that still can be a very big unlock for the company, and that is the LLM enables a completely new way of interacting with the information internally inside, the organization. So it it allows people to make use of information that is already there, maybe not even AI generated, but apply it situationally, contextually in a very easy and natural manner. So it makes it much easier to bring it to life and to really turn it into a business outcome. Just to to your question on optimization, if you look at from a purely from a business perspective, I think one of the pain points usually all our businesses have is lack of speed and scale. And I think that could be the anchor in terms of the use cases for AI to begin with. And, of course, the rest depends on your imagination, how you want to extend that. And I think the way we are approaching it at Mondelez as well is how we are able to reduce content at scale and then how we are able to distribute that content as well pretty quickly. Because there will always be expectation from the business to do more with less, and at quick speed and be reactive. And I think AI provides us a, I would say, a solution to look into that, and that’s how we are approaching it to assess what the current pinpoints that business have and how can AI can be a tool to help help that. And for us, speed and scale is a critical element of that. I think just to touch on the point that you made, there was, a survey done by the Digital Marketing Institute, which stated about sixty eight percent of marketeers still see automation being the biggest benefit for AI. So actually, there’s a huge opportunity with processes and process improvement that I think, you know, there’s a creative aspect of, AI. There’s the content aspect, but actually, there’s a huge opportunity around improving processes and actually removing some of those mundane tasks that we as marketeers actually have to do day to day. So I think there’s a the you know, we have a complex role as marketeers. You know, we we get tasked with everything, and, and we need to know about everything as well. So I think that’s the biggest, opportunity, I think, that we see is, around the automation aspect. I think I can follow-up on that. Someone outside just asked me a question ago. So you I’ve had a talk earlier and we said we have, increased, our speed or our workforce within seventy percent. So already, you know, through automation and integrating AI. And And then someone asked me, so what are you guys then doing if you if they don’t spend time on that? And so we’re actually creative, and we’re focusing on the problems and the pain points and the strategy. But we really embrace, you know, putting in the automation. We’re using it for content scaling, for content generation. That allows us to be more strategic, more critical, and for me, again, more creative. Yeah. And, I I like what you said a lot about, you know, I what excites me about AI is the fact it’s not even that it that it can take existing content and then start to create something new internally. So it’s not really necessarily pulling from other pieces of content to create something. It’s able to take existing things that you have and then create something new and bring that to the forefront. And that in itself, I could sit in a think tank for hours and hours and come up with nothing. And this here just gets me a really great ground floor running type of scenario. You know? So when when we think about AI and the fact that it can help you, you know, it can save so much time. And so you anybody can chime in here. I’m definitely leveraging AI across so much. And what I’ve noticed is I’ll use let’s just use generative AI and crank out a lot better articles. You know? So instead of five hundred word articles, maybe fifteen, two thousand, twenty five hundred word articles that have a a lot more depth to it. But I actually end up still spending just as much time creating because now I have a better piece of content to work with. So I’m so I’m still using the same amount of time. I’m just getting a better outcome from what I created. Is anybody seeing something like that? Anybody want to start there? Yeah. Yeah. I would I would truly agree. I would we tend to think about it like AI helps you to do seventy percent of the job in literally close to no time, But you still then need to finish the job, but you have more time to finish it. And you can do better quality, you can go deeper, you can, just do more quality work with the same resources. And that’s how to think about it, not so much effectiveness, is more important than than rationalization and and, cost squeezing in this regard. Can I add to that? I think it’s also a mindset shift. What I have started implementing just so that I have more AI in my workflows is that taking a pause before starting any activity and sort of thinking, is this something that I can do manually? Is that something that I can completely outsource to AI or some tools? Or is this something that I can do as a combination of both manual intervention and AI? And having that mindset has really helped me. Like, even if I’m writing a brief, how about asking those very initial questions to the or even perplexity to do some research and then use that as a base and enhance it rather than staring at a blank document for hours. Similarly, if you read, like, a really easy mood board to start, you know, your thinking ahead, these tools can be really helpful. Again, having that balance of how much do you let it to be automated versus how much you want to start from scratch, really helps. So, you know, this is a general rule of thumb. Even if you’re writing an email or if you’ve got a really long email thread, use AI to summarize it and see if that helps your work, and then slowly try and, add it to more workflows of your life. Yeah. I think everything has been said. The AI, regardless of how you’re using it, is eventually going to elevate the outcome of the product that you are looking for. Because of the time, because of the effort, it will give you a good starting point, but then you have to kind of because, you know, in a lot of cases, you will still know your brand or your problem or your pinpoint better than the AI. Until and unless you have a peaceful LLM for your brand, and then you can fix those things. But till then, I think you you just have to use this more effectively to elevate your product or outcome that you are looking for. And that’s exactly, so far, we have been leveraging AI internally as well just to make sure that we fine tune the eventual outcome that we are looking for. I think if I can just touch on skills being one of the the biggest things that we’ve had to look at is actually Joe we are really focusing on upskilling our existing teams, bringing in new talent to help us understand how we can leverage the tools that we have an opportunity to use. We personally also work in a very highly regulated environment, so that comes with a huge amount of challenge. And also being part of a large corporate, large global organization as an adds another layer of complexity. So we’re having to navigate all of this, but at the same time, thinking about the skills that we actually need because to be able to even use the tools that are available today. Like, if I just go online and, you know, there’s a number of, vendors here that provide fantastic tools. I need to know what I’m going to get out of it. So I think that is so critical. So that’s one of the challenges when I speak to marketing leaders that everyone is saying, I’d actually don’t know what skills I need to bring in or think about for the next two to three years. Because after three years, we’re gonna need a whole load of new skills. So that’s one of the biggest challenges we have at the moment is making sure we have the right skills right now even if it’s for a twenty four month runway to to get us into the, get into an embedded AI into our functions and in our teams. Mhmm. And you and I, we had a great conversation, a pre but I’m gonna jump back up to you. But go ahead, Olaf. Alright. Thanks. So but just chiming in on that, I think there is for sure the need for the upskilling for understanding how to use the tools. And you said before you start to set out on a job, you would think about how you can use the tools. What we see as a software vendor, however, is many people don’t always think ahead in that regard. And then even though there are these tools available, you kind of don’t use them as much as you could use them if you would consciously think about it. So there is also a big opportunity to do this upskilling, to do education and awareness, but then itself is also another opportunity for AI to be part of your workflow and advise you when to do what and what the limits are. So today, you have Microsoft Copilot, but in fact, that’s not a Copilot. It’s a chatbot. If you had a true Copilot that would understand what your job to be done is and that could help you navigate through that, then you could already do so much more also with the tools that exist today. Absolutely. And and I I will admit I was a bit confused about Copilot because I because I’m like, alright. What? What is where? So, you know, and that sometimes where they might even be dropping the ball a bit and getting ahead of themselves. So let’s talk about adoption here. And I wanna go with Ayesha here because we talked about, you know, you know, especially with the bigger brands, legacy brands, sometimes we can move they can move a lot slower. And you said something to me that was really shocking that the health care industry and and what you’re doing actually jumped on this a lot faster, which I thought was extremely exciting to see a a a brand or or an industry decide to come out a lot faster than normal because they can sometimes move at the speed of, a turtle. I think one of the challenges that we and probably other brands have is that you have marketeers, and other, I guess, talented people that are jumping on wanting to use AI straight away. So the challenge that we had was to make sure that we had the framework and the guidance in place right at the outset to help our marketeers globally understand what can and can’t be used, what you should and shouldn’t be putting into an NLM. We’ve looked at custom LLMs as well to help our marketeers actually use the tool, without having to feel restricted. So I think there is a, the the governance piece is absolutely critical, and I think, you know, the government early our UK government earlier this year issued some guidance around AI for regulators. This is a new space even for regulators, whether it’s FCA, PAGB, ASA. Everyone’s still navigating what it means in terms of, content, original images, or copyright, all of those things. I am sure your your brand is experiencing some of those as well. So I think the the challenge around having the frameworks and the principles in place, leveraging the guidance that’s out there, looking at responsible AI. So we talk a lot about about responsible AI, especially for the next one to two years. Once it’s embedded, hopefully, it just becomes AI because it’s good practice. But actually giving guidance to our marketeers on what they can and can’t do, should and shouldn’t be doing, I think has really helped. But we came at it from a risk angle, which is probably why it it kind of got the traction that that it did. Yeah. I I agree with, I, with Ayesha. I think, there’s a major pivot happening, across organizations. Whether you are a big brand or a small brand, everybody wants to leverage AI in some shape or form, and we went through the same process. All our big brands, whether it’s Cadbury or Milka Oreo have leveraged AI for different sort of set of campaigns. And for us, the challenge has been that how to make sure that we don’t end up having into a legal issue or a or a IP issue of some sorts. And therefore, we do like, currently, the process is such as that there is a manual intervention that happens to make sure that if it if there’s an idea if there’s a campaign idea that’s come from the agency, we go through the process to check if everything is sorted. Eighty percent of the campaigns do go through, but twenty percent of the campaigns don’t because there are certain elements which we can’t approve. We we just don’t know how it’s gonna land. And I think you spoke about capability. I think there is, like, fifteen years ago when digital was become was becoming mainstream, there was a lot of focus from the organization to upscale their marketers to make sure they understand. But I think this time, I’m seeing there’s much more pull from the marketers as well to be part of this game, and I think they are much more keen and interested to learn this space. And, you know, internally at our end, we always try to preach, like, you know, learn by doing. Right? So do the stuff so you know how it works and what doesn’t work. So I think it’s a it’s a great shift or the great pivot that’s happening currently, and I I would say it’s a it’s a great exciting time to be a marketer actually, you know, for sure. Joe, yeah, I fully agree. You know, responsible AI, legal frameworks are extremely important. At the same time, it’s great to see the marketers are leaning in rather than just sitting at the, you know, back end. That that’s exciting because, you know, when you when you say it like that, you think about change management where you always kinda have to force people to try something new. And and and and I’m excited to definitely do something because, again, I’m looking at productivity. I’m looking for a better opportunity to get things done. So this is exciting. So anybody I have another question, but I wanna make sure okay. We’re good. So let’s dive into the tools. Hey. It’s my homie right there. It’s me and her. We bonded yesterday over at at the happy hour. Glad to see you here. Okay. Here we go. So let’s talk about let’s dive into actual tools. If anybody can share some of the tools that they that they really like when it comes to getting things done. Who wants to go? Yeah. Don’t laugh at me. That’s right. I got I was worried about that question because, you know, when I came to this panel and we read about the question, what are the current tools? I mean, that looked totally different, you know, two months ago or a month ago because everything’s changing all the time. I mean, again, we’re we’re looking at the creative angle. I mean, look from the, motion side of you. I mean, everyone’s waiting for open AI as a Soarer and seeing what it does within video. I mean, mind blowing. But then suddenly, you’ve got Illumina AI coming out, obviously, with Dream Engine, which we so I’m not quite sure if they just publish it because they wanted to be there before OpenAI finally comes out because there are a few interesting bits coming up. But, nevertheless, mind blowing. Or did he look runway? You know? Now I think feel like every time when you look at or read anything, there’s a battle between runway and dream engine going on. You know? Who’s doing well first? Who can do a few more seconds? Who is is more intelligent behind it. There are so many tools, but also it comes to know looking into you think you said it earlier is there are so many providers out there. And interestingly and when you go through here, there’s so many brands and everything. Suddenly, everybody just throws AI in their tagline. Just like, okay. You’ve been here, but everyone’s doing AI. It’s, you know, really be critical about it. I think tools changing all the time. But the the way we’re doing is we’ve got got AI task task force in house. Joe we really are embracing the new tools. We are embracing different tools, but we’re also really, really critical about it. And but it goes back to we’re at the beginning of it. They they’re changing weekly, you know, monthly. And it’s how can we use it? What’s the best? Some of them, we just can’t simply use, but that doesn’t mean we’re just closing our eyes or saying we don’t want to. It’s just we’re learning and growing with them, and we’re just keeping our eyes open on a weekly basis, on a daily basis. You know what? Someone else throws something else out there. Speaking from a provider’s perspective, also, the the upstream technology is, of course, changing at a dramatic, pace, and it is very difficult to even keep up with what is going on. So I think from our side, but also from from our customer side, it’s more like really focus on the the concrete use cases you wanna address and focus on that and shut out a bit the noise of everything that’s happening there. But at the same time, it it it also means constant evaluation and constant challenge of what what the best tools really are. So, it is a very exciting time to be, in this in this space, but I do very strongly believe even though it’s very engaging and personally interesting, and that’s, I believe, why also many folks in the in the organization are keen to get their hands into it, it’s very important to be always structured and reflective of what is my use case and what do I really wanna achieve and validate that and not be only dazzled by maybe the, fantastic charts, words, or images that come out. I think just to echo what you’ve just said there, I think there’s so many tools out there and platforms. It’s really hard to actually navigate through what what is going to be useful for your organization. We’ve taken a use case approach very much, which is based on business value, and actually effectiveness. So we are starting small. A lot of the times, the pilots start in individual markets on a very small scale, either through a vendor or through our internal teams. And if there’s validation then for that use case, then we will scale it up. That seems to be Joe far working as the most sensible approach, but there is so much out there. You know, we have marketeers that are wanting to try things with their agencies. We also have, content creators that want to try, you know, content, type AI tools. So I think there’s a lot that we are trying to navigate, but we are trying to focus more on use case value and, and really looking at it with a sandbox approach. So try it. If it doesn’t work, doesn’t matter, we pause, try something again. And that flexibility and having that ability to be able to start and stop, I think, is really, really key. I agree to everything that Sabrina said. I feel like since the the rise of Chat GP3, I’ve been tracking all the AI tools, and there was a time when I counted more than twenty thousand tools that were available there. And being the known myself, I actually tried and tested two hundred different tools, both free and paid version. But exactly as you said, I feel like the space is changing so much. I do have a feeling that they will also be, like, a consolidation of some of these tools. But having, use case based approach really helps. Like, there are so many tools, but what is your main need, and what are some of the tools available there that you can try? I probably would share some of the ones that I use on a daily basis. You know, your old friend, Chat GPT, is a really great one to start, and based on certain ideas. You have Perplexity as well, which also helps you get links, to some of the queries that you’ve asked, which you can refer to later. If you’re looking to write copy, Anyword dot a I is a great start or even Copy dot a I is, again, a great start. If you’re looking to sort of, like, automate your emails, Mailchimp’s AI is really great. And if you’re looking for just some, you know, text based help, Jasper dot a I is, again, really great tools. But, again, exactly as you said and Kasif said that it all boils down to, how much input you’re able to add in, from a brand perspective. Does it align with your brand voice? Does it align with your tone of voice? Does it align with your ethics and values? I think that’s really important. And, again, similar to what Aisha said, it’s that small pilot, test and run sort of an approach to it. So test it, see if it works, positive need by, take learnings, and then scale. And I and I think just to add to that, our approach has been very much around the kinda can you build it yourself? Can you buy it in? Can you adapt it? So I think that approach has really helped us to work with our, internal governance structure and really look at actually what we need to bring in. Do we need to bring in something, or can we adapt to what we have? Can we work with existing enterprise vendors because they have something coming up rather than having to build something ourselves? So I think this this process of almost elimination is helping us to focus and really focus on on the tools that we will need for the future because things take time to develop if we’re going to invest in in developing, in tools in house. So we we’ve really taken that build by approach, borrow approach, for our internal focus. Yeah. The same. Not only take us it takes time to build those tools. It requires a lot of investment from the from the enterprise as well. I think in the next and the the the speed at which we are going right now, I think in the next probably eight to ten months, you know, we will continue to work with independent partners. But I think at the same time, we have worked with OpenAI. We have worked with Google Gemini as well a lot. At the same time, for certain campaigns, we have used some new startups as well. So depending on the use case, depending on what the need is, and I think for the next eight to ten months, probably the the strategy will remain the same. And then let’s see where the space heads. I think, what we are seeing with OpenAI and what we are seeing from Google Gemini as well are two big players, see whether whether they can become one stop solution for everything or something else altogether. So I think it’s an exciting space. As I said, you know, do stuff to learn, and then let’s see where the, you know, space finally Joe, and then we’ll we’ll take it from there, Joe be really honest. K. So we got getting close to I mean, I have so many questions. So maybe we can have a little powwow over there after this so we can continue this conversation. You guys are more than welcome to join us over there. But, anyway, I have one more question. I’m gonna kinda open it up here, and I’m gonna start with you all off here because you you you have a note about, criteria. And I wanna talk about, you know so what are the criterias for, you know, choosing these type of tools? Yeah. Well, again, of course, you start out as we just discussed with the use case with what you’re trying to accomplish and, of course, need the success criteria for that. But then there’s a number of, I would say, overlay topics that you need to take into account. It starts, of course, with, especially in the Gen AI field, being able to ensure transparency. Where does the data come from? Whatever it comes up with as a result, can I trace back the origins of that information and validate that it’s correct? And I should, even if I can, better do that exercise once to, convince myself that this works. And then, of course, there’s other aspects like, can I I think you said, can I, take this tool and instead of just using it as a black box, can I overlay it with my context, with my business specifics without now going down a rabbit hole of becoming an AI, expert necessarily? And then, of course, at least all the larger corporations we work with have a huge range of other criteria in terms of compliance, ethics, etcetera, that that also need to be taken into account. Okay. So let’s go ahead and open it up for questions, and I hope we have quite a few. I believe that hands are up already. I like this. Name and company and your question. Thank you. Hi. My name’s Nina, and I’m from Digital Gold HQ, a marketing agency. Thank you so much for sharing your experiences and your knowledge so far. I wanted to find out whether any of you would take a stab at outlining a broad road map that the average middle sized organizations I’m not talking about huge global bands with massive budgets, But the average medium sized organization, what would be their AI marketing roadmap and the core areas within the marketing function that they should look to start integrating some degree of AI functionality or resource. Okay. Oh, wow. What question? There you go. We just went for it right for the gold on that one. Okay. I’ll have a stab. I think that my recommendation for anybody looking at a road map, and I’m sure there’s there’ll be a few people out here today that will have have the brief from their CEO or their MD to say, go and sort it out. I think the the first thing I would say is really understand what your business needs are. You need to understand you need to have the outside in approach. So what is it that will enhance your customer experience? What will enable your then your marketeers to be able to respond to that and enhance that customer experience at speed? So you look at what your customer needs are, then you look at your marketing skills. I would then look at assessing off the back of that then what do you have today, what technology do you have in house, what can you bring in. Because I think with a medium sized company, there’s still flexibility to bring in and leverage external existing tools already. So really looking at either getting in some expertise to help you understand what technologies are available. But for me, the biggest thing which no one else can learn is really what will enhance that customer experience and what your current marketing skills are because they’re the questions that only you can answer as a business. Once you have those answers, then you can leverage, the expertise around you who can hopefully help you pull together a road map. Thank you so much for that question. I think it’s a brilliant question, and this is why I feel AI has so much to offer because it’s almost creating a level playing field. You know, we’ve worked for big corporates. We know we have big budgets and big claims, but with AI’s power and tools, now it’s made it easier for medium size or even start ups to sort of, like, play in the same field as a big corporate. I have, a three phase golden rule, for any of the AI adoption. And the first rule, this is the most important one, is start with the consumer. See how this AI tool is gonna be helping your customer’s journey. You know, a lot of CMOs and marketing leaders look at AI as a way to, make something more efficient, as a way of cost reduction, faster to market. But instead of that approach, start with how does this AI tool is gonna help us better our customer experience either by sending them a targeted offer or either by, you know, enhancing the customer journey. The second one would be having, like, a short term as well as a long term approach. So sort of looking at what are the ecosystem, what are the priorities right now, and where can that be scaled using AI. And then as a long term, think about how you can work with, you know, some of the developers around these AI tools to make it customized for your brand, whether that’s custom GPT or sort of like an integrated way into your existing workflows. And the last one would be, sort of like creating that atmosphere of people to be able to test and learn and not for your AI, but also feel that they’re they’re ready to upscale, also comes down from management in a lot of ways. So if you encourage that in your time, that’s when people are also open to try it in their daily work, in their smaller work streams, and then scale that. Hi there. My name is Louie. I work for We Grow Startups. We’re a digital marketing consultancy agency. I guess my question, and it might overlap with the last question a little bit as well, is it’ll be interesting to know from your experience what steps you took to sort of champion this implementation of, like, where this shift in ways of working to be part of just your consistent flow in your internal teams as well, and if there were any kind of obstacles in the way and how you kind of went around, those obstacles as well. So if I if I understood your question correctly, I think first, it’s extremely important to have, senior level sponsorship around AI also in every organization. Otherwise, things will not move. Because as we’ve been talking about, not only you need, resources, but you look require a lot of investment as well depending on the size of your company. I I think in in our experience, the the obstacles or the challenges have been mostly around responsible AI and also making sure that we are in sync with the compliance of the legal framework that we have developed. Because that’s the most critical thing for the business. We don’t want to do something which tomorrow end up creating issues for us. And apart from that, there’s been much more, as I said, leaned in marketers who really want to do this work. They want to experiment the tools as well. But at the same time, we know there are there are guardrails that we need to follow. And apart from that, I don’t see any other challenge apart from in the long term, you know, if you really want to bring some of this thing in house, if you want to build your own custom l LLM, etcetera, that will come at a cost. And therefore, you need a senior level sponsorship before you venture into anything like this. Yeah. I would agree. And I think also it’s important related to that question to make sure there is alignment on a broader level what the initiatives are and how they feed into one another. Because especially in these early days, there is a lot of activity, and we often then project stall, halt, and and things need to be sorted out. So to to have a clear vision on that from the get go is, super critical. Alright. So please put your hands together for this esteemed panel, ladies and gentlemen. We really appreciate every single one of you and your perspectives.