Wella Company and SAP Emarsys

The future of marketing:
Unlocking the power of AI

In our next session, we're going to be taking a deep dive into the power and potential of AI. Please welcome our brand leaders, Susan Jones from Diageo and Aaron Bradley from Wella Company. Susans and Aaron, a huge welcome to you both, thank you for joining us. And Susan, I'd love to start with you. Tell us a little bit about how Diageo thinks about AI and some of the efficiencies that it's bringing to the business. Hi, everybody. We're thinking about AI in a couple of different ways. I'd say the first is productivity and how it can bring efficiency into the business. But increasingly, I think we're thinking about how it can inspire, how it can inspire innovation, how can help us drive insights at scale much faster and much more depth. And then we're really starting to lean into how we can use it to improve consumer experience. And that to me is sort of the next, you know, dimension for us to go into. And in terms of workplace productivity, how are you using AI within your teams and what are some of those results that you're seeing? The productivity within workplace, we have our version of ChatGPT that's within Diageo and more and more people are coming to rely on it and see how it can make our jobs easier every day. But a lot of the productivity we're doing is within our marketing practice, really just using pretty simple tools to do programmatic bidding at scale much faster, much more competitively developing tools that help us do the things that have been really difficult for us to scale. So as an example, we recently co-developed a tool that helps us with compliance to our marketing code. As you can imagine, in the spirits industry, our responsibility to our consumers and to marketing is a big task and it's complex. So training every marketer on the code and having confidence that they can interpret every part of it is is pretty difficult. But what we found is we can use AI. So we've developed this tool where marketers can upload the content and we've trained the AI based on the expert that we have in compliance and found that not only does it take a bit of a heavy lift to give people a little hint to say, "Hey, you may want to look at this, you may want to look at this aspect" but it's a great training tool. So when we first rolled it out and tested it, people agreed with the tools assessment about 50% of the time. So therefore they agreed with the expert assessment about 50% of the time. After using it for a couple of months, they agreed with the assessment 97% of the time. Wow. So we find time and again that's kind of the extra benefit that the AI's bringing, not just consistency of how we're performing, but as a bit of a learning and a training tool. And Aaron, how are you using AI as a sort of learning and training tool? Have you seen its impact on workplace productivity? Yeah, of course. Firstly, Hi, everyone. So, at Wella Company, we've run a number of different pilots and we've implemented workplace support tools and the areas that we're seeing a huge amount of efficiencies are the uses of AI, which enhances how you as an individual within the workplace operate. It's not a separate tool as such as we're using. It's how do you enhance or accelerate how someone's doing their work. So the areas that we're seeing already very quickly are the meeting arrangements, the transcripts, the automation in terms of sentiment analysis on some information that comes through, the summarization of a lot of those those those long emails everyone gets and people think, "I don't want to spend 20 minutes reading those", uou can you can summarize those very quickly. But the clever part of the AI tools is the next step, which is it makes recommendations on actual actions for those employees. So it will read, it will read through maybe some meeting notes it will read through the emails, and it will then make suggestions on setting up a meeting for this person. So for me, we're not using it so much in the training aspect, but we're really pushing it out for those every day sort of efficiencies which which we're seeing, we're seeing huge amounts of benefits of already and we're only scratching the, scratching the surface. So it's removing some of those more admin tasks that somebody would have to do. Yeah, I think one of, when we looked at and I'm sure we'll talk about this throughout the conversation. But when we looked at deploying AI, it really comes down to the the problem that we're trying to solve and the value that we're trying to add either to the individual, to the company or the customer and consumer. And for the workplace efficiencies, it's really about how do we open up more time so that they can do more creative things. And so one of the big areas that we found were the administration of everyone's day to day lives. But the other thing is the augmentation reporting of of data. And again, that's in multiple different ways but even just as an individual trying to gather information within the organization around basic pieces of content or data. So instead of having to go into the data team and asking for these things or searching for your folder system of 400 PowerPoint presentations, which you've been storing over the last few years, we're using AI to quickly find, augment and then provide that information to you. And I think that's what we're seeing a huge amount there. Yeah. And Susan, I know that you're also using AI within Diageo to really boost innovation and creativity and one of those areas is in packaging design. Can you tell us a little bit about that? Yeah, one of the really amazing programs that our breakthrough innovation team piloted was using AI in collaboration with an artist for Johnnie Walker. So if you're a consumer and you go to our Johnnie Walker experience in Edinburgh, which I highly recommend to everybody and buy and during the program, if you bought a bottle of Johnnie Walker Blue Label, you could co-create the label with Scott Naismith, an artist. It guided consumers through choices to make and then crafted a bespoke label for the consumer. So, you know, just an example of, I think, AI bringing creativity to the consumer experience. And, you know, consumers tell us every day they want more personalized, more memorable experiences. And I can do that in a way that would be hard to replicate sort of conventionally. Yeah. Can you give us some examples of some of those other customer-facing initiatives that you're using to really create that sort of personalized experience through AI? Yes. One of the other experiences that we have started with a program called "What's Your Whiskey?". So, you know, consumers told us it's a little difficult to navigate the the many, many whiskeys that are out there. So we have our own system that guides consumers through their choices and asks them a series of questions about sort of common taste profiles of "Do you like lemon, do you like vanilla, do you like this? Do you like that?" and uses AI to then give recommendations. We've taken that system and now used it to other effect. So the more recent example is "What's your cocktail?" which helps people find the cocktail that would be perfect for their occasion, for their taste profile, etc.. And what we're finding is obviously it gives amazing recommendations. But also 45 million people have gone through it. And the wealth of data and information that that's bringing to us the insights on taste profiles, on food pairings, on everything is really valuable to us. Actually, I tried as part of my research, I tried the "What's your cocktail" website. What did you get? I got a Captain Morgan Mai Tai. Yeah, it's really fun. Everyone should have a go. And you mentioned about this wealth of data that you've got from from that sort of campaign. And have you seen any uplift in sales as a result of this? Yeah. I mean, what we see is every time we use consumer insights and append what we're doing with data to personalize it, we see better results. We see cost efficiencies. We see it driving more sales. It just works. Yeah. Great. And Aaron, how are you using or planning to use connected data and AI and to improve the customer experience? So Wella Company, the customer's always at the forefront of, what we do. So when we're looking at AI to increase and improve that experience, it touches on what Susan mentioned earlier around this personalization at scale. And what what we're trying to do is take that step further in terms of hyper - this is sort of hyper personalization. And that means treating individuals as individuals. Understanding them and using a huge amount of information to be able to provide them with the right services, the right products, the right tailored experience, which wasn't really possible before such a scale to so many people. So when we're looking at improving the experience for our customers, we have such a range of different customers from independent hair stylists up to big salons. And of course, we've got the consumer business as well. But one of the big things for Wella is we grow if you grow. So it's really important for us to connect to that customer and help them grow as a as a business in partnership with us. So we're using the information in the data. Can we provide more services for them as individuals, so the individual hair stylists or the salons? But then how can we use the data to provide insights which are going to be tangible, real value adds to those individuals? Of course, we've got the product recommendations andstuff like that, which we actually use with Emarsys as well. But it's that mass information of how how would you added an additional service? How do you improve that experience? Again, recognizing the individual and providing them the tailored experience. Yeah. And can you share any of the results that you've seen as a result of providing that hyper personalization? Yeah. So we've got I mean, we're in various stages, as, which part of that, but we've seen a significant uplift in, in using the, the AI product recommendations and things like that because it's actually, the clever element of it is showing and recommending products or services which aren't natural for us internally: to say X goes with Y. Actually using A to be able to identify opportunities for those individuals when there's product recommendations. Whether it's education courses, whether it's, you know, other services around that. So we do see a conversion uplift. Absolutely. I'd love to know from both of you what are some of the biggest challenges you face with integrating AI into your operations? I think what we've learned is you have to understand what you're trying to do with it at a fairly granular level, almost a hands on level first. When you put the technology first it and don't understand what you're trying to solve, it can lead you to some strange places. And I think, look, this is an area that it's changing so rapidly. We've had to rethink and build systems that are a bit more flexible so that we're open to when the next generation of technology comes in, we can connect it through and and not really be stuck to a technology that might be obsolete in a bit. Yeah. So it's also understanding the why you're using it, not just using it for the sake of AI. Yes, exactly. Aaron? So for us, it's an interesting one because the challenges of integration. And I'll make this point because actually positive and negative is that everyone is so keen to use AI, which is amazing, right? Because we've already bought, you know, they're already bought in the the challenge on there is to make sure that we're executing the right AI tools for the right reasons, and we focus on the ROI, on what those solutions are. And there's a real value to us as an organization and to our customers or consumers. There's no shortage of new tools which keep getting asked to be put in. But the I'd say operationally, the first challenge that we had to face was to actually have a data science strategy, because without that foundational data strategy, there everything above that, when it's executing using AI is restricted or slightly more difficult. So it's about the connectivity of of all the different systems, but it's about the foundational data strategy and it's about identifying the right tools for the right job and the real reasons why I'm focusing on the ROI because it gets lost sometimes when everyone's so excited about doing something new. And, you know, we need to make sure there's real tangible value. Yeah. I mean, you mentioned that everyone's so enthusiastic and excited about using AI. Is one of the challenges, finding the right talent and someone who blends that creativity with the technological knowhow? Yeah, it's definitely a different talent profile than we might have been looking for ten years ago. But I think it is about finding that blend because it is about the marketing know how, the knowhow about our business and then the curiosity to pursue new ways to to do that. I think one without the other all of the new without the context of why are we doing this? What does it mean to our business? Is not as valuable. Yeah. Yeah, I absolutely agree. And I think one of the things that is not a challenge, but an opportunity is to make sure that they, the marketers or the people around the business are embracing how the AI works and upskilling and doing the prompt training and making sure you know how to use it. Because it's very easy to say, "Well, we'll just get our agencies to use it", which is, you know, completely valid and that's fine. But what we want to do is ingrain the use of AI in the operations of how you do your job. So it's not this separate thing - "Well, we'll get AI to do that." Actually, it's"I'm using that part of AI as how I'm operating" and for that to happen, they need to upskill, they need to understand prompt training, of course, is so, so important as well as just understanding that the the whys and the ROI. Because I mean, alongside this explosion of AI tools and we've also seen a real growth around regulation as well, particularly recently we've had the EU AI Act coming into force. So what are you doing to make sure that your employees are using AI correctly? Aaron? And so is a really good question because the answer is something which isn't usually liked or appreciated by most and it's governance. It's the horrible governance which gets put in place, which people think is restrictive and no one likes. But it's really important that, as I said earlier, you've got that data strategy in place. But the layer above that is having a what we refer to as AI steering group, our AI governance. And what that means is it's not you can't do things, it's there's the right policies, procedures, there's right people there to support and doing it. But it's also building that list of vetted tools which have, you know, we know where the data's going, we know what the security controls are. And then it's about enabling people. And for that to be successful for us, we've needed to involve people around the whole organization. This isn't it's not an IT or a Digital or a Legal or Security or or a Marketing or HR Team. It's that having all of them around the table because for us to appropriately govern, there's so many different expertise actually around the table needed to make sure that we've ticked all the boxes but also we've done the right thing and we've got these, those these right tools. So we'veset up a governance structure with policies and processes, which again, sounds boring, but it's not because it's AI and it's really cool. So it's, so it works and I think that's it, I think it's important to try and, you know, have that agility for the future, even though I know governance doesn't sound agile, but it can be. Yeah AI governance can be cool. Susan. Yeah. We have set up cool AI governance as well. So much the same things. I think the thing I'd add is we started with sort of principles of how we believe we can ethically, legally use AI. And those governing principles, I think added with some more specifics, really help people to get into the right mindset to start the journey because it is complicated. And, you know, the regulatory requirements of not just knowing what you're doing, but understanding the models underneath it all – it's a lot. So we have, like you, brought together cross-functional teams to try to help people think through all the issues because, you know, we want to encourage that experimentation, we want to encourage people to find new uses, but we want to do it within the bounds of of our responsibility. And I think the education piece around it as well, because I think the intent is always really good. That, you know, marketeers in particular will, you know, they'll start to use a AI new tool and it will give really good results, which is amazing. But I mean, I didn't think originally where does that, where does that go? I just upload loads of creatives from our brand teams and they're going to help us generate some maybe some really cool, interesting content, but who owns the IP on that once it enters that system? Who else has got access to that? So it's that that whole governance and that education piece, actually, it's not because we you can't do it. We just want to make sure that either in it in the right way. Yeah. And Susan, you just touched on experimentation and finding new uses. I'd love to know what's exciting you about potential uses of AI within Diageo for the future. Yeah, I think the really exciting place is, is how we augment consumer experience. And, you know, our mission is to help people celebrate every day and we have great brands to do that. So connecting that mission to our brands and consumers is really exciting and we are thinking about what are the problems that consumers tell us they're having and how can we play a part in that? Whether it's, you know, organizing a party, you know, creating a more memorable experience. These are all areas that we're trying to hone in on, the consumer wants and needs and really develop new ways to solve those problems. Have you got anything coming up that you can tell us about? We have piloted and experimented with something that is a party planning tool, an at home party planning tool. As an example, you could put in prompts about this is a real prompt for someone, I didn't make this up off the top of my head. It was back when Barbie was the popular movie – "I would like a punk rock Barbie party." And it comes back with suggestions. Now, to me, the interesting thing was what it came back with, you could see how it got there. But it was really interesting as the suggestions came forward and then you continued to prompt it where it could lead you and how you could customize it. And in understanding like, "oh wow, why did it recommend that?" it actually led you to think about different ways to, you know, really achieve what you want. It was really interesting. So it's actually sparking further innovation? Yeah, absolutely. Aaron? And so for us and well, for me personally, there's a few there's a few really interesting areas. And the first one is being able to really analyze external information at such a scale that it can provide us with more insights as an organization, as a digital function, as a brand, as a brand functions to really understand the external knowledge and insights. So it's like CMI on steroids? Like, how did how do we how do we find out more so that we can better understand what the external conditions are and how we can tackle and support them and and how can we provide the right services and stuff. And then in turn, and then the more internal facing aspect of it is "how do we give more services to a customer?" as I mentioned earlier, around like hyper personalization. And I think that goes to even from ads and creation of content for ads like how do you create on a scale for 1 to 1 for every single person in this room to have a tailored piece of content, which is actually more attractive for you in terms of conversion, but relaying the correct message but you feel like this this is really for me and not just a segment of 20 or 20 million people or 20,000 people. And then the other area is these insights of peer to peer understanding and stuff. And I've spoken about this previously and in other talks about this knowledge of "how can our customer grow and how can we provide that 1 to 1 experience that we give from a business perspective we have set we have the most amazing (I'm biased) the best sales reps in the world, but they and they will spend so much time with customers, really understanding them. We have some sales reps who are now being with the same customer for so long that they go to., they, you know, they're invited to their children's weddings and stuff – these are real relationships. But how do we scale that same level of love, care and that essence of being part of the Wella family, as we call it, how do we scale that? And we can't do that without AI. So it's about how do we how do we really relay our culture and all our values, but on a more scale, because unfortunately we don't have a million employees to go around and spend time with everyone. But how do we how do we do that? Yeah. So that's both the opportunity and the challenge in many ways. Yeah. An interesting question for both of you, given how AI can really spark creativity and innovation and can take away some of those heavy admin tasks. Would you let AI write your brand's manifesto. Susan? I don't think we would. In thinking through how we use AI, I think we have a commitment to there are some things that we need to own, that need to be personal, that need to be created by our humans. And that feels like one of them. That feels like it goes a bit too far. Now we might use AI to get insights at scale to inform what that manifesto should be. But I think the actual creation act in this case should be people. Aaron? And for me whilst I completely, I completely agree. I did get AI to do it. Now, just to be clear, this isn't this isn't this is not the official company one. This isn't, and as a as a company, no, we would – like I agree with you – we would not get AI to build the values of our company. But what what I would do is and I think you mentioned this earlier, is you can use AI to spark information. You can use AI to support you in what that could look like and stuff of that. But I cannot and will not take away the amazing talent of my corporate, of the corporate agency and the amazing brand teams we've got. And they're not going to be replaced, the creativity that they've got is not going to be replaced. But in this instance, yes, I did ask it. Yes, so can we have a listen? Yeah. Yeah. So this is my Achilles heel... Does it say cool governance at any point? It was using our internal, fully governed AI tools. And but yeah, this is actually a childhood fear of mine to read in front of of people so this is a big milestone for me. Wella Company embodies a rich heritage of innovation and excellence in the beauty industry. The brand values, creativity, quality and sustainability, striving to empower individuals to express their unique beauty. Wella's personality is vibrant, sophisticated and inclusive, reflecting a commitment to diversity and self-expression with a focus on cutting edge technology and professional expertise. Wella inspires confidence and transformation, making every customer feel valued and beautiful. The brand's dedication to education and community support further enhances its reputation as a trusted and forward-thinking leader in the world of beauty. Pretty cool! It's, I thought it was actually quite good. Now, I have to say, we know that AI tools do crawl the internet and they do pull information from different areas. So I would actually say it's probably a summarized version of what we have on our corporate site. But that being said, it's not the same. It has amended it, it has shortened it and some of that. So do I think it's completely creativity there? No, I don't think it is. Yeah. But nonetheless it's a it was pretty good. Yeah. Brilliant, thank you for reading that out and any questions from the audience. So I run an independent creative agency and I would just love to know given on the conversation we just had, how has it changed what you are looking for from your creative partners? I think it's only sharpened what we need, which is amazing ideas and and true creativity that can really help us develop platforms for our brands. I will say I think some of the tasks will change of, you know, production tasks. We have, you know, divisions in every country in the world and we need to provide enough content for all of them. So we are going to use am I to adapt and trans create? So that is a function that I think could probably happen through AI, although with always having people involved. But I, I think it's almost sharpened what we need from agencies given the last thing you'd want is given it crawls the internet, everything would start to look the same. So I think if you don't have that originating creative, it just becomes all bland. Yeah, I agree. And where we've used is actually, we've used AI to help brief our creative agencies a lot stronger. And because there's one thing during the whole document of the brief but to really have a visualization which isn't complete but it helps a lot of the more visual and creative people actually really start to interpret what what we're asking for which I thought was an interesting way to actually use it. So we didn't use for the end result, we actually used it to try and communicate what we wanted more. And then similar to what Susan said, it's we're seeing the production element and the the transcreation, the reformatting of the creatives. And I think that's the bit which do we need to go to an agency and say can you convert this into like 6 or 7 different formats? Can you make sure that is optimized for, for mobile? Can you, can you make sure it's optimized for all of our E commerce and all media? I think I can see that taking a shift as well. And like Susan said, I don't think the creativity will be gone because we still need that that creativity. But these admin or operational bits, I think if we use AI to remove that, we can spend more time on the creativity. And I think that's that exciting because the more time we've got for creativity and thinking, who knows what we'll come up with. And I think that's how we see it changing with third parties. As a consumer psychologist, I will always look at what customers actually do. And the challenge with AI when we look at all the research is the customers will tell you that they really like the products that are co-created or driven by AI and that they will value it higher. But when you actually look at their behaviors or decisions, the product perception is actually devalued considerably and the conversion is reduced mostly due to reduced trustworthiness and emotional value. How do you manage that mixed mindset around the consumers, around AI and do you therefore choose to occasionally not communicate that AI has been used? First, we're always transparent about when we use it. Full stop. I think what we've tried to lean into is less about what consumers might say and more about what they do. And what AI's allowed us to do is harness the actual actions they're taking via search, via e-commerce, etc., etc. to understand that pathway that they're taking to interact with our brands. So I guess in summary, we tend to look at action versus only what they tell us. And I think that serves as a pretty good guide. Yeah and just I mean, the only things for me to add to that is I think authenticity is, is super important. And with the AI tools which exist now, there's no reason why, you know, a lot of companies could end up creating the same perception or the same types of creativity. And for us, we're a beauty company. and that authenticity of the output is so important. So if we're going to be using AI for anything like Susan said, there's no question you have to be fully transparent. But back to the authenticity, if we're selling something or providing something to you, it doesn't matter if we've got AI to create the best like a sales creative – if your hair doesn't look absolutely incredible when after you've used the products, then we've failed. And I think that's super, super important is the authenticity of the value that the customer or consumer gets at the end of the day. I have a question about the smart recommendation of drinks or beauty products. Do you think there would be a risk of narrowing the selection, therefore the brand's offering with providing highly precise, personalized offers? Because like when we were all buying CDs and listening to the entire album, sometimes I feel it helps you broaden your taste by listening to the sounds that you wouldn't proactively handpick for your own playlist, but by offering to customers exactly what they know they want, do you think it might affect the potential discovery of unexpected new favorites? Great question. It's a really good question. I think for now, what we're focused on is almost enhancing that discovery. Because, you know, for any occasion there are so many great drinks that you could have. And our portfolio is amazing and broad. So I think there's a lot of land to cover just in helping people discover what's right for the thing they're trying to do right now, which could be different from the thing they're doing tomorrow and the occasion they're in there. What we've seen so far is it actually leads to more discovery because it brings up things that people never would have thought of. And then there's a journey from there of like, "Oh that's really interesting" recommendation. What about this? What about that? And the prompting that AI involves allows for that. So I mean, for now, no, I haven't seen that is a big issue. But I think it is something to to consider as we move forward. Yeah, I'd agree. And there is an element of going way too far on some of these things, right? But I would say for us, being able to provide the insights on what we think you may, which may trigger a question, which may trigger a thought. And I think when it comes to style and things like and just to be clear, I am not an expert in fashion or beauty or anything just by working for Wella it's a different department. But but to be able to inspire and recommend and try to have something relevant for that person, even if it's wrong. Has it has it triggered a question? Have they considered something that they haven't before? And for us, because we want everyone to feel their best selves, it's we can't we can't make those recommendations unless we we really try to understand you as an individual. Is it going to be right all time? No. Great thing about AI is it's going to learn over time and maybe we'll get better. But in regards to the hyper personalization, I think this has been done by banks for that for many times, and that's offering a service which is best for you. And in their situation, it's like loans and stuff like that. But like for us, it's how do we make sure the trade terms possibly right, and things like that are actually built for you as a as a business because your business isn't exactly the same as someone else's. And when you've got hundreds of thousands of businesses that you're working with, it's not always easy to to do that.